Atomized Incorporation: Chinese Workers in the Aftermath of China's Rise 1009161202, 9781009161206

Citation preview

Atomized Incorporation Atomized Incorporation examines why the Chinese regime selectively tolerates workers’ collective action within single factories and what this means for the country’s long-term political resilience. It investigates the implications of state–labor relations in contemporary China and suggests that it has evolved away from overt coercion to limited incorporation. Based on two years of in-depth fieldwork, Rho uncovers how ordinary workers think, believe, and behave in this changing socio-political environment. She demonstrates that labor grievances have become more politicized and finds that the current approach to economic grievance resolutions demobilizes the emergence of labor movements by rewarding those with collective action resources within individual workplaces. Rho argues that though this limited state of incorporation allows workers to express discontent at wages and working conditions, it also denies them the opportunity to make claims about structural problems and does not effectively enhance political loyalty in the long run. Sungmin Rho is Associate Professor of International Relations and Political Science at The Graduate Institute of International and Development Studies. She also conducts policy-relevant research by collaborating with international organizations such as International Labor Organization (ILO).

Atomized Incorporation Chinese Workers and the Aftermath of China’s Rise

SUNGMIN RHO The Graduate Institute of International and Development Studies

Shaftesbury Road, Cambridge C B2 8E A, United Kingdom One Liberty Plaza, 20th Floor, New York, N Y 10006, U S A 477 Williamstown Road, Port Melbourne, V I C 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467 Cambridge University Press is part of Cambridge University Press & Assessment, a department of the University of Cambridge. We share the University’s mission to contribute to society through the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781009161206 D O I:

10.1017/9781009161183

© Sungmin Rho 2023 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press & Assessment. First published 2023 A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data N A M E S : Rho, Sungmin, 1984– author. T I T L E : Atomized incorporation : chinese workers and the aftermath of china’s rise / Sungmin Rho. D E S C R I P T I O N : 1 Edition. | New York, N Y : Cambridge University Press, 2023. | Includes bibliographical references and index. I D E N T I F I E R S : L C C N 2022048466 | I S B N 9781009161206 (hardback) | I S B N 9781009161183 (ebook) S U B J E C T S : L C S H : Industrial policy – China. | Labor disputes – China. | Industrial relations – China. | Working class – Political activity – China. C L A S S I F I C AT I O N : L C C H D 3616. C 62 R 46 2023 | D D C 331.80951–dc23/eng/20230224 L C record available at https://lccn.loc.gov/2022048466 ISBN

978-1-009-16120-6 Hardback

Cambridge University Press & Assessment has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Contents

List of Figures

page vii

List of Tables Acknowledgment

ix x

1

1 3 6 12 16 19 21

Introduction 1.1 1.2 1.3 1.4 1.5 1.6

2

The Political Costs of Labor Coercion: The Changing Socioeconomic Environment since the 2000s 2.1 2.2 2.3 2.4

3

Central Arguments Atomized Incorporation Untangling Grievances Particularistic Bargaining Notes on Research Methodology Structure of the Book

Political Economy of Labor Coercion Labor Coercion in Post-reform China Declining Effectiveness of Labor Coercion Socioeconomic Conditions for Labor Assertiveness: Long-Term Consequences of Labor Coercion

Atomized Incorporation: Regime Response to the Changing Environment 3.1 3.2 3.3 3.4 3.5

China’s Labor Control Strategy in a Comparative Perspective Toward Atomized Incorporation Atomized Incorporation in Post-reform China Political Economy of Atomized Incorporation Conclusion

v

24 25 28 36 43 46 47 51 62 76 93

Contents

vi

4

Politicization of Labor Discontent and Blame Attribution 4.1 4.2 4.3 4.4

5

Workplace Mobilization and Collective Action 5.1 5.2 5.3 5.4

6

The Rise of Interest-Based Collective Action Firm-Level Patterns of Labor Protests Characteristics of Interest-Based Collective Action Firm-Level Patterns of Blame Attribution Conclusion

Discursive Opportunities and Collective Action at Law-Abiding Firms 7.1 7.2 7.3 7.4

8

Barriers to Collective Action for Unorganized Workers Protest Initiators in Chinese Factories Understanding Heterogeneous Protest Intention Conclusion

Interest-Based Collective Action and Firm-Level Patterns of Labor Protests 6.1 6.2 6.3 6.4 6.5

7

Debate about Migrant Labor Discontent Politicization, Atomization, and Blame Attribution Patterns of Blame Attribution Conclusion

Media Reports of Labor Protests Perception of Media Exposure at Different Firms Cases of Labor Protests Broader Implications

State–Labor Relations in the Xi Era and Beyond 8.1 Atomized Incorporation in the Xi Era 8.2 Beyond Protests

95 97 98 105 122 124 125 128 135 159 161 163 166 177 181 184 186 187 193 198 211 214 216 227

Appendix I: Survey Dataset References

231 235

Index

257

Figures

2.1 Number of migrants page 33 2.2 The ratio of available jobs to job seekers 37 3.1 The number of mentions of “peasant worker” in the title of People’s Daily 64 3.2 Profits of select industries in PRD 82 3.3 Value-added taxes of select industries in PRD 82 4.1 Reported perception of blame attribution: Who is to blame? 106 4.2 Reported perception of blame attribution by job position 108 4.3 Effect of economic and social grievances on blame attribution: Line supervisors 117 4.4 Effect of economic and social grievances on blame attribution: Rank-and-file workers 117 4.5 Effect of economic and social grievances on blame attribution: Alternative independent variable 119 4.6 The effect of protest behavior and information on blame attribution 121 5.1 Predicted effect on perception of better outside option 142 5.2 Protest intention by blame attribution and job position 152 5.3 Dependent variable: Have you participated in collective action? 157 5.4 Predicted effect of past participation on protest intention and determinants 158 6.1 Percentage of demands by year 173 6.2 Effect of extra-legal demands on protest tactics 178 6.3 Effects of protest demands and size on police presence 180 6.4 Effect of law score on blame attribution 183 7.1 Effect of law score on perceived group efficacy 194

vii

viii

List of Figures

7.2 Marginal effect of expected media exposure on perceived group efficacy 7.3 Marginal effect of law score on expected media exposure by job position 7.4 Does your firm produce internationally renowned brands? 8.1 Cases of labor protests by ownership 8.2 Cases of small-scale protests 8.3 Cases of labor protests by tactic A.1 Geographic coverage of survey firms

195 197 198 217 218 219 233

Tables

2.1 4.1 4.2 4.3 5.1 5.2 5.3 5.4 5.5 5.6 5.7 6.1 6.2 6.3 6.4 6.5 6.6 6.7 7.1

Percentage of rural workers in nonagricultural sectors page 39 Perceptions of upward mobility, inequality, and fairness 107 Multivariate analysis of blame attribution 109 Multivariate analysis of social grievances 113 I share the same interests with [. . . ] 131 Are you willing to protest for a pay increase? 136 Sources of preferences for collective action (% agree) 138 Demographic characteristics of workers by job position 141 Determinants of protest intention 148 Effect of blame attribution on protest intention 153 Why did you not join the collective action? 155 Firm-level determinants of labor protests 169 Determinants of strike incidence at the firm level 171 Ownership and legal/extra-legal demands (% of protests) 174 Firm size and legal/extra-legal mobilization (% of protests) 175 Multivariate analysis of protest demands 176 Protest tactics 177 Determinants of police presence 179 Effects of firm ownership and size on workers’ perceived chance of media attention 196 A.1 Basic firm-level characteristics by ownership 234

ix

Acknowledgment

When I began working on my doctoral dissertation, I never imagined that it would take nearly a decade to publish this book. This book reflects my individual journey of striving to comprehend who we are and why we are here, as much as my academic pursuit as a learner and researcher. I was blessed to have tremendous amounts of support, love, and encouragement throughout this journey. I owe my utmost appreciation to my PhD advisor at Stanford University, Jean Oi. My drive and determination alone would have been inadequate to finish my dissertation, let alone this book! Jean’s constant enthusiasm for my research inspired me to turn my dissertation into this book. I sometimes feared her emails inquiring about the progress of my work, but she always gave me a safe environment when I felt disoriented and alone. I would also like to extend my gratitude to Michael Tomz for his unceasing guidance. When I was frantically running around the factories for my first survey and was teetering on the edge of a mental breakdown, I could go on because of his assurance and support. He met with me online even at a very late hour in California, examined the preliminary data with me, and assured me that I was not wasting my life. During my time at Stanford, I was fortunate to work with Professor Beatriz Magaloni, whose feedback was indispensable in building my theory and in situating my work in a comparative context. I am indebted to her advice and kindness. Professor Lisa Blaydes was always there to provide me with encouragement even after I moved to Geneva. Her advice about book publishing made the publication of this book possible. My fieldwork in China is truly unforgettable and filled with gratitude and appreciation. First and foremost, I am deeply indebted to the x

Acknowledgment

xi

interviewees who took the time to answer my questions and share their experiences with me. Any errors are solely my responsibility and I hope this project lives up to their kindness. Sung-hwa Jung at KOTRA went out of his way to help me. He assisted me not only in connecting with interviewees but also in exploring the new city. Young-hak Yim took care of me like his own daughter. He acquainted me with many of his peers and friends in China and expressed genuine support, which is still deeply ingrained in my heart. I would like to express my gratitude to the scholars that I met in China, including Ngok Kinglun, Wan Xiangdong, He Gaochao, Pun Ngai, Ellen David Friedman, and others. I was fortunate to receive their generosity and be part of stimulating discussions. My fieldwork days were often tiring and draining, but I always found comfort in seeing my friends. I am thankful for all the friends I crossed paths with in China. This project was made possible by the generosity of numerous organizations. The Korea Foundation for Advanced Studies gave me financial assistance during my Ph.D. studies, and I could dedicate myself to research. The fellowships and grants provided by the Freeman Spogli Institute for International Studies, the Center for International Security and Cooperation, and the Stanford Center at Peking University helped me to fund my fieldwork. I am grateful to the Walter H. Shorenstein Asia-Pacific Research Center at Stanford University and the Center for the Study of Contemporary China at the University of Pennsylvania for the predoctoral and postdoctoral fellowship. Through their generous support, I was able to devote my time to writing and be immersed in an intellectually stimulating environment. Without the encouragement of my colleagues and friends, it would not be possible for me to complete this long journey. I thank my friends at Stanford University and the University of Pennsylvania. My friends in Geneva have accompanied me through the exhausting days of writing this book and have always supported me. My colleagues at the Graduate Institute of International and Development Studies gave me constructive feedback and showed their trust. I feel truly blessed to be part of a community that is both supportive and inspiring. Many people read different versions of this book and helped me strengthen my arguments. I am grateful to the audiences at Stanford University, University of Pennsylvania, University of Bern, University of Geneva, and International Labor Organization. Didi Egerton-Warburton gave excellent editing that significantly improved the book’s writing. I would also like to extend special thanks to the book workshop participants, Mary Gallagher, Xiaobo Lü, Peter

xii

Acknowledgment

Lorentzen, and Jean Oi. They read the entire manuscript and provided detailed comments even amid the chaos of the COVID- 19 pandemic. I thank Robert Dreesden at Cambridge University Press for supporting this project as well as three anonymous reviewers for offering detailed and thoughtful feedback. Robert Judkins and Sunantha Ramamoorthy provided me with assistance and instruction in the editing and production of the book. Finally, I would like to thank my family. My sister Seunghwa Rho was an incredible source of moral support, and I could always count on her, even when I had no clue about statistical models. My parents, Booho Rho and Hyojoo Yoh, gave me a safe place to come back to. I embarked on this journey because I was curious about a different society and way of life. This curiosity also led me to obstacles and challenges – many of them unexpected – but I was able to take risks and make mistakes because I knew they were always there for me. Through them, I came to understand a life of love, humor, and devotion. I dedicate this book to them. Sungmin Rho

1 Introduction

In 2011, there were labor strikes at more than sixty automobile factories across Guangzhou in China. While there was no evidence of cross-firm coordination, it was clear that the high-profile strike at Honda Nanhai in Foshan inspired a wave of unrest across the region. In response to the protests, however, state authorities displayed surprisingly lenient attitudes. The People’s Daily, the mouthpiece of the Communist Party of China (CCP), commented that during the incidents “both labor and management did not display any extreme or irrational behavior,” signaling that the government did not perceive the protests to be drastic.1 According to Chen Wei-guang, the then chair of the Guangzhou Federation of Trade Unions, the strike wave was not met with state violence because workers stayed away from politics and only sought job-specific economic gains: [W]e believe that the demands of the workers are justified. But we hope that such economic disputes do not develop into political incidents and will not disrupt social order. This is our bottom line. [. . . ] We have to say that our strikes have been very orderly, there were no walkouts from the factories to the streets, no destruction of machinery, no playing of politics. Everything remained in the framework of disputes within factories.2

To a lay observer, the Chinese regime’s tolerant response to the protests might be surprising given its authoritarian nature. Yet in China and other authoritarian states, such as Vietnam and Egypt, political leaders have permitted limited forms of popular protest while repressing politically threatening mobilization. As Chen Wei-guang’s remarks show, there is a 1 2

Li, 2010b. Lüthje et al., 2010.

1

Introduction

2

critical distinction between economic and political contention that has been central to defining the Chinese regime’s response to workers’ collective action. While the regime tolerates what it defines as “economic” protests within individual workplaces, it punishes those that breach the physical boundary of the firm. This distinction has created silos for workers’ collective action, because economic protests must be contained within individual firms. Labor protests in China have thus been characterized as “cellular activism.”3 Despite the prevalence of labor unrest in China, it has generally been assumed that the regime is resilient to worker discontent. Most labor protests seek job-specific economic gains, such as higher wages and better working conditions, and exhibit little political activism. With severe restrictions on independent labor organizing inside and outside the firm, workers have limited resources to generate and sustain a labor movement. A growing body of literature suggests that these relatively frequent, but apolitical, protests could be a sign of the regime’s political resilience rather than of its instability.4 An apolitical protest allows the regime to collect information on the location of grievances and gives workers space to air their discontent: This combination of information gathering and de-escalation undermines revolutionary threat from below. Apolitical protests are assumed to function as a “safety valve” or a co-optative device, by rewarding disgruntled workers with material benefits. Yet we still have a limited understanding of what these apolitical protests actually mean for the Chinese regime’s political resilience in the long run. The prevalence of apolitical labor protests is puzzling, given the many historical examples of collective action of disgruntled workers becoming politicized and morphing into a pro-democracy movement.5 From the early episodes of democratic transition in western Europe and Latin America to more recent waves in Asia, rapid economic changes such as industrialization and globalization have birthed an industrial labor force that is both interested in and capable of promoting democratic transition.6 The pro-democracy labor activism in these regimes also started out as small-scale economic protests about wages and working conditions, but these protests failed to deradicalize labor discontent. Even though independent union activities were repressed, workers in many 3 4 5 6

Lee, 2007a. Lorentzen, 2013; Dimitrov, 2013; Lorentzen, 2017; Li, 2018. Rueschemeyer et al., 1992; Collier and Mahoney, 1997; Wood, 2000; Koo, 2001; Beinin, 2009; Dahlum et al., 2019. Rueschemeyer et al., 1992; Koo, 2001; Silver, 2003; Seidman, 1994.

1.1 Central Arguments

3

authoritarian regimes developed strong antiregime identities and transitioned the job-specific economic protests into militant labor activism. Has the regime’s policy of tolerating isolated economic protests within workplaces enhanced workers’ overall political loyalty? Do protests that only focus on economic issues imply that workers do not link their everyday grievances to the regime? Why is it that Chinese workers appear to be empowered enough to stage relatively frequent protests, but rarely turn their collective action toward a social or political goal? In this book, I look inside China’s factories and examine workers’ thoughts and behavior to uncover the political implications of state–labor relations in contemporary China. 1.1 CENTRAL ARGUMENTS

My central argument is that the Chinese regime has been successful at demobilizing the labor movement but has not been able to depoliticize labor discontent. I offer a theoretical framing of the CCP’s approach to state–labor relations as atomized incorporation and show that, by allowing workers to make job-specific economic claims but denying their political voice as citizens, the regime has only incorporated workers in a limited way. This limited state of incorporation means that protests only express small economic issues and offer a channel to materially co-opt those individuals and groups who are capable of demanding economic gains. It, however, also means that workers’ social and political grievances are not tempered in a lasting way. The protests grounded in narrow economic issues reduce the immediate threat of broad mobilization, but labor discontent in China is increasingly politicized and in the long term could challenge the stability of the regime. In the “economic” realm that is defined by the state, workers appear to have been permitted substantial leverage as economic agents. They are able to use their market power to target their employers and extract concessions. The state’s leniency toward the more aggressive style of protest that proliferated in the early 2010s, as seen at the Honda protest, revealed the regime’s toleration of workers that use the market to discipline their employers. Workers’ demands often exceeded the legal minimum conditions but remained apolitical; most of the demands concerned job-specific economic targets such as wage increases or wage arrears. The argument that the Chinese regime undermines social organization and individualizes contentious behavior is not new.7 Scholars who study 7

White et al., 1996; Goldman and MacFarquhar, 1999.

Introduction

4

popular contention in China have emphasized citizens’ atomized state as a barrier to broad mobilization. It has been argued that the regime has succeeded at breaking collective grievances down into individual ones by deploying a wide range of tools such as law, state-controlled union, and nongovernment labor organization.8 Even when citizens engage in collective action, their actions have been characterized as “cellular,” “fragmented,” and “parochial.”9 It has also been stressed that in order for popular protests to remain contained, the regime has to accommodate the demands of protesters to a certain extent.10 In other words, the incorporating element enables atomization. What has been less explored is the inherent challenge in trying to achieve these two apparently conflicting goals: incorporation and atomization. How effective is the incorporation if it should accompany atomization? What is the nature of incorporation that enables the atomization of workers, and can this type of incorporation appease workers’ discontent? In this book, I focus on theoretical and empirical investigations of this contradiction. In doing so, I highlight the importance of connecting macrolevel theories to microlevel analyses. In macrolevel theories, the prevalence of apolitical contention has been understood as a sign of the regime’s success in maintaining political resilience. These analyses offer useful insights about the advantages of allowing small-scale contention. In this book, however, I also show that the underlying assumptions in the existing arguments need to be examined at the microlevel, from the workers’ perspective. On the basis of two years of in-depth fieldwork, I study and describe how different workers behave and think about their firm, society, and government in this evolving economic and sociopolitical environment. A close look at the dynamics of collective action within and across firms shows that only a fraction of workers have resources for collective action, such as social networks, the support of external actors, and intelligence about the horizontal and vertical organization of the firm. My analyses demonstrate that atomized incorporation has been successful at reducing the immediate threat of broader mobilization because it allows those with such resources to make claims and receive concessions. Material co-optation of those who are capable of initiating collective action could be important given prevalent grievances about the central government and broader society among migrant workers that I find. 8 9 10

Lee, 2007a; Su and He, 2010; Friedman, 2014; Gallagher, 2017; Chen and Gallagher, 2018; Fu, 2018. Perry, 2002; Lee, 2007a; Chen, 2018. Lorentzen, 2017; Chen, 2018.

1.1 Central Arguments

5

Yet I also show that the absence of political claims in worker protests is not necessarily a sign that the regime has succeeded in appeasing workers. Many workers who have political grievances do not use protests as a channel to express their demands. The majority of workers who lack the resources for collective action are even more discontented with their position in the broader social structure and frustrated with the central government. The concerns of the majority, however, are not contested, because atomized incorporation does not provide a platform for these workers. This suggests that many workers with substantial discontent are left out. To understand the implications of atomized incorporation, therefore, I argue that it is also important to understand what is not currently contested in labor protests. The distinction between economics and politics is arbitrary in the sense that workers’ economic struggles are a product of the political system and national policymaking.11 Low-wage workers’ economic circumstances in contemporary China are in part due to a legacy of institutional discrimination. Rural-to-urban migrant workers who have been the backbone of the low-skilled labor force in manufacturing and service industries have not enjoyed equal access to residency rights and public goods. The household registration (hukou) system in China segregates citizens according to rural or urban origins, as well as by local and nonlocal registration, and determines eligibility for public services that are fundamental to quality of life, such as housing, medical care, and schooling. In practice, this has meant that unskilled rural workers who move to urban areas to work in rapidly industrializing centers are denied access to the benefits of the modernizing economy that their labor makes possible. Even though the institutional elements of discrimination have disappeared over time, its social legacies persist and continue to undermine migrant workers’ career progression, life chances, and integration into urban society. By treating economics and politics as distinct, however, the regime has framed workers’ everyday struggles as the discontent of a few “bad apples” than a systemic flaw. While this framing might have delayed the formation of workers’ critical consciousness about the structural origins of their plight, I find that many workers do eventually make this connection. Because the hukou-based restrictions make it difficult for migrant workers to integrate into urban communities, they are likely to have high levels of labor mobility. This mobility undermines the communitybuilding that would support sustainable labor activism, but at the same time, it has exposed workers to the idea that their situation is the product 11

Wood, 1995.

Introduction

6

of a system, maintained by the central government, that prevents them from fully participating in the economic and social affluence enjoyed by the urban population. The grievances that are targeted at the central government might not pose an immediate threat to the regime. Yet the widespread discontent suggests that the regime’s capacity to monitor and punish defectors will become increasingly important. As many historical cases of prodemocracy movements demonstrate, labor grievances are easily fed into alternative platforms of antiregime mobilization. Due to the labor force’s size and capacity to disrupt production, it is often perceived by antiregime activists as an attractive group to mobilize. This means that as labor grievances increase but are not channeled into available platforms of claim-making, it could become ever more necessary for the regime to isolate workers from each other and from other groups and suppress potential insurgencies. I suggest that the Xi regime’s seemingly outsize response to recent cases of labor mobilization reflects these concerns. At the same time, even when disgruntled workers do not directly join antiregime activities, there are other more subtle forms of noncooperation that workers can engage in. These actions are more difficult to police and could pose a challenge for state–labor relations, which I elaborate in Chapter 8. 1.2 ATOMIZED INCORPORATION

1.2.1 The Setting: Political Costs of Labor Coercion China’s rise in the world economy has been touted as a model case of economic development. Scholars have argued that economic success is a major factor in the CCP’s ability to maintain authoritarian rule concurrently with rapid economic growth. China’s path of economic reform has been characterized as “reform without losers” and postulated to have provided “performance legitimacy” for the regime by delivering a better quality of life for its citizens. Contrary to the conventional assumption that growth creates a middle class that desires political freedom, many argue that Chinese citizens perceive that the CCP is the facilitator of economic development and have “accepted authoritarianism.”12 This book, however, commences with the observation that China’s rise also created socioeconomic conditions that could lead to conflict in state–labor relations. China’s economic transformation was driven by rapid industrialization that lured large numbers of migrant workers 12

Wright, 2010.

1.2 Atomized Incorporation

7

from rural China to select geographical regions. Economic reform clearly created losers, and one of the most disadvantaged groups was migrant laborers. In the 1980s and 1990s, when the regime promoted foreign direct investment and export-oriented industrialization, workers’ wellbeing was largely ignored. It is well documented that the rural population has not experienced the same benefits of economic growth that the urban population has enjoyed. Rural-to-urban migrant workers, in particular, have been the backbone of the growth of the urban economy, but they experience social, economic, and political marginalization in urban areas.13 Many factors could have contributed to the extent and nature of marginalization, but it is clear that the Chinese regime’s institutionalized exclusion of rural migrants has played a critical role. The hukou system was introduced as part of the planned economy as a means of mobility control, but it persisted in the postreform era. In the 1980s, rural residents were permitted to leave their farmland for urban jobs, but they were denied permanent residency rights and access to the welfare benefits available to urban residents. Similar to the apartheid system in South Africa, this semipass system helped labor-intensive sectors to minimize labor costs and accelerate capital accumulation14 ; at the same time, it contributed to the emergence of migrant workers’ collective identity as those who “dagong,” which can be translated as “work for someone else” but is used almost exclusively to indicate migrants’ low-skilled employment and marginalized status.15 The effects of structural discrimination persist, restricting the social and economic mobility of migrant workers and their children in urban areas. Migrant workers’ shared experience, Pun and Huilin (2010) find, “precipitated anger, frustration, and resentment conducive to the emergence of the workers’ consciousness.”16 While labor discontent grew, the success of export-oriented industrialization absorbed the surplus labor from the countryside and the labor market tightened. From the mid-2000s, labor-intensive sectors experienced acute labor shortages and high turnover rates. The more competitive labor market meant that bargaining power shifted in favor of migrant workers. Making use of their enhanced bargaining position, disgruntled workers began to express both passive and active resistance, 13 14

15 16

Chan and Senser, 1997; Solinger, 1999; Wang, 2005. See Wintrobe (1998) for a political economy analysis of the apartheid system that can also be applied to the Chinese hukou system. Alexander and Chan (2004) make a direct comparison of the two. Pun, 1999. Pun and Huilin, 2010, p. 496.

8

Introduction

from job hopping – “voting with feet” – to labor strikes and street demonstrations. Once considered a docile and cheap source of labor, migrants now make up the majority of workers who participate in protests.17 Another outcome of this tight labor market was that the efficiency and contribution of traditional labor-intensive industries to the Chinese economy declined. Migrant laborers’ wages, which had stagnated until the early 2000s, began to rise rapidly.18 The increased labor costs reduced the competitiveness and profitability of labor-intensive sectors and the regime responded by reorienting its growth model. The central government began to promote industrial upgrading, high-technology development, and domestic innovation. The socioeconomic conditions that contributed to growing labor assertiveness, therefore, are endogenous to the logic of the regime’s earlier development strategy. China pushed for exportled growth and adopted a coercive approach to migrant labor, but as a result it faces a changed environment for state–labor relations: There are widespread worker grievances, workers’ bargaining power in the labor market is increasing, and the two-digit growth rates that previously gave workers hope that the benefits of economic growth would be universally realized are no longer possible. How the Chinese regime manages state–labor relations in this critical period has important political implications. Many authoritarian regimes have faced pro-democracy labor movements when these socioeconomic conditions were present in the aftermath of rapid economic expansion: Labor militancy was “manufactured” in Brazil, industrial workers’ antiregime identities emerged in South Korea, and continued labor insurgency triggered a political negotiation that led to democratization in South Africa.19 Observing China’s economic transformation, Silver (2003) predicted a strong labor movement would emerge and argued that it is “also likely to play an important role in widening and 17

18 19

While official statistics on labor protests are not publicly available, various sources suggest that migrant labor has become the major group of contenders. In 2010, for example, migrant workers’ labor disputes comprised nearly seventy percent of all labor disputes in Beijing (“Migrants make up seventy percent of labor dispute cases (laodong zhengyi anjian nongmingong zhan qicheng)” 2011). A more recent analysis of protest incidents on social media finds that migrant workers have engaged in the largest number of protests in recent years (Goebel, 2019). There are also some patchy, but informative, data to show that migrant workers’ protests have increased in number. For example, Wen (2012)’s study indicates that migrant workers’ protests in Guangdong almost doubled between 2000 and 2004, reaching 4,008 cases in 2004. Feng (2008) shows that the number of migrant workers involved in collective action increased from around 160,000 in 2001 to more than 250,000 in 2005. Wang, 2010; Cai and Du, 2011. Seidman, 1994; Koo, 2001; Wood, 2000; Wintrobe, 1998.

1.2 Atomized Incorporation

9

deepening processes of democratization.”20 The structural origins of rising labor unrest suggest that the regime needs to adopt a fundamentally different approach to state–labor relations to achieve political stability in the long run. 1.2.2 Toward Atomized Incorporation How do authoritarian regimes dissipate labor discontent and ensure cooperative state–labor relations? I argue that in response to the changing socioeconomic environment, state–labor relations in China have evolved away from labor coercion to what I theorize as “atomized incorporation.” Atomized incorporation is a form of incorporation because it brings the interests of a formerly excluded group into policymaking and produces policy benefits for the group. There is incorporation in the form of channels of claim-making for the marginalized group, which offers a pathway for political participation. However, incorporation is atomized because the channels are only available to atomized subgroups and not to the collective. Atomized incorporation differs from the most commonly discussed strategies of labor control in authoritarian states – state corporatism and market-based repression – because it neither fully represses labor nor does it acknowledge its organized interests.21 Comparative theories on state–labor relations in authoritarian regimes propose a variety of tactics that authoritarian regimes adopt to enhance workers’ loyalty, but many have focused on the regimes’ control of labor unions. The seminal works discuss state corporatism as a model of labor control, which presumes organized labor.22 Ruling elites might use semidemocratic institutions such as parties and legislatures as a platform to negotiate labor demands and reduce labor unrest, but this also assumes that labor’s interests are organized and represented in an institutional framework.23 When there is little labor organizing activity, it is postulated that the regime represses labor.24 Yet, in a substantial number of nondemocracies, workers are unorganized but not always repressed.25 In theory, Chinese workers are organized and represented by the All China Federation of Trade Unions (ACFTU). Similar to corporatist 20 21 22 23 24 25

Silver, 2003, p. 73. Valenzuela, 1989. Valenzuela, 1989; Collier and Collier, 2002; Cohen, 1982. Kim and Gandhi, 2010. Valenzuela, 1989; Deyo, 1989. A few studies stress that we have a limited understanding of state–labor relations in nondemocratic regimes where labor is not organized. See, for instance, Robertson (2010).

10

Introduction

labor unions, the ACFTU has a monopoly on labor representation at the regional, industrial, and enterprise levels. However, the ACFTU has not been able to mediate state–labor relations in a meaningful way because it largely functions as an extended body of the state under the control of the CCP. Because it is subjugated to state interests and has limited resources, the ACFTU has not been able to function as an organization to represent and control workers. In particular, migrant workers were not acknowledged as part of the working class and were only recognized as an official constituency by the ACFTU in 2003. When labor disputes do arise, the ACFTU usually channels them back to legal and administrative bodies that undermine workers’ collective action.26 As a result, few workers consider the ACFTU to be a meaningful representative of their interests. While the party-led trade unions are not able to represent workers’ interests, the regime strictly limits autonomous labor organizations out of fear of workers becoming radicalized. There have been some attempts at independent labor organizing, but they did not survive state repression, particularly since the suppression of the Tiananmen Square democracy movement.27 Since 2003 the regime has instead adopted pro-labor policies that offer channels of claim-making to marginalized workers, as long as they remain unorganized. When an authoritarian regime declines to engage in centralized bargaining with organized groups, these claim-making channels can provide information about the groups’ demands, as long as the channels remain limited in scope.28 The Chinese regime’s tolerance toward protests for job-specific economic gains, which I call “atomized protests,” has been an important element of atomized incorporation, in addition to the establishment of the rule of law and channeling of labor grievances into committees and courts.29 The tolerated space of contention has been codified in informal rules with an implicit warning that those who cross

26 27

28 29

Chen, 2004; Chen and Gallagher, 2018. During the Tiananmen Square protest in 1989, some independent labor organizations such as Beijing Workers’ Autonomous Federation (BWAF) got involved (Walder and Gong, 1993). Their appearance in Tiananmen Square was supposedly an important factor leading to the leadership’s decision to crack down on the movement. Workers’ involvement fed into Deng’s fear that the Tiananmen protest could turn into a broader political movement such as the Solidarity Movement in Poland (Feigon, 1990; Perry, 1994). The Chinese leadership since then has maintained tight restrictions on independent labor organizing. Dimitrov, 2013; Lorentzen, 2017. Gallagher, 2017; Chen, 2018.

1.2 Atomized Incorporation

11

the boundary of acceptable protests will be repressed.30 While introducing pro-labor policies, the regime has maintained, and even intensified, strict restrictions on independent labor organizing and social mobilization across workplaces or regions. Atomized incorporation can be theorized as a cost-effective way for an authoritarian regime to manage a social group, when the group is not powerful enough to pose an immediate threat but is nonetheless important for securing long-term resilience.31 It might have something in common with redistributive schemes that authoritarian regimes use to materially co-opt disgruntled social groups, but it is more selective and flexible. Under atomized incorporation, there is bargaining between the state and workers, but in a particularistic way.32 As Gallagher (2017) points out about the legal labor disputes, the state has not been able to impose its labor policies from the top down but largely relied on workers’ complaints and protests to initiate monitoring and enforcement. Atomized incorporation can thus be better characterized as a fire alarm system rather than a police patrol.33 One advantage of relying on bottom-up claim-making, compared to centralized bargaining, might be that atomized incorporation can dispense resources more selectively and flexibly. Given that not everyone is motivated to, or capable of, engaging in popular contention, it requires a smaller commitment of resources from the regime. The regime also retains substantial flexibility in shaping the environment in which workers engage in claim-making. It can influence the quantity and nature of resources dispensed according to the changing conditions. This flexibility could be particularly important in state–labor relations, since the regime’s commitment to increased labor protection and welfare could 30 31

32

33

Perry, 2010; Li, 2018. It has been assumed that a social group needs to have sufficient political leverage to extract resources from the state (Bueono de Mesquita et al., 2005). An unorganized group often lacks such political leverage; yet recent studies demonstrate that authoritarian leaders adopt more comprehensive redistributive policies, even if the recipients are not politically powerful or well organized (Wallace, 2014; Albertus et al., 2018). China scholars also suggest that the Chinese regime has long been managing its relations with unorganized citizens, while undermining their collective capacity to use their voice (Stockmann and Gallagher, 2011; Howell, 2015; Gallagher, 2017). In other words, there is a wider range of state–society relations than commonly assumed even in the regimes that repress political activism. The term particularistic has been used to describe the private nature of goods negotiated in a political setting (Mayhew, 1974; Volden and Wiseman, 2007). The introduction of pro-labor laws such as the 2008 Labor Contract Law could be understood as a type of public good, since the law is supposed to impact all marginalized workers. McCubbins and Schwartz, 1984.

12

Introduction

impact state–capital relations. In fact, a large portion of the costs of atomized incorporation have been borne by individual firms, not by the state. As I explain in Chapter 3, this also means that atomized incorporation manifests itself differently across industries, regions, and time periods. The bargaining that occurs under atomized incorporation does not reflect workers’ organized interests as a group but rather particularistic demands specific to a worker or a single workplace. It can be politically safer than centralized bargaining because atomized claim-makers do not have as much leverage as organized groups. Yet, due to the particularistic nature of the claims that can be made, atomized incorporation cannot reflect the voices of those who lack resources or intend to raise an alarm about conditions more generally. As McCubbins and Schwartz (1984) argue, a fire alarm system encourages particularistic bargaining with a focus on “the interests and rights of individual citizens and small groups” rather than those of the public at large.34 It is not just that this approach to collective claim-making is prone to exclude more marginalized groups; it also affects the nature of demands being negotiated. As I demonstrate in this book, this can be both a strength and a limitation of atomized incorporation in the pursuit of cooperative state–society relations. 1.3 UNTANGLING GRIEVANCES

For an authoritarian regime, small-scale contention grounded on economic grievances might be a tolerable option because it de-escalates citizen discontent. It is postulated that any type of grievance – even narrow economic grievances – can easily be politicized and channeled into rebellion against the regime, unless there are opportunities to engage in small-scale contention.35 In China scholarship, the metaphor of the “safety valve” has been used to describe the benefit of permitting popular contention.36 Giving disgruntled citizens the chance to express their anger can distract them from building a movement that would demand more radical change. Others emphasize the material concessions provided to protesters.37 For instance, Lorentzen (2017) highlights the rewards that the regime grants to protesters as an important condition of the strategy’s success. The study uses the analogy of a “controlled burn” and argues that “once aggrieved groups have called attention to their problems and 34 35

36 37

McCubbins and Schwartz, 1984, p. 172. Comparative literature on democratization also assumes that economic grievances predict preferences for political reform. See Boix (2003) and Acemoglu and Robinson (2009). Hassid, 2012; Chen, 2012, 2016; Li, 2018. Lorentzen, 2017; Chen, 2018.

1.3 Untangling Grievances

13

had them addressed, they will become less interested in participating in anger-venting riots or even broader challenges to the state.”38 Whether as an anger-venting opportunity or a device of material co-optation, protests for short-term economic gains are assumed to reveal the most threatening members of the group – those most likely to have an interest in seeking political change. Yet workers with narrow economic grievances do not always display preferences for broader social or political change. This idea can be traced back to Lenin’s concept of economism. Borrowing Lenin’s argument, Ahlquist and Levi (2013) suggest that workers often display “a focus on the narrow economic interests bound up in the job.”39 They argue that job-specific grievances alone do not lead workers to seek broader change beyond their individual workplace unless they envision a wider “community of fate.” Workers do not automatically believe that broader change is the way to address their own struggles. Similarly, social movement theorists emphasize “system attribution” – the recognition that their hardships originate from stable and external causes – as an important cognitive precondition.40 Short-term economic grievances about job-specific conditions are not necessarily a sign of workers’ system attribution because they can be attributed to workers’ internal qualities or their employer’s personal qualities. Atomized protests that are motivated by narrow economic grievances do not always signal a potential political threat. It is true that public rhetoric about the plight of migrant labor has focused on job-specific economic issues such as low wages and poor working conditions. As migrant workers are mostly unskilled and have only a basic education, their wages are usually determined by the local minimum wage.41 Yet starting salaries for college graduates are not substantially higher than migrant workers’ wages. In 2018, a migrant worker’s average monthly wage in the manufacturing sector was 3958 yuan, while the starting salary for college graduates was about 4317 yuan per month.42 Many college graduates start their career in a white-collar job that pays less than factory work and about 15 percent of graduates are estimated to be unemployed.43 If grievances about wages predict 38 39 40 41

42 43

Lorentzen, 2017, p. 476. Ahlquist and Levi, 2013, p. 3. McAdam, 1982; Ferree and Miller, 1985. The total wages exceed the local minimum wage because of the overtime compensation, but most migrant workers on the production line are paid the local minimum wage as their basic wage (dixin). Jung, 2018; PRC Bureau of Statistics, 2019. Sharma, 2014.

14

Introduction

antiregime attitudes, it is difficult to conclude that migrant labor poses a greater risk than other social groups. A more substantial grievance for migrant workers, however, is their limited upward mobility. For ruralborn workers with limited education – which is a product of the Chinese regime’s urban bias and unequal investment in education between urban and rural regions – their starting salary might seem sufficient, but their low-wage jobs are not the first step on an upward-sloping career path. Take Xiaoyu, a male migrant worker from Sichuan province, who I met in a factory zone in Shenzhen. He was working in the testing department of an electronics factory, and it was already his fifth job after graduating from high school. Even after changing jobs multiple times, and moving from one city to another, his life did not seem to get better. The jobs he could access, regardless of his work experience, were low-skilled jobs that did not provide career mobility. Because most firms do not value migrants’ work experience in low-wage jobs, Xiaoyu’s tenure in a firm did not lead to a substantial increase in wages. On his current income, Xiaoyu was not sure when he would have the financial security to get married. While workers like him did not talk specifically about the central government or political reform, it was clear that their discontent concerned more than a specific workplace. Their work experience led them to perceive the inequality and unfairness of Chinese society and form grievances about migrant workers’ marginalized status. “As a migrant,” Xiaoyu said, “no matter how hard you try there is just no exit.” Short-term economic grievances are not so easily politicized, compared to broader social and structural grievances. It is only when workers perceive that their economic struggles within the firm are part of a wider, discriminatory social structure, and that workers in different firms and regions experience similar issues, that they begin to attribute them to the central government. This perception of interconnection is what the regime has been trying to discourage with atomized incorporation. Without integration into urban communities, and with high levels of labor mobility, migrant workers are already atomized in urban areas.44 This atomized state helps the regime to frame workers’ hardships as a consequence of workers’ lack of character or “internal” issues within a few bad firms.45 In addition to restricting social mobilization across 44 45

Becker, 2014. Social organizations and gatherings that span different groups of citizens produce political effects by facilitating cognitive and attributive processes that generate political consciousness (Verba and Almond, 1963; McAdam, 1982). Atomized incorporation discourages such voluntary associations.

1.3 Untangling Grievances

15

workplaces and regions, atomized incorporation promotes particularistic claim-making that could further undermine the chance of cognitive transformation. Media reports of migrant labor disputes, for instance, have stressed their episodic nature46 rather than presenting them as highlighting a systemic flaw.47 However, the example of workers like Xiaoyu shows that many workers eventually attribute blame to the central government. As migrants accumulate experiences in different jobs and regions, they generalize the patterns of their plight and perceive the systemic nature of their issues. They develop broader social grievances – about limited upward mobility, economic inequality, and social unfairness – and become more likely to direct blame toward the central government. High levels of labor mobility present both an opportunity and a challenge for the regime. Whereas the rise of industrial labor has historically been associated with a regional concentration of economic activity and a reduction in the costs of mobilization for industrial workers,48 migrants in contemporary China cannot easily engage in community-based mobilization. Despite the concentration of migrant labor in a few industrial regions such as the Pearl River Delta (PRD) or Yangtze River Delta (YRD), the cost of mobilization across firms remains high.49 The high levels of mobility, however, also present a different pathway to the development of political grievances, because they expose workers to new sources of information that reveal the systemic – and political – origins of their economic struggles. The experience contrasts with migrant workers in other authoritarian regimes, such as in Mexico under PRI, where workers who were able to settle down in urban areas came to support PRI more as a result of their migration experience.50 Among other factors, workers’ perceptions of upward mobility and social inclusion are crucial for shaping their attributional beliefs. This parallels the argument in the literature that citizens in authoritarian states 46

47 48 49 50

Studies find that the way that media covers social events has important implications for political attribution. Iyengar (1990), for example, finds that “episodic” framing of social problems discourages citizens’ attribution of blame to the political system or leader, compared to “thematic” framing that stresses the systemic nature of the problem. Stockmann and Gallagher (2011) stress that the media reports of migrant labor contention can be characterized as “a few bad apples but happy ending” (p. 445). Rueschemeyer et al., 1992. One exception might be labor NGOs that proliferated in the late 2000s and early 2010s, but they have not survived the state repression, as I discuss in Chapter 8. A classic study by Cornelius (1975) shows that rural-to-urban migrant workers under PRI were able to settle in cities and integrate into urban communities. Their experience of urban integration seems to have enhanced their support for PRI.

Introduction

16

do not form their attitudes toward the regime solely based on current conditions but take into account future payoffs.51 This suggests that authoritarian regimes cannot always buy citizens’ political loyalty in a quid pro quo, unless the regime makes a long-term investment in the broader system to induce a perception, even if illusory, of upward mobility and inclusion. The Chinese regime, however, may find it difficult to shape these perceptions for migrant workers as the era of rapid growth comes to an end, and the continued marginalization of migrants has left them with few ways to achieve upward mobility and social integration into urban areas. As one of the factory managers argued, “there is not much the CCP can do now to make migrant workers happy.”52 1.4 PARTICULARISTIC BARGAINING

Atomized incorporation generates substantially different short-term and long-term outcomes, depending on who actually engages in the particularistic bargaining. The central question becomes whether protesters actually channel their broader grievances about society and the central government. Even with the relatively tolerant environment, workers still suffer from acute collective action problems; not everyone has the resources to generate collective action and anticipate its success. At the same time, those who have the resources to organize labor are not always enthusiastic about protests. In this book, I use empirical analysis to identify the patterns of protests at the individual and firm levels and investigate how they map onto the underlying social and political grievances. If the workers with social and political grievances are the ones to actually protest in the workplace, their visibility means the regime can gather information about who might seek broader change, and potentially co-opt them materially; at the same time, the regime faces the risk of atomized protests turning into political labor activism. It is in the regime’s interest to identify and suppress any shift toward broader mobilization, as potential regime challengers could appropriate the limited space for contention to seek their political goals. If workers with social and political grievances do not use atomized protests to raise their discontent, on the other hand, the regime faces a different challenge. There is a smaller risk of atomized contention developing into a broader political movement; the risk to the regime in this case is that it cannot identify or materially co-opt those with more substantial grievances. In other words, atomized protests 51 52

Blaydes, 2018, p. 40–41. Interview, 07-2012.

1.4 Particularistic Bargaining

17

do not provide the information the regime needs about where the most intense grievances lie. They also do not function as an effective “safety valve” if atomized protests only reflect workers’ job-specific grievances and fail to channel their grievances about society and the regime. In other words, there might be a gap between what is visible in the tolerated space of contention and the broader social and political concerns that are less visible, either because the aggrieved workers are not active or the issues are not being expressed in that space. 1.4.1 Labor Turnover of the Aggrieved I find that the restrictions on access to mobilizational resources outside of individual firms in contemporary China make it difficult for workers to channel their social or political grievances via protests. In the workplace, workers do not have an independent organization to represent their interests and suffer from acute collective action problems. Those who initiate collective action rely on their social networks, as there is no organization that can coordinate collective action. High levels of labor turnover mean that most workers do not have a deep social network within the firm to facilitate collective action. When workers exit a firm, they lose their main resource for mobilization. Workers’ turnover behavior originates from their dissatisfaction with the status quo but results in the loss of mobilizational resources. Those who experience upward mobility within a firm – a small minority among migrant workers – exhibit higher firm-embeddedness; their junior-level position is often a firm-specific asset that is difficult to earn. Those workers’ longer tenure also implies social integration into the workplace and neighborhood, which makes them hesitant to leave the firm. This embeddedness grants them an advantage in initiating collective action, but at the same time makes it more costly for them to engage in collective action. Those who have not experienced any upward mobility, on the other hand, choose to change jobs frequently, but soon realize that job hopping does not improve their economic prospects. As they accumulate social grievances and start to perceive the role of the central government in creating their situation, they also become less capable of launching collective action within firms. This mismatch shows how atomized incorporation can be effective at demobilizing the labor movement by materially co-opting those with firm-level mobilizational resources. At the same time, it means that those with more substantial grievances could be left out.

18

Introduction 1.4.2 Discursive Resources at Law-Abiding Firms

Patterns of labor protests analyzed at the firm level confirm that atomized protests do not necessarily reveal the location of the most substantial grievances. The literature has examined the conditions under which workers participate in collective action, but mostly by analyzing either individual-level factors such as emotion, class identity and social ties53 or macrolevel factors such as labor market conditions and the regime’s attitudes toward labor issues.54 Although they generate important insights, these factors cannot fully explain why collective action is more likely to take place at some firms, but not at others. I examine how within-firm individual heterogeneity interacts with firm-level characteristics to shape the patterns of collective action that we have seen. Since the high-profile strikes at large-scale factories, scholars have questioned why migrant workers – who have been assumed to lack the political consciousness to make demands outside the legal boundaries55 – engage in “interest-based” protests where they aim to maximize their interests, rather than seeking the legal minimum.56 Protests have also become more organized and planned, thanks to the younger generation of workers who are more versed in using online social platforms. These “offensive” or “proactive” protests, scholars and observers argue, could develop into politicized activism because they demonstrate workers’ growing political consciousness.57 Others disagree and point out that even the interest-based protests are apolitical and remain short-lived.58 I find that in the early 2010s when interest-based protests were prevalent, law-abiding firms – such as foreign-owned and large-scale enterprises – experienced more collective action. This outcome challenges the dominant theory of “rightful resistance,” which posits that Chinese protesters use state laws and regulations to legitimize their demands and pressure local officials to correct their behavior.59 Protesters at lawabiding firms, however, neither rely on the law nor target local authorities. Instead, they focus on non-state actors – such as consumers, the public, and the lead firm in the value chain – to pressure the target directly. Unlike rightful resisters who use disruptive tactics such as blocking roads 53 54 55 56 57 58 59

Pun and Huilin, 2010; Chan, 2009; Chan and Siu, 2012; Becker, 2012. Elfstrom and Kuruvilla, 2014; Cai and Meiyan, 2012; Butollo and Brink, 2012. Lee, 2007a. Chan, 2009; Butollo and Brink, 2012; Leung, 2015; Elfstrom and Kuruvilla, 2014. Chen and Tang, 2013; Zhang, 2014. Lee, 2016. O’Brien and Li, 1999.

1.5 Notes on Research Methodology

19

and marching in the street to attract the attention of state authorities, protesters at law-abiding firms rely on innovative but less disruptive tactics.60 As long as the workers use market-based resources to pressure their firm – such as using media to cause reputational damage to the firm’s brands – the local authorities treat the protests as “internal” matters and do not intervene heavily. Workers at law-abiding firms have an advantage in producing collective action due to their discursive resources, which enable them to attract the attention of non-state third-party allies. Yet, these interest-based protests at law-abiding firms could also be less regime-threatening, as their contention rarely breaches the firm’s physical boundary and the workers hold fewer political grievances. These findings together show that atomized incorporation in contemporary China has the effect of revealing the workers with better resources for collective action, which helps the regime to reduce the immediate threat of an organized labor movement. Due to the high levels of social and political grievances among migrant workers – particularly among the younger cohort – gathering information on those with the resources to mobilize firm-level collective action is important. At the same time, the high-visibility firms that are most exposed to the public could also pose a serious threat to the regime if workers at these firms become radicalized. Those who join atomized protests might not be those with most severe grievances about society and the regime, but they could still have considerable grievances and the capacity for collective action. Atomized incorporation, therefore, helps to prevent the radicalization of resourceful workers. My findings, however, also bring into question the assumption in the literature that offering limited channels of contention helps the regime to ensure cooperative state–society relations. Atomized incorporation still excludes a large number of workers with growing social and political grievances and offers them no channel for the expression of their demands. 1.5 NOTES ON RESEARCH METHODOLOGY

The main evidence presented in this book is drawn from two years of in-depth fieldwork in China. I conducted interviews with migrant workers, factory managers, labor activists, scholars, government officials, and trade union cadres in two major industrial regions between 2011 60

In protests that seek to provoke state intervention, protesters often use “trouble making” tactics to appeal to government officials at higher levels. See Chen (2009) and Cai (2010).

Introduction

20

and 2017. I made regular visits to local NGOs where migrant workers received aid and training in labor dispute resolution, and I worked as a volunteer at a community center for migrant workers. I carried out participation observation and interviews with participants at twelve protest scenes. The two regions of fieldwork were the Pearl River Delta area in Guangdong Province and the Yangtze River Delta area around Shanghai. These two regions are the largest manufacturing hubs in China and draw more than half of interprovincial migrant workers.61 They are also the most contentious regions in terms of labor relations in China. The insights from these regions were complemented with follow-up fieldwork in an inland province, Henan, where some of the labor-intensive industries were relocated to from PRD and YRD. The findings in this book, however, do not represent the labor dynamics in China as a whole. The socioeconomic changes that I focus on as a driver of state–labor relations are much more pronounced in the PRD and YRD regions, but they are ideal case studies for examining what has happened in the regions that reaped the most benefits from the regime’s coercive control of labor. The quantitative evidence in this book is mainly drawn from an original survey that I designed and implemented during the summer of 2013 in Shenzhen, Guangdong Province. The survey contains responses from 618 factory workers at 155 factories. Shenzhen was selected as the survey site for two reasons. Shenzhen is at the center of low-wage manufacturing in China and has been considered the most contentious city in PRD. Shenzhen was also the first city to open its doors to foreign capital, while remaining a primary center for global production. The dataset is based on a random sample of manufacturing firms in the city. Due to the higher frequency of labor disputes at foreign-invested firms documented in the early 2010s, I oversampled foreign-owned firms to match the number of domestic private firms. It is difficult to generate a random sample of migrant workers since they are not officially registered. Firm-based sampling enables a rigorous analysis of firm-level factors but can exclude migrant workers without a formal job. I detail the sampling method and its potential disadvantages in Appendix I. The major data collection took place between 2012 and 2013 when migrant workers’ interest-based collective action was at its peak. As explained in Chapter 5, the incidence of interest-based protests increased up to 2014 and decreased thereafter. This period coincided with the China–Japan dispute over the Diaoyu/Senkaku Island that spurred a 61

Chan, 2011.

1.6 Structure of the Book

21

strike wave at Japanese-owned factories. The data from 2012 and 2013 are thus ideal for examining the changing nature of labor protests and workers’ innovative tactics to improve their wages and working conditions beyond the legal minimum. Some labor NGOs played an active role in training workers not only to use the law but also to establish a system for collective bargaining in workplaces. Unfortunately, many of these NGOs closed during the 2015 government crackdown. The changing political atmosphere under the Xi leadership could be one of the reasons why interest-based collective action has become relatively uncommon. Documenting the dynamics of labor contention in the early 2010s, however, could reveal an important insight into the evolution of state– labor relations. Studying which factors enabled workers to overcome collective action barriers to make interest-based demands could reveal why interest-based collective action used to occur so frequently but has declined in recent years. My findings suggest that the conditions that produced workers’ perceptions of group efficacy could have changed over time. For one, the Xi leadership asserted a stronger grip on social activism, grassroots organizations, propaganda, media, and the Internet.62 While the leadership still tolerates atomized contention, its consolidated approach to repressing civil society could have weakened the power of the third-party allies; I find that these allies were particularly important for generating interest-based collective action. The findings from the early 2010s are complemented by the evidence from follow-up fieldwork in 2017 as well as analyses of labor protest incidents in more recent years. Although labor protests do not seem to have decreased in number, their nature has changed. I discuss the continuities and differences of the Hu and Xi eras and their implications for state–labor relations in Chapter 8. 1.6 STRUCTURE OF THE BOOK

The book has three parts. The first part sets out the theoretical and empirical foundations of my analyses. In Chapter 2, I describe the socioeconomic changes in the post-reform era that have contributed to growing labor assertiveness. I contend that the regime’s coercive control of migrant labor in the 1980s and 1990s created the structural conditions for labor assertiveness. In Chapter 3, I introduce the theoretical framework with an examination of the regime’s changing approach to labor control. 62

Fu, 2017; Repnikova, 2017a; Lei, 2017.

22

Introduction

I explain why the Chinese regime has moved away from overt coercion and adopted “atomized incorporation,” and argue that the change could be understood from a political economy perspective. My empirical findings show that the central government and the local governments in developed industrial regions have a new incentive to implement pro-labor policies, even when they undermine the profitability of export-oriented sectors. I contrast the specific components of the new strategy with the strategies of authoritarian labor control observed in Latin America and East Asia. In Part 2 of the book, I examine the conditions under which workers politicize their discontent. Chapter 4 is an examination of workers’ blame attribution, looking at when workers direct their grievances to the central government vis-à-vis other actors. I demonstrate that migrant workers’ social grievances about limited upward mobility, income inequality, and unfairness grow as they gain experience as migrants. While atomized protests focus on economic grievances pertaining to a specific job, I show that social grievances pose a bigger threat to the regime, since they change the direction of blame attribution. In Part 3, I look at how workers overcome collective action problems in workplaces. Chapter 5 is a study of within-firm mobilization during collective action, and I explain why those with the resources for mobilization have weaker preferences for collective action. Due to high levels of labor turnover, the majority of the workforce lacks strong social ties in the workplace, and those who do have mobilizational resources perceive collective action to be highly costly. Collective action occurs when the workers with mobilizational resources expect a high chance of success. In Chapter 6, I analyze firm-level patterns of collective action and find that law-abiding firms are more likely to experience collective action for interest-based demands. Using the strike map dataset of the China Labour Bulletin, I show that interest-based protests are less likely to invite state repression, in part because they do not target state authorities. In Chapter 7, I argue that law-abiding firms’ concerns for reputation generate discursive resources, which contribute to workers’ expectations of success. Unlike collective action for legal rights, interest-based protests rarely use disruptive tactics that physically expand the scope of conflict. Instead, workers use publicity tactics to attract the attention of third-party allies who exercise direct influence over the target firm’s policies. These two chapters show that even in the more favorable environment for atomized protests, not all workers have the resources to engage in collective action. By limiting social mobilization, the regime has been able to manage the

1.6 Structure of the Book

23

frequency and nature of atomized protests. At the same time, workers with the resources to engage in atomized protests are much less likely to hold the central government responsible for the situation they are in. Chapter 8 is a review of the main arguments of the book, with a discussion about how they can be applied to the new political environment in the Xi era and beyond.

2 The Political Costs of Labor Coercion: The Changing Socioeconomic Environment since the 2000s

An enduring question in the study of political economy is how some authoritarian regimes are able to promote rapid economic growth while maintaining regime resilience. The unprecedented economic development in authoritarian countries like China defies the commonly held belief that democracy is necessary for economic growth. An alternative view is that China and other successful authoritarian regimes have been able to achieve economic success because of their nondemocratic political system. According to this view, autocracies enable a more efficient economy because they can suppress civil society’s demands for immediate consumption and for redistribution. This view has been particularly relevant for the political economy of labor. Many nondemocratic regimes in the developing world are guided by the notion that economic growth – at least in the initial stage of development – requires political exclusion of the masses, especially labor. The Brazilian miracle and the rapid growth of the Four Asian Tigers under authoritarian rule have been associated with those regimes’ coercive control of labor. A similar view prevailed in China in the post-reform period. Labor coercion was deemed necessary to attract foreign direct investment and accelerate export-oriented industrialization (EOI), while maintaining labor peace. The regime chose this approach to promote economic growth and ensure political stability. China’s economic success during the last three decades might be taken as evidence that the strategy works. In the post-reform era, China experienced an unprecedented transformation from a planned, backward, agricultural, and closed economy to “the World’s Factory.” Within a short period of time, China embraced industrialization, urbanization, 24

2.1 Political Economy of Labor Coercion

25

marketization, and globalization. Despite the subpar working conditions, low wages and other labor abuses, factory workers in export-oriented industries were relatively acquiescent until the 1990s. Their low wages and limited access to social benefits contributed to rapid capital accumulation in export-oriented regions, enabling both industry and local government to reap the benefits of industrialization with minimum costs. The growing labor unrest in the last decade, however, forewarns that the strategy of labor coercion is becoming less effective. In this chapter, I analyze why, and find that sustained labor coercion in the post-reform period created structural conditions for labor assertiveness in the long run. I focus on the hukou system, a mechanism that has been instrumental to labor coercion in post-reform China. I show how the hukou system enabled the state’s coercive control of labor, and that the laborintensive growth achieved in the 1980s and 1990s induced socioeconomic transformations that have rendered the hukou system ineffective. China’s example shows that using labor coercion to promote economic growth is an inherently unstable strategy for an authoritarian regime. It not only breeds grievances among marginalized workers but also grants them structural power in the long run. It is a strategy that undermines its own logic and makes the regime more vulnerable as the potential for political instability grows. 2.1 POLITICAL ECONOMY OF LABOR COERCION

Labor coercion is a political strategy that enhances employers’ bargaining power vis-à-vis labor. Historical examples of labor coercion abound, ranging from slavery and serfdom to strike bans and wage controls. The state does not need to be directly involved in coercing workers to accept the terms of employment offered: in capitalist societies, employers can use the threat of dismissal to induce workers’ effort without providing commensurate compensation. However, it is difficult to argue that coercive labor relations at the firm level are independent of state policies. For example, a law banning collective action is a form of market intervention, even though it could move wages closer to the market wage. Thus, this study focuses on the state’s use of political methods to reduce workers’ bargaining power.1 The term “labor repression” does not capture 1

In this study, I use ‘labor coercion’, but the literature uses other terms as well. Moore (1966) uses “labor repression,” while Wood (2000) introduces the term “extraeconomic coercion” to emphasize its reliance on nonmarket control. Wintrobe (1998) uses a more generic term, “economic exploitation,” but he defines it as the situation in which “one group uses its political power over another to appropriate its earnings” (p. 163).

26

The Political Costs of Labor Coercion

economic marginalization of labor, which is central to the analysis here. Exclusionary corporatism, for instance, could be described as repressive because it depoliticizes labor and weakens its political voice but does not always reduce labor income.2 2.1.1 The Developmental Logic of Labor Coercion It has been suggested that authoritarian regimes have a better set of political tools for labor coercion than democracies, putting them in a better position to promote economic growth. Wintrobe (1998), for example, argues that capitalist authoritarian regimes, such as Chile under Pinochet, South Africa under Apartheid, and South Korea under General Park, achieved rapid economic development because they had the capacity to redistribute the rights to earn income3 : [These regimes] essentially redistribute by altering rights in the workplace: They shift the rights of labor to a management backed by the state, they raise the cost of job loss to workers, they remove or outlaw collective bargaining rights, and in other ways they generate a labor force willing to work for low wages.

In other words, the “machinery of dictatorship – rewards, sanctions, and instruments of monitoring” in authoritarian regimes might deliver a comparative advantage in labor coercion.4 There is some empirical evidence that nondemocratic governments treat their workers worse than democratic governments do and may achieve economic development via labor coercion. For instance, the labor share of income in general is higher in democracies than in nondemocracies.5 Przeworski and Alvarez (2000) suggest that while growth in advanced democracies is based on technical progress, “[w]ealthier dictatorships grow by using a lot of labor and paying it little” (p. 179). In other words, successful authoritarian regimes tend to rely on labor-intensive growth. The idea that labor coercion aids industrial growth has a long academic history. The literature on “bureaucratic authoritarianism” has argued that expanding industrial production is not possible without suppressing labor’s demands for redistribution.6 Studies on rapid economic 2 3 4 5

6

See Collier (1982) for her explanation of Mexico and Brazil, contrasting them as inclusionary and exclusionary corporatist systems. Wintrobe, 1998, p. 160. Ibid., p. 142. Studies such as Acemoglu and Robinson (2009) and Przeworski and Alvarez (2000) find that labor share is higher in democratic states. Rodrik (1999) also demonstrates that workers with comparable levels of skill and productivity receive a better wage in democracies. See also Palley (2005). O’Donnell, 1988.

2.1 Political Economy of Labor Coercion

27

growth in Asia have also pointed to labor coercion as a major contributor to the success of EOI in the region. According to this view, the kind of industrialization a regime pursues is also important. EOI, in particular, requires strong government intervention to hold wages at low, internationally competitive levels. Import-substitution industrialization, on the other hand, has been associated with a more inclusionary approach to labor.7 Some scholars argue that the Four Asian Tigers – Taiwan, Singapore, South Korea, and Hong Kong – owe much of their success to their authoritarian labor control.8 The relationship between labor coercion and economic growth is the subject of much debate.9 Yet, history provides many examples of regimes that pursued coercive control of labor, believing that it would attract more capital investment and accelerate industrialization. The “wage squeeze” policy in Brazil after the military coup in 1964 was sold with the idea that the exclusion of populist demands – particularly labor’s demands for wage growth – would contribute to economic growth and stability.10 Some regimes designed their labor policies explicitly around the idea that higher wages would, by making exports less price competitive and reducing inflows of foreign direct investment, disrupt economic development. For example, Bangladesh has placed strict restrictions on workers’ rights to associate and engage in collective action in export processing zones. Park’s military regime in South Korea banned strikes in foreign-invested sectors out of fear that labor unrest would deter foreign direct investment flows. The changes to the structure of international production since the 1990s seem to have created more incentives for developing economies to suppress labor demands. As global value chains have enabled developing states with little technological sophistication to enter, the “join-butnot-build” model has induced more severe competition among these economies to reduce labor costs.11 Nondemocracies in South Asia such 7 8 9

10

11

See Kuruvilla (1996). There are also scholars who argue against the thesis. Deyo, 1989; Haggard and Kaufman, 2008. It has been argued that NICs grew rapidly, not because of, but despite labor coercion, since labor coercion distorted the labor market and created economic inefficiencies (Fields, 1994). Some point out that there is no significant relationship between labor repression and economic growth among non-democracies (Geddes, 2003). Other studies cast doubt on the notion that nondemocratic regimes are better able to maintain labor peace (Teitelbaum, 2011; Robertson and Teitelbaum, 2011). In Brazil, the “wage squeeze” policy lowered workers’ real wages by as much as 35 percent between 1964 and 1970. See Seidman (1994, p. 59) and Geddes and Zaller (1989). Mosley and Uno, 2007; Anner, 2015.

28

The Political Costs of Labor Coercion

as Malaysia, Thailand, Cambodia, and Vietnam have been major participants in global value chains in labor-intensive industries such as apparel, textiles, and electronics. These regimes embraced EOI and the depoliticization of civil society to tap into the new opportunities created by the internationalization of production.12 Labor coercion was an integral part of China’s post-reform strategy of pursuing growth via EOI and FDI. The regime’s tight grip on labor was deemed instrumental in the creation of low-cost, disciplined, and relatively acquiescent labor until the 1990s. Yet, from the early 2000s, signs emerged that factory workers in China were no longer submitting to the coercive control of labor. Why is the state’s coercive control of labor declining in effectiveness? What are the implications of the Chinese experience for other labor-coercive regimes? To answer these questions, I begin with a brief analysis of how the household registration system, hukou, functioned as a major instrument of labor coercion. Then I examine the ways in which labor coercive growth brought about structural transformations in China.13 2.2 LABOR COERCION IN POST-REFORM CHINA

2.2.1 Overview of the Hukou System While its historical roots lie in Imperial China, the CCP’s hukou system was established nationwide in 1958.14 Throughout the pre-reform period, it functioned as a “de facto domestic passport system.”15 It heavily restricted individual mobility to the point that “peasants could be arrested just for entering the cities.”16 Rural-to-urban migration was tightly controlled and the central government imposed strict regulations on nongzhuanfei, the re-classification of a hukou from rural to urban. The hukou system reflects the strong urban bias of the Chinese regime. It has created and maintained two different classes based on geography: The privileged urban class and the marginalized rural class. Urban residents enjoyed many privileges during the Mao years. They were organized 12 13

14

15 16

Haggard and Kaufman, 2008; Human Rights Watch, 2009. It is still debated whether labor coercion induced economic development in post-reform China. The term “labor-coercive growth,” therefore, does not assume causation, but denotes labor-intensive growth accompanying labor coercion. The system has its roots in the traditional baojia system that was a means of social control in Imperial China, particularly during the Ming and Qing era (Cheng and Selden, 1994; Wang, 2005). Chan, 1996, p. 136. Young, 2002, p. 32.

2.2 Labor Coercion in Post-reform China

29

under a system of work units (danwei), which took care of their members from birth to death. Danwei provided lifelong employment, housing, social insurance, and other welfare benefits such as education. The state procured agricultural produce from the countryside and distributed it to urban residents via rationing. Rural residents were organized by rural collectives and were obliged to send agricultural output to the state, but did not enjoy much social protection. As a result, urban residents had a substantially better quality of life than rural residents.17 By restricting rural-to-urban migration, the hukou system protected urban privileges. The best illustration of the urban–rural gap in this period is that most of the famine-related deaths were in the countryside. During the Great Leap Forward, for example, the rural collectives had no food left to feed their members after meeting the state quotas, while the level of urban consumption was generally maintained through rationing.18 Despite the often dire situations in the countryside, farmers could not move to the cities because they could not access basic goods and services such as grain coupons, housing, or employment. Some scholars suggest that restrictive migration control was influenced by the Soviet model of industrialization. The Soviet Republic and China both pursued “Big Push” industrialization that focused on developing heavy industries. Agricultural products had to be provided at a low price to urban areas to finance industrialization.19 By keeping the majority of the population in the countryside, the hukou system could be used to suppress agricultural prices and direct the savings to urban sectors.20 When the CCP leadership initiated the economic reform in 1978, the majority of the population – 87.5 percent – was classified as rural. 2.2.2 Opening Up and Migration When economic reform started in 1978, the Chinese economy was suffering from slow economic growth and low productivity. The promotion of heavy industries and suppression of light industries in the Mao era wasted the economy’s comparative advantage in abundant labor, distorted factor prices, and led to excessive surplus labor.21 The early agricultural reforms, including decollectivization and the introduction of the Household Responsibility System, gave rural households reason to be more 17 18 19 20 21

Yang and Cai, 2003. During the Great Leap Forward, about 20 to 30 million people died of starvation. Most of the deaths were in the countryside (Lardy, 1987; Yang and Cai, 2003). Naughton, 2006. Chan, 1994; Naughton, 2006; Fan, 2007. Lin et al., 2003.

30

The Political Costs of Labor Coercion

productive and released surplus agricultural labor.22 According to various estimates, in the 1980s, China had 100 to 200 million surplus laborers.23 To reduce the surplus labor, the regime started to allow farmers to work in nonagricultural sectors. In 1985, the central government introduced the “temporary resident” system, whereby rural workers with a guaranteed job could receive a short-term residency permit (zanzhuzheng) in the region of their work. However, the central leadership only relaxed, and did not eliminate, the institutional dualism between urban and rural residents. Rural workers seeking employment in a different region had to secure consent from the labor bureaus at both origin and destination. Workers were also expected to return to their village of origin at the end of their short-term labor contracts. To encourage farmers to “leave the land, but not leave the village” (litu bulixiang), the state actively promoted rural industrialization. The Township and Village Enterprises (TVEs) – collectively owned enterprises at the township and village level – rapidly expanded and became a dominant source of employment for rural workers. TVEs induced some level of migration to nearby towns, but it was small scale and remained local. From the 1990s, migration grew rapidly in volume and distance. A reform that had an important role in shaping the level and patterns of migration was “opening up” (kaifang) and China’s pursuit of EOI. Ruralto-urban migrants became the primary source of labor for export-oriented industries. It was not just that the local workforce could not meet the growing demand in labor-intensive industries, but outward-oriented firms preferred rural migrant labor because it was cheaper. Studies show that in the early years of reform, firm managers were prejudiced against migrant workers, who were considered “not used to the labor discipline,” “slow to pick up technical skills,” and “undesirable.”24 Still, factories that produced for an overseas market were expected to hire low-wage migrants to maximize their price competitiveness. Expansion of export-oriented industries, therefore, increased the scale of migration.25 The sequential process of opening up drew a disproportionately large volume of migration from inland to coastal provinces. During the Mao era, particularly after the Sino-Soviet split, the CCP emphasized self-reliance and maintained a closed door policy. It was after Deng’s 22

23 24 25

These two reforms changed the unit of production from collectives to households. Households were granted the freedom to plan and carry out the production independently and to keep the residual income. Li, 1996. Siu, 1990, p. 77. Zhang and Song, 2001; Fu and Balasubramanyam, 2005.

2.2 Labor Coercion in Post-reform China

31

return to the central leadership that the Party started to implement substantial opening up reforms. The reforms did not receive full support from the Party’s conservative members, so the reformist leadership adopted an experimental approach, starting with a few coastal cities. In 1979, it nominated four coastal cities in the South – Shenzhen, Zhuhai, Shantou, and Xiamen – as “Special Economic Zones” (SEZs), which were allowed to accept foreign investment and engage in export activities. The regime expanded its open door policy to fourteen other coastal cities in 1984; to three Delta regions – Pearl River Delta in Guangdong, Min River Delta in Fujian, and Yangtze River Delta around Shanghai – in 1985, and eventually to the entire coast in 1988. The coastal regions took advantage of the new economic opportunities and quickly tapped into the international economy, while inland regions lagged behind. The uneven geographical development induced interprovincial migration between inland and coastal provinces. Guangdong Province has been the top destination for migration, unrivaled since the mid-1980s. The province received nearly one-third of all inter-provincial migrants in China between 1985 and 2000.26 Located next to Hong Kong and Macau, Guandong took advantage of the close proximity to ethnic Chinese investors overseas and developed export-oriented industries early on. For instance, even when migration remained relatively small scale and short distance in the 1980s, the Shenzhen SEZ – the first and largest SEZ in China – experienced explosive population growth. Shenzhen was a small town with 314,100 residents in 1979, but by 1994, it had a population of 3.3 million. Population growth came mostly from migration: More than 70 percent of residents in the mid-1990s were migrants.27 These migrant workers were mostly employed in foreign-invested enterprises: In 1995, 56 percent of migrant workers in Shenzhen were working at foreign-owned enterprises, 14.4 percent at sanlaiyibu enterprises,28 16.7 percent at SOEs, and only 12.6 percent at private enterprises. Since the 2000s, with the rise of the YRD region as a modern manufacturing base, the PRD and YRD have accepted more than half of interprovincial migrants. On the other hand, inland regions such as Sichuan, Anhui, and Hunan have been top exporters of migrants.29 Migrant workers’ contribution to industrialization in China is reflected in their sector of employment. By the 1990s, the majority of migrant 26 27 28 29

Fan, 1996; Chan et al., 1999; Chan, 2011. Liang, 1999, p. 123. sanlaiyibu enterprises are processing facilities that produce final goods with the raw materials and designs provided by an overseas partner. Chan, 2011.

32

The Political Costs of Labor Coercion

workers were in secondary industry. According to the 1 percent population census in 1990, about 60 percent of migrants were industrial workers and the number went up to 70.5 percent for rural migrants.30 Recent statistics also demonstrate that manufacturing still attracts a substantial fraction of migrant workers: In 2019, for instance, about 27.4 percent of rural migrants worked in manufacturing, while 18.7 percent worked in construction. Migrant labor has become the main source of labor for labor-intensive factories. In export-oriented cities such as Shenzhen, the majority of migrant workers were employed as production workers in factories.31 2.2.3 The Hukou System as a Labor Coercive Institution Despite the increasing flow of migration, the principle of litu bulixiang persisted in a way throughout the post-reform era. The regime permitted migrant workers to find jobs in cities but limited via quota the percentage of the population that could transfer from a rural to an urban hukou.32 Such transfers are extremely difficult, although the restrictions on receiving an urban hukou in small- and medium-sized cities have been relaxed since the mid-1990s.33 As displayed in Figure 2.1, due to the restrictions on the official conversion, the number of migrants who formally transferred their hukou status remained small, while the number of unofficial migrants without hukou conversion increased dramatically. 2.2.4 Workers’ Dependence on Employers In its restrictions on labor mobility, the hukou system shares many features with Apartheid in South Africa.34 Like Apartheid, the hukou system controlled the influx of migrant workers to cities, only granting temporary residential status in cities to those with proper documents. These documents included, but were not limited to, the temporary residence permit, work permit, migration permit, marriage (or single-status) certificate, 30 31 32 33

34

Chan et al., 1999. For instance, according to a survey in 1993, 74.19 percent of migrants in Shenzhen were factory workers (Liang, 1999, p. 125). Wang, 2005, p. 50. In the 1990s, large cities like Shanghai introduced the “blue-seal” hukou that worked in the same way as green cards in the United States, but it was only available to people with wealth, skills, and education. See Feng et al. (2002). For instance, Alexander and Chan (2004) discuss similarities and differences between the Apartheid and hukou system. They conclude that, despite their dissimilar origins, both systems were a mechanism to control labor mobility using influx control and lower wages for migrants.

33

0

20

40

Millions 60

80

100

120

2.2 Labor Coercion in Post-reform China

1980

1990

2000

2010

Year Official migration

Unofficial migration

FIGURE 2.1 Number of migrants Source: Chan (2011, p. 83). Unofficial migration size captures the number of peasant workers.

and family planning permit. No matter how long the migrant workers resided and worked in a city, they were rarely granted the urban hukou. Due to their temporary and unstable residential status, migrants without the local hukou have been called “floating population” or “temporary workers.” The central consequence of the hukou system as a labor market institution was that it increased workers’ dependence on their employers for survival, sometimes quite literally. When explaining the economic consequences of Apartheid, Wintrobe (1998) shows how Apartheid increased the costs of job loss by tying workers to their employer: [Under Apartheid,] the loss of a job also implies the loss of a pass – and therefore of the right to remain in the white manufacturing sector. This gives (white) employers considerably more power over their employees than they possess in a free market. In a free market the maximum sanction available to an employer is dismissal from that firm. The influx-control-pass law system formally empowers the employer to dismiss an employee from the entire manufacturing sector, considerably raising the cost of job loss to the worker (p. 184).

The hukou system likewise had institutional features that tied migrants’ residential status to their job. Until 2003, those migrants without a job,

34

The Political Costs of Labor Coercion

housing, and an identification card – the “without-three” population (sanwu renyuan) – could be placed in a local deportation center and sent back to their rural place of origin. Under this system, losing a job meant losing both income and legal residential status. City police would stop people in the street if they appeared to be migrants, to see if they were carrying all the required documents. Survey data show that these random stops were common. According to a survey conducted between 2004 and 2005 in four major cities, about 41.2 percent of migrant workers had had their identification checked by police and 18.3 percent had been fined. While a relatively small subset – 5.5 percent – of migrants reported having been deported, a more substantial number had an indirect experience: 26.5 percent of workers had seen their friend, relative, or fellow villager be detained and deported.35 Sometimes, migrant workers were literally dependent on their employers for their physical safety. Since local governments did not provide much protection for migrant workers’ communities, workers considered their gated factories to be a protected area. When police intensified random checks, for example, employers directed their workers who didn’t have proper documents to stay inside the factories, to avoid arrest.36 Workers’ dependence on their employers for everyday survival induced their acquiescence and willingness to work hard, if not outright support for some firms’ disciplinary practices.37 In other words, under the hukou system, a job provided residential, and sometimes physical, security for migrant workers and made migrant workers more loyal to their employers. Losing a job was also more costly for migrant workers than for locals, because they did not have access to social security. Until the late 2000s, employers were not required to provide social insurance to migrant workers who didn’t have the local hukou. Nor could migrant workers access public goods and social services offered by the city government, such as unemployment allowances and public assistance for job search. Despite the prevailing belief that migrant workers’ job turnover has always been higher than locals’, the data on job mobility from 1998 show that migrant 35 36

37

Zheng and Huang, 2007, pp. 76–77. Pun (2005), for example, notes that those workers who did not hold proper documents were worried about getting arrested in the street and listened to the firms’ advice to stay inside the factory. The “dormitory regime” in Chinese factory zones allowed employers to control workers even outside of the working hours. Peng (2011), for example, shows that workers supported firms’ imposition of curfew because it was seen as a mechanism to ensure workers’ safety.

2.2 Labor Coercion in Post-reform China

35

workers enjoyed fewer job opportunities, had higher job search costs, and stayed in the same job for a longer period than their local counterparts.38 Migrant workers’ dependence on their employers, not surprisingly, weakened their bargaining power in the workplace. One immediate consequence of reduced bargaining power is lower labor compensation. As many studies have documented, migrant workers’ wages have been significantly lower due to their hukou status. A survey conducted in 1996, for example, found that migrant workers received between 74 and 89 percent of urban workers’ wages even when they had similar labor productivity.39 Some show that the real wages of migrant workers did not meaningfully increase until the early 2000s, despite rising labor productivity.40 Thus, wages lagged well behind migrant workers’ marginal productivity of labor. In their study of the wage gap in the 1990s, Knight et al. (1999) show that migrant workers’ marginal product of labor was nearly four times higher than their wage level.41 As a result, migrant workers had a much lower return on their human capital than urban workers.42 The pay gap between migrant workers and urban workers becomes even greater when nonwage payments are taken into account.43 On top of their economic marginalization, migrant workers experienced other mistreatment in the workplace, especially in export-oriented regions. Migrant workers have been reported to suffer from long hours, an unhealthy and dangerous work environment, and physical and verbal abuse.44 According to a survey from 2004, 15.3 percent of migrant workers experienced physical and verbal abuse, 27.2 percent had to pay a deposit to their employer, and 22.2 percent experienced wage arrears within the previous six months.45 Some factories adopted military-style control of workers and used rules to control behavior such as talking, drinking water, or going to the restroom.46 Due to migrant workers’ lack of access to housing in cities, most factories house migrant workers in dormitories, which enables “the dormitory labor regime.” This 38 39 40 41

42 43 44 45 46

Zhang, 2010. Knight et al., 1999, p. 84. Meng and Bai, 2007; Cai and Du, 2011. Urban workers, on the other hand, received more than their MPL, although the premium was not large. While the MPL for migrant workers was 20,076 yuan per year, their average wage was only 5,368 yuan. By contrast, urban workers received 6,956 yuan per year when their MPL was 5,597 yuan. Knight et al., 1999. Lee, 2012. Chan and Senser, 1997; Pun, 2005; Chan and Siu, 2010. Zheng and Huang, 2007, pp. 57–69. Chan, 2001.

36

The Political Costs of Labor Coercion

regime increases the control employers have over workers inside and outside the factory.47 2.3 DECLINING EFFECTIVENESS OF LABOR COERCION

The exponential growth of the Chinese economy and the success of EOI based on low-wage labor seem to reaffirm that – at least in the short term – labor coercion as a developmental strategy could be effective. Economists have suggested that low-cost labor in China was key to rapid capital accumulation in the post-reform era.48 Political scientists and sociologists have also suggested that the hukou system is an important contributor to China’s economic growth and its ascendancy to the position of the world’s top exporter.49 In spite of their mistreatment, migrant workers were generally considered to be compliant and obedient in the 1990s. For instance, when firm managers from 118 enterprises were surveyed in 1996, the most common reason given for preferring migrants to locals was that migrant workers “can bear hardship and are easily manageable.”50 The rising labor unrest since the 2000s, however, shows that “the days of cheap, docile labour [are] coming to an end.”51 This section shows that socioeconomic transformations in the post-reform era generated two structural conditions that undermined the effectiveness of labor coercion as a tool for labor discipline: the structural empowerment of labor and deepening labor discontent. 2.3.1 Labor Shortage The logic of labor coercion presumes that workers would acquiesce to coercive control by their employers because the circumstances make them dependent on their employers for survival. However, in the early 2000s, there was a shift that empowered migrant labor and changed the dynamics of the market: a labor shortage emerged. What had been an unlimited supply of labor from the country side began to contract, and by 2003, there were reports of acute labor shortages in export-oriented regions. According to the quarterly surveys by the Ministry of Human Resources and Social Security, the ratio of the newly created jobs to job seekers increased drastically from the early 2000s, exceeding 1 in 2010 (see Figure 2.2). Looking at Guangdong Province alone, the ratio was already 47 48 49 50 51

Smith and Pun, 2006; Pun and Smith, 2007. Woo, 1996; Lin et al., 2003. Alexander and Chan, 2004; Wang, 2005; Chan, 2010a, 2012b; Pun and Chan, 2013. Knight et al., 1999, pp. 91–92. China Labour Bulletin, 2012.

37

.6

Job/seeker ratio .8 1

1.2

2.3 Declining Effectiveness of Labor Coercion

2001q3 2003q3 2005q3 2007q3 2009q3 2011q3 2013q3 2015q3 2017q3 Quarter FIGURE 2.2 The ratio of available jobs to job seekers Source: Ministry of Human Resources and Social Security.

greater than 1 in 2002, meaning that the province had more job positions than workers. The ratio reached 1.89 in PRD during the first quarter of 2008.52 The labor bureau dispatched a research team to study the shortage, which reported in 2004 that the shortage of migrant labor had become a serious challenge to sectors using low-skilled labor in PRD and YRD. The PRD region, according to the report, was in need of 2 million workers.53 Among the cities in PRD, Shenzhen had a shortage of 400,000 workers even though the city already had 4.2 million migrant workers. The labor shortage has become a major challenge for export-oriented businesses. According to a survey conducted by the Guangdong Provincial Statistics Bureau, more than 72 percent of companies experienced difficulties in recruiting low-wage workers.54 Self-reports in surveys by the US China Business Council and the Hong Kong manufacturers’ association in China found that labor shortage and high labor costs were their members’ biggest challenges in China.55 My interviews corroborate this 52 53 54 55

Li et al., 2012, p. 69. Labor Bureau Research Team, 2004. Shenzhen Daily, 2005. Rapoza, 2011; CMAHK, 2010, 2011.

38

The Political Costs of Labor Coercion

finding. For instance, a firm manager at a high-end wig factory in Shenzhen noted the drastically different labor market situation compared to the 1990s56 : When we first established our production facilities in 1989, it was so easy to recruit workers. During the 1990s, we did not have to run a big advertisement. We just put up a small recruitment poster on the factory wall, then floods of people would line up at the factory gate. At one point, the factory wall collapsed as too many workers leaned against it. However, now it is not easy to find workers who are willing to work in a wig factory. Over the years, we have downsized our factory. We used to employ up to 3700 workers, but now we only have 80. We moved most of our production facilities to Indonesia, because the Indonesian government promised to help us recruit 10,000 workers.

As the term “labor scarcity” (yonggonghuang) recurs in national headlines, scholars have debated whether the Chinese economy has already exhausted the surplus labor from the countryside and reached the Lewis Turning Point.57 While there exists some disagreement over whether the nonagricultural sectors have completely exhausted the surplus labor from the countryside, various data sources suggest that China can no longer be characterized as a labor surplus economy.58 Particularly among the younger cohort who can enter the urban labor market as unskilled migrants, most are already participating in the off-farm labor market, as demonstrated in Table 2.1. Studies show that both the real and nominal wages of migrant workers have risen rapidly since 2004.59 The shortage of migrant labor means that migrant workers have gained what Wright (2000) refers to as “structural power” – the power that stems from “the location of workers within the economic system” (p. 962).60 Migrant workers have a weaker incentive to acquiesce to employers’ coercive control, because high labor demand makes it easier to switch jobs. A majority of migrant workers believe that they can easily find a new job if they leave their current employment. This contrasts with 56 57

58 59 60

Interview 09-2013. When an economy transforms itself from a traditional economy relying on agriculture to an industrialized modern economy, the initial stage of this transformation is characterized by a nearly unlimited supply of low-skilled labor. The Lewis Turning Point occurs when the expansion of nonagricultural sectors exhausts the supply of surplus labor and labor shortages emerge. Some suggest that surplus labor still exists in the countryside, pointing to rural unemployment (Golley and Meng, 2011; Knight et al., 2011). Wang, 2010; Cai and Du, 2011; Lu, 2012. The author uses the term “structural power” in contrast to another source of labor power, which is “associational power.” How these two affect migrant labor’s contentious behavior and the state’s strategy of managing it is discussed in Chapter 3.

2.3 Declining Effectiveness of Labor Coercion

39

TABLE 2.1 Percentage of rural workers in nonagricultural sectors

Age

1990

2011

16–20 21–25 26–30 31–35 36–40 41–50

24 34 29 27 21 21

88 90 88 84 74 61

Source: Li et al. (2013, p. 11).

workers’ perceptions in the 1990s that job switching was not easy.61 In a tight labor market, it becomes more difficult to coerce labor. Reflecting this change, labor turnover has increased substantially. As the job options increase, dissatisfied migrant employees are more likely to move between jobs. Migrant workers’ frequent turnover, according to some firm managers, is a bigger obstacle to production than the labor shortage itself. A firm manager at a large-scale textile factory in Shandong said he could find workers to fill posts but they left too quickly.62 Another firm manager who runs a small-scale paper packaging factory stated in 2013 that he typically needed to go through one hundred workers to retain just two.63 In other words, migrant workers are able to “vote with their feet” when it comes to their choice of employment. 2.3.2 Labor Discontent The literature reveals that sustained labor coercion also created an environment where workers felt aggrieved about their marginalized conditions. While migrant workers in the 1980s and 1990s were acquiescent despite their poor treatment in the workplace, now they are more likely to resist. This argument is largely based on the observation that the younger generation of migrants expresses growing resentment. Scholars and observers differentiate the “new generation of peasant workers”

61 62 63

Daming and Xiaoyun, 2014, p. 63. Interview 08-2012. Interview 03-2012.

40

The Political Costs of Labor Coercion

(xinshengdai nongmingong) from the previous generation.64 The new generation, born in the 1980s and 1990s, is often characterized by “three highs and one low” (san gao yi di): high levels of education, career aspirations, and demands for material and psychological enjoyment, but low ability to endure hardship. According to the PRC National Bureau of Statistics,65 a substantially higher percentage of this generation has an educational qualification beyond technical school.66 The new generation is also less likely to send money home and more likely to spend on their personal enjoyment. Lastly, even though members of the younger generation are still referred to as “peasant workers” (nongmingong), most have no experience of farming, and a substantial portion (37 percent) express their desire to stay in the city.67 According to Huntington’s “gap hypothesis,” modernization accompanied by “urbanization, literacy, education, and mass media” raises people’s social aspirations.68 Social aspirations could induce social frustrations and political instability if there is no economic or social mobility. The rising social aspirations of the “new generation” of migrant labor could be a product of the rapid industrialization and modernization of the post-reform era. During the three decades of reform, the percentage of urban population grew from 17.9 percent in 1978 to 57.96 percent in 2017.69 The literacy rate jumped from 66 percent in 1982 to 95 percent in 2010.70 Modes of communication changed as mass media and the internet flourished. The younger cohort also benefited from the Chinese government’s initiatives to implement nine-year compulsory education in the late 1980s. These workers born in the post-reform era might have a different set of benchmarks against which to evaluate their current conditions, since their starting point is higher than that of the previous generations of migrants. Due to their rising aspirations and expectations, the younger migrant workers seem less likely to perceive their unskilled jobs in labor-intensive 64

65 66 67 68 69 70

The National Bureau of Statistics estimates that the number of the new generation migrant workers reached more than 100 million in 2013, which was roughly 60 percent of the entire rural migrant population (PRC Bureau of Statistics, 2014). PRC Bureau of Statistics, 2011. While only 3.5 percent of the older generation attended technical school (zhongzhuan) or above, the ratio increases to 15.4 percent for the new generation. Data from 2011 show that nearly 90 percent of rural residents under 30 were already working in off-farm sectors (Li et al., 2013). Huntington, 1968, pp. 53–56. World Bank Data. UNESCO Institute for Statistics.

2.3 Declining Effectiveness of Labor Coercion

41

sectors as a valuable opportunity. Their attitudes toward factory jobs are reflected in firm managers’ descriptions of younger workers as unprofessional, impatient, and self-oriented. One manager at an electronics factory in Dongguan, for example, complained that “the young migrants do not know how to work hard.” “If their shift ends while they are in the middle of a task,” he explains, “they just leave without completing the task, even if it would only take two extra minutes.”71 In addition to the perception of the young workers’ “inability to bear hardship,” firm managers generally share the belief that they have become more conscious about their own interests. “Ten years ago,” one manager at an apparel factory in Guangzhou recalls, “workers did not even ask me how much they would get paid for overtime hours or if at all. Now, workers don’t want to sacrifice even one yuan.”72 From the workers’ perspective, there might not be a benefit to exerting extra effort in workplace. One important factor behind this perception seems to be the limited opportunities for upward mobility within the organization. Promotions from the low-skilled positions available to migrants are rare, which means that even if they exert themselves more, most would not gain a better position or substantially better wages. The training they receive as unskilled workers also does not appear to help them to find a better job in the future. Studies show that instead of aiming for promotions, migrant workers switch jobs and geographical locations to pursue upward mobility. Yet, even with job switching, migrant workers do not significantly improve their wages and economic status, demonstrating that work experience does not translate into career progression.73 This is also in large part related to the rural–urban gap in public goods provision. The big differences in the quality and availability of public education to urban and rural residents mean that it is more difficult for rural residents to acquire the human capital for upward mobility.74 Despite the central government’s recent efforts to improve the quality of public goods provision to rural areas, a substantial rural–urban gap in education still remains. Outside the workplace, migrant workers’ status in the cities has not improved substantially. The expanding economic opportunities in the 1980s and 1990s meant a better quality of life for the first generation

71 72 73 74

Interview 05-2012. Interview 03-2012. Tian and Xu, 2015; Daming and Xiaoyun, 2014. Knight and Shi, 1996; UNDP, 2005; Fu and Ren, 2010.

42

The Political Costs of Labor Coercion

of migrant workers; they could experience horizontal mobility by migrating to the cities. Most of them were also seasonal or transient laborers who had deeper roots in the village. The main motivation for migration for this generation was to earn enough money to return home and build a house. A 1996 survey of migrant workers in urban enterprises revealed that a majority of migrant workers at that time were relatively satisfied with their job and migration experience as a whole.75 By contrast, the new generation grew up in relatively good economic conditions. Their benchmark for comparison is the city, not the countryside.76 Yet, due to the hukou restrictions and continued social exclusion, it is difficult for migrants to integrate into urban areas.77 The ongoing discrimination against migrant labor brought about social prejudice and negative social identity, contributing to their lack of social mobility. Studies of discriminatory institutions such as the caste or slavery system reveal that the ruling members of the society constantly create social narratives about the marginalized group to justify the discrimination.78 These negative social characterizations affect the social identity of the marginalized and have long-lasting effects on their economic and social status, even after the institutional barriers disappear.79 Migrant workers in cities constantly face prejudice and are stereotyped as “outside people” (waidiren) who engage in low-skilled jobs that locals avoid, commit crimes, and are of low quality.80 Thus, the term dagong – which literally translates into “being employed” but is almost exclusively used for migrant workers’ low-skilled employment – has social connotations of belonging to the lowest social class. Some migrant workers began to express their frustrations about the dagong life, reflected in this migrant worker’s poem:81 I haven’t made it to the 21st century’s low slope of prosperity the mountains are so high, but the body rots; how many years will it take to reach utopia, I pity myself as I age unable to squeeze aboard communism’s last train but living in a scorching workshop in a sweatshop (...) time begins to defect it laughs at our memories and enthusiasm as they slip away 75 76 77 78 79 80 81

Knight et al., 1999, p. 88. PRC Bureau of Statistics, 2011. Zhan, 2011; PRC Bureau of Statistics, 2011. Gupta, 2000; Loury and Loury, 2009. Hoff and Pandey, 2004. Solinger, 1999, pp. 100–107; Wang, 2005; Tse, 2016. Walsh, 2017.

2.4 Socioeconomic Conditions for Labor Assertiveness

43

While most of the migrant workers’ protests only concern their direct employer’s wages and working conditions, there seems to be growing resentment about the dagong life in general. In interviews, some workers lamented that “there is no exit” (meiyou chulu) from their dagong life. While their initial aspirations as migrant workers were high, after years of dagong their wages did not substantially improve, and they could not find an alternative path out of the dagong life. Some scholars argue that the high levels of labor shortage in part result from these frustrations with life as a marginalized migrant in the cities. One young migrant worker, for instance, explained that she returned to her hometown because she was “no longer willing to put up with the hardship in the city like [her] father did.”82 The mismatch between workers’ growing aspirations and their ongoing experience of marginalization raises the question of how an authoritarian regime that had relied on labor coercion could justify its strategy. For authoritarian leadership, a rapidly growing economy could deliver more resources that could be used to maintain citizens’ political support. However, if rapid economic growth is accompanied by persistent or growing inequality, the benefits of economic expansion will not be experienced by the excluded social group. Migrant labor drove the expansion of urban economies in China,83 but the ongoing exclusion of migrant workers has prevented them from realizing the benefits of economic growth. The experience of marginalization does not immediately turn migrant workers into political activists, but it increases the risk of labor instability and changes the environment for state–labor relations.

2.4 SOCIOECONOMIC CONDITIONS FOR LABOR ASSERTIVENESS: LONG - TERM CONSEQUENCES OF LABOR COERCION

This discussion suggests that China’s path to becoming the world’s top exporter created a socioeconomic environment for labor assertiveness. Growing resentment and the increasing structural power of workers in post-reform China were the endogenous consequences of the regime’s use of labor coercion as a development strategy. Labor-intensive growth ultimately drained the pool of surplus labor and structurally empowered workers. At the same time, the sustained coercion of migrant workers 82 83

Hong, 2010. “Migrant workers contribute 16 percent GDP growth” 2006.

44

The Political Costs of Labor Coercion

fed their resentments and made them more willing to resist the coercive control. The declining effectiveness of labor coercion as a labor discipline device, in other words, derives from its own developmental logic. These observations raise questions about long-term political costs of coercive labor control in authoritarian regimes. Labor coercion has long-term consequences for both workplace-level disputes and broader regime resilience. While the hukou institution might be specific to China, the underlying patterns of socioeconomic transformations parallel the experiences of many other labor-coercive regimes. There are many historical examples of authoritarian regimes facing a substantial political risk at this critical juncture. For instance, studies find that before the demise of Apartheid, the mining and manufacturing industries were suffering from resistance among Black workers and high rates of labor turnover. As better-educated Black workers entered the labor market, for example, they “were less amenable to the old military-style discipline, barking out orders and wielding the sjambok (stick or whip).”84 Wintrobe (1998) attributes the origins of increased resistance to the rapid expansion of the industries that relied on Black workers. The author concludes that “the growth of the economy under Apartheid is, oddly enough, part of the explanation for the ultimate demise of the system” (p. 193). Seidman (1994) suggests that Brazil and South Africa demonstrate a strikingly similar pattern of “manufactured militance.” Both states pursued economic growth based on “savage capitalism” that prioritizes a favorable investment climate over workers’ welfare and demands (p. 259). The Brazilian industries – particularly consumer-durables – experienced massive growth and an expansion of the industrial workforce. Foreign direct investment in Brazil, unlike China, went more into semi-skilled industries such as automobiles, so the economy began to experience a shortage of experienced workers. The author writes, the very logic of “authoritarian industrialization” empowered workers to challenge the state. The closest example to the experience of China might be newly industrializing countries (NICs) in East Asia. As with China, their economic growth depended on EOI. Studies show that these economies also experienced growing labor shortages after expanding export-oriented industries absorbed the surplus workers: “If early EOI development was based on the mobilization of cheap, disciplined, low-skill labor for export manufacturing,” Deyo (1989) writes, “the very success of such development 84

Lipton (1985, p. 131), requoted from Wintrobe (1998, p. 193).

2.4 Socioeconomic Conditions for Labor Assertiveness

45

eventually eroded the labor supply that made it possible” (p. 24). In South Korea, the wages of workers grew rapidly in the mid-1970s.85 Despite the wage growth, however, Korean workers have become much more discontented about their wage level and economic conditions. The social stigma attached to factory workers as low-class and inferior also contributed to the mounting perception of injustice. While few analyses demonstrate exactly how these changes contributed to labor activism, scholars note that labor conflicts have become more aggressive and organized in the 1980s.86 The common denominator in these cases is the effect of the socioeconomic transformations, which were brought about by sustained labor coercion, on political instability in the authoritarian regimes. These examples, however, do not mean that labor coercion automatically leads to pro-democracy labor activism, even in the presence of labor-intensive growth. They merely point to the increased potential for political instability. The question for post-reform China is, how can the regime manage state–labor relations in this new socioeconomic environment to prevent the emergence of politicized labor activism, and what does this imply for state–labor relations in the long run? While some have argued that migrant workers have increasing potential as political actors,87 others disagree, pointing to the prevalence of apolitical protests.88 But do apolitical protests mean workers are disinterested in political change? Despite the widespread grievances about socioeconomic conditions, has the regime successfully preserved the political loyalty of migrant labor? In the remaining chapters of this book, I examine these questions by analyzing how the regime’s top-down approach and workers’ behaviors and beliefs have co-evolved in this new socioeconomic environment.

85 86 87 88

Song, 1990. Koo, 2001, Chapter 5. Chan and Pun, 2009; Shirouzu, 2010; Chen and Tang, 2013. Chan and Siu, 2012; Lee, 2016.

3 Atomized Incorporation: Regime Response to the Changing Environment

The unstable labor relations and structural changes that drive them are a warning to the regime that state–labor relations might need to be reformed. This is a difficult task for the regime: When an authoritarian regime lacks any political or organizational link to labor, how can it politically co-opt the unorganized group? The CCP has been reluctant to allow labor to organize because it fears radicalization, but without political or organizational ties, how can an authoritarian regime manage labor demands? In this chapter, I show that, since the early 2000s, the regime has evolved away from labor coercion and instead adopted what the study calls “atomized incorporation”: Incorporation of migrant labor into the state’s labor policymaking, but only as atomized members and not as an organized group. The regime has introduced pro-labor policies to strengthen the protection of individual labor rights and tolerated isolated, within-firm labor protests, but efforts to incorporate labor have been accompanied by suppression of any labor mobilization that might be coordinated across firms, regions, or with other individuals or civil groups. Here I suggest that this form of incorporation could be a tool for depoliticizing labor struggles but with its own limitations. By limiting labor discontent to economic concerns and individual workplaces, the central government strives to keep its distance from the growing labor struggles of migrants. The regime has offered channels to workers for making job-specific demands, but not to make broader social and political claims against the state. So what explains atomized incorporation and what are the potential risks? In this chapter, I lay out the theoretical expectations about the conditions for success of atomized incorporation to ensure 46

3.1 China’s Labor Control Strategy

47

long-term political resilience. This discussion will guide the analyses in Chapters 4–7. Atomized incorporation has been implemented in a way that raises questions about why the regime would enact pro-labor policies that could undermine business profits. The alliance between state and business is assumed to generate incentives for labor coercion; so why has the state, both central and local, tried to enhance labor protection despite the negative effects on firms? I analyze the changing relationships between state and business in different industries and conclude that the introduction of pro-labor policies reflected the regime’s declining interest in the old growth model. This meant that labor coercion was no longer part of the strategy for economic growth and the regime began to let individual firms absorb labor contention. Atomized incorporation would not have been possible if the state had not reoriented its development strategy away from export-oriented growth to a more sustainable model of economic development. The findings demonstrate that the regime’s apparently pro-labor approach since the early 2000s does not empower labor in a fundamental way, since the extent to which local governments incorporate labor demands can be adjusted according to the local economic conditions. 3.1 CHINA’S LABOR CONTROL STRATEGY IN A COMPARATIVE PERSPECTIVE

The most widely discussed form of authoritarian labor control in the comparative literature has been “state corporatism.”1 Corporatist arrangements increase workers’ structural dependence on the state and make it possible for the state to control labor. These regimes create unions with leaders who are directly appointed and controlled by the state and make union membership compulsory, giving themselves a monopoly over labor representation.2 By controlling the unions and their leaders, the regime can restrict the labor movement.

1

2

The seminal work by Schmitter (1971, pp. 94–95) defines corporatism as “a system of interest representation in which the constituent units are organized into a limited number of singular, compulsory, noncompetitive, hierarchically ordered and functionally differentiated categories, recognized or licensed (if not created) by the state and granted a deliberate representational monopoly within their respective categories in exchange for observing certain controls on their selection of leaders and articulation of demands and supporters.” Cohen, 1982; Valenzuela, 1989; Collier and Collier, 2002.

48

Atomized Incorporation

State corporatism can take different forms, and authoritarian regimes have varied widely in the political space they allow for labor. Some corporatist regimes are more “inclusionary” toward labor, engaging the working class in political activism and including them in the supporting coalition. Others are more “exclusionary,” restricting labor’s political participation and minimizing their influence in national policymaking.3 Another distinction is who is mobilizing or controlling labor. Collier and Collier (2002) argue that when the party seeks to mobilize labor as a support group – called “party incorporation” – the nature of corporatist arrangements is more inclusionary, as it is in Mexico. When the state controls labor unions directly via bureaucratic and administrative channels – called “state incorporation” – its focus is more on depoliticizing labor, as it is in Brazil. State corporatism, therefore, cannot be viewed as a unitary phenomenon that “is either present or absent,” but should be treated as “a set of dimensions.”4 Although corporatism is discussed as an important mechanism of labor control in authoritarian regimes, in fact, a majority of modern authoritarian regimes do not allow labor to organize. In many authoritarian regimes, workers are either unorganized or only weakly organized even if labor unions exist, as in the case of China. The Chinese regime appears to follow the corporatist model of labor incorporation. The CCP has granted a monopoly to the All China Federation of Trade Unions (ACFTU) to supervise and control all trade unions at the regional, industrial and enterprise levels and control the ACFTU by directly influencing leadership appointments. The formal representation of workers is completely incorporated into the state structure. The ACFTU is not an official government body, but it functions as a quasi-government bureau, which is why some scholars have discussed the role of the ACFTU within the framework of state corporatism.5 Yet, as some scholars have noted, it is difficult to characterize state– labor relations in China as corporatism. The ACTFU does not have any substantial influence over workers that would enable it to function as a corporatist organization.6 State corporatism works as a mechanism to control labor because workers depend on the corporatist union for subsidies and welfare benefits. Leaders of corporatist unions are subject to the state’s control because of “inducements,” in the form of 3 4 5 6

Stepan, 1978, Chapter 3. Collier and Collier, 1979. Chan, 1993; Unger and Chan, 1995. Friedman, 2014.

3.1 China’s Labor Control Strategy

49

career advancement, material gains, or organizational resources.7 While the strength of the corporatist union can vary depending on the political and economic contexts, the key to state corporatism has been to induce workers’ organized dependence on the union and on the state.8 While the Chinese regime’s control over urban workers during the Mao period resembled the state-corporatist approach – except that workers were largely incorporated into the system of work units (danwei) instead of trade unions9 – the Chinese regime in the contemporary era has not empowered the ACFTU in a meaningful way. Scholars have pointed out that trade unions in China have been structurally constrained in a way that prevents them from effectively mediating state–labor relations.10 The way the state designs career incentives for trade union leaders means that the leaders neither perceive themselves as labor leaders nor are they invested in labor issues. The higher-level leadership of the ACFTU typically holds positions in government agencies before and after working at the ACFTU. The system of joint appointments to both the trade union and the standing committee of the People’s Congress also subjects union leaders to a “double identity,” as Chen (2003) has noted, leading to subordination of union leaders to the interests of government bureaus. Because union leaders are structurally constrained from making sustained efforts to represent workers’ interests, this adds to the perception among workers that the official union is not a legitimate organization that can speak on behalf of workers. Therefore, despite the presence of the ACFTU, workers in China cannot be considered to be organized. 3.1.1 Labor Atomization State corporatism, therefore, is not a fitting description for state–labor relations in many regimes where workers are either weakly organized or unorganized. In this regard, Valenzuela (1989)’s discussion of the “market mechanism” of union control – a labor containment strategy that “weakens unions as bargaining agents to a maximum extent” – could be a useful starting point (p. 448). The author suggests that some regimes seek to undermine labor’s collective power by placing restrictions 7 8 9 10

Collier and Collier (1979) analyze constraints and inducements as two major instruments of state corporatism. Cohen, 1982. Walder, 1988. Chen, 2003; Friedman, 2014.

50

Atomized Incorporation

on workers’ collective action, making union membership voluntary, encouraging competition between trade unions, and repressing union leaders. While the discussion is largely confined to the regime’s control of labor unions, the analysis can be applied more broadly to describe state– labor relations where the regime actively undermines workers’ capacity to organize. I use the term labor atomization for the strategy that authoritarian regimes adopt to weaken labor’s collective capacity.11 The regimes that atomize labor do not acknowledge labor unions or other labor organizations as a legitimate mediator of state–labor relations. As the example of contemporary China demonstrates, the existence of labor unions does not mean that labor is organized, in particular if the labor unions do not meaningfully represent labor. Labor atomization has been a common strategy in authoritarian states, particularly in Asia, and is not unique to China. Some regimes allow some limited union representation at the enterprise level, as in the case of South Korea under the military regimes or Taiwan under KMT, but these regimes still forced the atomization of labor by preventing an overarching labor organization from emerging. Under labor atomization, workers remain as individual actors or small groups that are not able to represent a substantial number of workers. Labor atomization in East Asia has often been associated with labor coercion, reflecting the regime elites’ disinterest in gaining labor support in the first place. While discussing state–labor relations in Asian NICs – which largely adopted labor atomization – Song (1990) characterizes them as a system of “constraints without inducements.” Unlike state corporatism, no benefits accrued to the working class, neither to individual workers nor to labor leaders, in exchange for yielding to the political constraints imposed by the state. Since the Asian NICs from the outset viewed labor as an oppositional force that should be excluded from politics, the regimes did not consider co-opting workers to their cause. Thus, there was no incentive for regimes to offer inducements to the working class and labor leaders, nor to use labor organizations as a mediator between the state and labor. The Federation of Korean Trade Unions (FKTU) in South Korea was ordered to follow the state directives but was never offered corresponding benefits that would enable them to function as a corporatist organization. Korean workers, as a result, largely disregarded 11

In the China literature, the term “atomization” is widely used to refer to individualization. For example, Fu (2017) studies atomized action by examining the behavior of individual contenders. In this book, however, I use the term more broadly to refer to the breaking down of people into individuals or small groups.

3.2 Toward Atomized Incorporation

51

the FKTU as a guardian of their interests. Likewise, in Taiwan, while the KMT allowed the Taiwan Confederation of Trade Unions (TCTU) to play a limited role as a mediator, the role remained largely symbolic and did not empower the TCTU to function as a corporatist union.12 Thus, under the atomized control of labor, the working class were both economically and politically marginalized. The Chinese regime in the post-reform period has been clear about not allowing any form of organized labor. The regime’s reaction to labor activism during and after the Tiananmen Square protest in 1989 is a sign that the regime was acutely aware of the political risk of labor starting to organize. Evidence suggests that the Chinese leadership decided to use violence in the crack down on the Tiananmen protest due to the growing presence of workers and independent labor organizations.13 While the central leadership did not immediately repress independent labor organizations, such as the Beijing Workers’ Autonomous Federation (BWAF) – which protested the state’s failure to control inflation and political corruption – their continued engagement raised the alarm for Deng Xiaoping and hardliners.14 Deng remarked that “the movement might spread soon to workers and peasants, as in Poland, Yugoslavia, Hungary, and the Soviet Union,” calling for government action.15 After the Tiananmen protest, some workers still attempted to establish independent trade unions, but these organizations did not survive state repression.16 This suggests that what the ruling elites observed in the Tiananmen Square protest and the collapse of authoritarian regimes in Eastern Europe convinced them to consolidate the suppression of independent labor organizing. 3.2 TOWARD ATOMIZED INCORPORATION

How can an authoritarian regime maintain regime stability when discontent among unorganized workers is increasing? The discussion in Chapter 2 explored how sustained labor coercion has created structural 12 13 14

15 16

Kim, 1993. Pye, 1990. In the initial phase of participation, the ACFTU even encouraged state-controlled unions to “fully support workers in their fight against corruption” (Perry, 1994, p. 85). For a detailed discussion of BWAF and its role in the Tiananmen Square protest, see Walder and Gong (1993). Feigon (1990), p. 153. Re-cited from Perry (1994, p. 84). Many worker activists were arrested and charged with subversion of state power including Li Wangyang, who organized the Shaoyang Workers’ Autonomous Federation and received a thirteen-year sentence.

52

Atomized Incorporation

conditions for growing labor assertiveness. Labor protests in contemporary China are mostly isolated and narrow in scope, so they don’t pose an immediate threat to the regime’s authoritarian rule. There is potential for protests to grow into a broader political movement, however, if the regime loses control over labor. The regime’s reaction to workers’ participation in the Tiananmen Square protest showed that it is aware of the potential threat of a labor movement. As labor unrest increases, so too does the Chinese regime’s focus on preventing a broader labor movement from emerging.17 Because of the earlier legacies of labor atomization and political exclusion, the regime’s options are limited. Gerschewski (2013) suggests that nondemocratic regimes rely on three tactics to maintain political stability: repression, co-optation, and legitimacy. The Chinese regime could respond to the increasing labor unrest and discontent with more repression, but the structural origins of labor unrest suggest that repression alone would not be a long-term solution. When workers are structurally empowered and increasingly resentful, excessive repression can easily backfire. A more sustainable solution would be to co-opt labor by establishing a political connection. Scholars identify political co-optation, particularly through semi-democratic institutions such as parties and legislatures, as one of the most effective ways to reduce labor conflict.18 Comparing the nature of labor incorporation in Mexico and Brazil, Collier and Collier (2002) also argue that the PRI’s more integrative approach to labor was one of the reasons why Mexico was better able to stabilize state–labor relations in the long run. Political co-optation of atomized labor is a challenge, however. Cooptation – “to tie strategically relevant actors to the regime elite”19 – usually requires an organized group, whether formal or informal, with a leadership that has influence over its members. Institutional co-optation of labor requires actors or entities that can represent workers in negotiations with the regime. If the regime does not offer organizational space to workers and permit labor leaders to emerge, it has no opposing party to negotiate with. In other words, a substantive form of labor organizing is a pre-condition for political co-optation of labor. The political environment created by the earlier strategy of the Chinese regime makes it risky for the regime to simply allow labor to 17

18 19

Here, I use the definition of social movement from Tarrow and Tilly (2009, p. 442): “a sustained challenge to power holders in the name of a population living under the jurisdiction of those power holders by means of public displays of that population’s worthiness, unity, numbers, and commitment.” Kim and Gandhi, 2010; Teitelbaum, 2010. Gerschewski, 2013, p. 22.

3.2 Toward Atomized Incorporation

53

organize. First, political co-optation might not be effective, given China’s historically coercive and exclusionary treatment of labor. Labor coercion breeds discontent, and political exclusion enlarges the ideological distance between the regime elites and the working class; yet, with limited channels for bargaining and negotiation, the regime has few political instruments to control labor when labor discontent increases. The ideological divide also suggests that organizations that can substantially represent labor might not be easily co-opted, while those labor organizations that can be co-opted might not attract support from labor. In South Korea, the labor leaders that emerged under military rule formed a strong anti-regime identity; the lack of a political link between the regime and labor leaders, coupled with growing class antagonism, meant that labor leaders could be more independent of the regime’s ideological and administrative control.20 The South Korean example suggests that even if the Chinese regime were to allow labor to organize, these organizations would not be able to mediate state–labor relations in a way that maintains political stability.21 When the regime cannot create an organizational space capable of co-opting labor, labor organizations will be autonomous and beyond state control, which is particularly threatening to the regime. The combination of ideological distance and lack of organizational and political ties means that conditions are established for labor radicalization and any autonomous organizing space, regardless of its original intent, could transform into anti-regime machinery. The long-term consequence of labor coercion in post-reform China is that there might already be a substantial number of aggrieved workers who are structurally empowered to tap into any new organizational resources. The discussion echoes Collier and Collier (2002)’s conclusion that the effects of an early-period labor control strategy are not easy to undo, and generate enduring consequences for state-labor relations. For this reason, scholars and observers have argued that the Chinese regime needs to institutionalize organizational representation of labor. The argument is based on the prediction that, without protection of labor’s collective interests, labor unrest will 20 21

The opposition parties also excluded labor in South Korea during military rule. Both the incumbent and opposition parties maintained a distance from labor. An additional explanation for why the Chinese regime chose to sustain atomization, rather than to politically co-opt – as in the case of private entrepreneurs – or to implement meaningful organizational control, might be that the regime does not perceive migrant labor as a well-endowed social group that is needed for China’s future. While China historically has restricted social organization outside the state system, the state engaged with some organizations that represent, for instance, home owners and entrepreneurs.

54

Atomized Incorporation

continue to rise. For instance, Friedman (2014) argues that due to the regime’s unwillingness to permit organizational space for labor, China is in an “insurgency trap.” 3.2.1 Defining Atomized Incorporation Given these constraints, the Chinese regime should prefer to keep labor atomized. Instead of institutional co-optation of labor, the regime has adopted what I refer to as “atomized incorporation.” Incorporation in political science has carried different meanings and been used in a wide variety of contexts. Some use broad definitions, as in Browning et al. (1984): “the extent to which group interests are effectively represented in policymaking” (p. 25). Others focus on a more specific condition for political incorporation, such as participation in electoral politics, group mobilization, substantial policy benefits, or institutional ties to the state.22 Regardless of breadth, these definitions usually share two features. First, incorporation is a process in which a formerly marginalized group enters the policymaking space and gains policy concessions. Group needs that were previously neglected acquire some level of representation. Second, incorporation creates a channel through which the group can access the policymaking process. It could be an institutional channel, such as enfranchisement or party incorporation, or it could be less formal, for example a populist movement in which the group establishes a direct link with the leader.23 Atomized incorporation is a form of incorporation because (1) it brings the interests of a formerly excluded group into the policymaking process and yields policy benefits for the group and (2) it creates a new channel for claim-making and political participation. At the same time, incorporation is atomized because the channels of claim-making are only available for atomized members of the group, and not the collective. Atomized incorporation does not allow organized groups to participate in policymaking. Rather than engage with a representative organized group, the state gathers information on isolated cases of claim-making, filters the information, and selectively grants policy concessions. Therefore, while bottom-up political participation plays a role, the process of atomized incorporation is largely top down.

22 23

Pribble, 2010; Ramakrishnan, 2013. Mouzelis, 1985.

3.2 Toward Atomized Incorporation

55

For some authors, incorporation cannot be atomized, because incorporation itself implies representation of an organized group’s interests. In the case of labor incorporation, most of the discussion has focused on incorporation of labor unions and their capacity to represent the interests of the working class. For instance, for Friedman (2014), incorporation exists “if workers are able to resolve collective problems and contend with capital within rationalized, legal channels (especially collective bargaining) and if they recognize the legitimacy of their legal union representatives” (p. 20). The study by Collier and Collier (2002) likewise defines the initial incorporation of the labor movement as the state’s efforts to “legitimate and shape an institutionalized movement,” pointing to “state support and approval of the existence of unions as organizations, along with an effort to influence the role of the labor movement” (p. 783). Although the study does not provide a definition of incorporation itself, the discussion of incorporation is based on the regime’s changing relationship to labor unions. While I acknowledge that an ideal case of political incorporation should entail organizational representation, I adopt a broader definition that allows more nuanced variations, in which authoritarian regimes reflect the voices of formerly neglected groups. In the comparative literature on authoritarian survival, it has been assumed that a social group must be politically powerful in order to benefit from state resources. Bueono de Mesquita et al. (2005)’s selectorate theory, for instance, suggests that a social group needs sufficient political leverage to extract resources from the state. An unorganized group often lacks such political leverage. Yet, a growing number of studies reveal that many authoritarian regimes implement comprehensive redistributive policies, even if the recipients are not politically powerful or well organized.24 Using the term “atomized incorporation,” I show that unorganized – and thus politically weak – groups can influence national policymaking in authoritarian states, even if the extent of the influence might not be as great as that of an organized group. 3.2.2 Political Implications Existing studies in the China scholarship have advanced the idea that the Chinese regime atomizes citizens and discourages broad mobilization. This approach may also not be specific to workers alone or the 24

Wallace, 2015; Albertus et al., 2018.

56

Atomized Incorporation

contemporary period.25 As a result, disgruntled citizens often take individual actions, even if these individual actions approximate collective action as they share similar strategies and origins.26 When collective action does arise, it has also been stressed that Chinese citizens understand the informal rules that legitimize their collective behavior in the eyes of the state, which discourages broad mobilization.27 Studies have stressed the advantages of atomization in creating a relatively safe space for popular claim-making that does not threaten the regime’s one-party rule. Yet as I detailed and further elaborate in this chapter, it will be inaccurate to argue that the Chinese regime adopts the same approach to all social groups. In relations to migrant labor, there was a clear shift in the regime’s approach over time in the post-reform period. At the same time, some other social groups, such as private capitalists for instance, that gained political significance in the post-reform era were integrated into the state in a much more institutionalized way via the Party and Congress.28 The regime’s approach to different social groups relies on some combination of repression, accommodation, and use of culture, but each has unique advantages and disadvantages. While the previous section dealt with why the regime adopted atomized incorporation in relations to migrant laborers, this coming section discusses what the regime might be gaining and risking by adopting atomized incorporation. Atomized incorporation provides channels of claim-making only to small groups and individuals, which means that it accompanies proactive suppression of broad mobilization. Because these social groups seem to receive resources from the regime but also are subject to repression, the workings of atomized incorporation appear be contradictory; yet, the repressive capacity of the regime to undermine broad mobilization is the key that allows accommodation of claim-makers. Atomized incorporation also has its own limits that arise from its conflicting elements. In the extant literature, there has been relatively little discussion of the inherent conflict in allowing disgruntled citizens to express their discontent as a small group while staying unorganized, both theoretically and empirically. This chapter lays out the theoretical foundations before probing them empirically in Chapters 4–7.

25 26 27 28

Zhou, 1993. Zhou, 1993; Fu, 2018. Perry, 2002; Cai, 2006; Perry, 2010; Li, 2018. Dickson, 2017; Hou, 2019; Dickson, 2000.

3.2 Toward Atomized Incorporation

57

3.2.2.1 Material Co-optation Material co-optation is perhaps the most widely discussed benefit of allowing citizens to make small claims.29 It is targeted at those who might pose a threat, granting them material benefits to increase their support for the regime. It is suggested that when disgruntled citizens make claims about short-term economic grievances and receive benefits, they will demonstrate higher support for (or at least less resistance to) the status quo and become disinterested in challenging the regime’s political rule.30 For instance, Lorentzen (2017) suggests that toleration of limited protests might function as a “controlled burn” in that “once aggrieved groups have called attention to their problems and had them addressed, they will become less interested in participating in anger-venting riots or even broader challenges to the state.”31 This effect could be similar to increasing what Gerschewski (2013) refers to as “specific support” for the regime – the “quid pro quo of the fulfillment of the demands.”32 The key is that for the regime, popular protests reveal which citizens or groups have the most severe grievances; the regime can use this information to undermine the revolutionary threat from below. There has been limited attention paid to the type of information that popular protests provide to the regime. The extent to which material co-optation works should depend on the kind of information. Do these protests arise when grievances are most severe? Studies of social movements and contentious behavior have highlighted that protesters are not always driven to protest because of grievances; their collective dynamics and strategic decision-making matter.33 Individual-level studies of workers’ contentious behavior also suggest that factors other than grievances might be more critical in determining protest behavior, for example, the presence of a social network,34 legal knowledge,35 or perception of group efficacy.36 This means that atomized incorporation might not guarantee 29

30 31 32 33 34 35 36

Material co-optation is different from political co-optation, which I used to describe the state’s incorporation of organized labor. Political co-optation denotes absorption of potential opponents into the ranks of regime elites, largely by establishing an institutional link. Material co-optation here describes the granting of material benefits to those who are a potential threat, in return for their loyalty. Scholars increasingly use the term co-optation to refer to material co-optation. Wallace, 2014; Lorentzen, 2017. Lorentzen, 2017, p. 476. Easton (1979), p. 268. Quoted in Gerschewski (2013), p. 20. McAdam, 1982, 1988; Chong (1987) and Lichbach (1998). Becker, 2012, 2014. Wong, 2011. Cai, 2006.

58

Atomized Incorporation

that the most aggrieved citizens have the access to claim-making channels. Co-opting protesters might still be important – because, after all, they are the ones who are able to engage in collective action – but it might not secure the group’s political loyalty. Even when grievances contribute to the incidence of claim-making, not all grievances matter equally. As I show in Chapter 4, some types of grievances could be more detrimental to the regime than others. If shortterm economic grievances are not necessarily attributed to the regime, for instance, atomized protests that arise out of those grievances might not provide the most useful information. This undermines the regime’s ability to detect those who would join anti-regime mobilization if not for the opportunity to engage in atomized protests. For migrant labor in China, there is no clear evidence that those who are most dissatisfied with the political status quo are the ones who are engaging in atomized protests. To understand whether atomized incorporation can be a tool of material co-optation, it is paramount to investigate under which conditions atomized protests occur. As I show in Chapters 5–7, the type of information provided by atomized protests varies according to the social, economic, and political environment that shapes the collective action problem on the ground. It also means that the political implications of atomized incorporation are not necessarily fixed; understanding the variations helps us to reach a more nuanced conclusion about what protests imply in an authoritarian context. 3.2.2.2 Depoliticization Atomized incorporation could also be more than material co-optation if it can successfully depoliticize labor discontent in the long run. Material co-optation might not work as a long-term device to secure citizen loyalty if it cannot provide a credible signal about the regime’s commitment to citizens’ future well-being. The extant theories of democratization postulate that the revolutionary threat from below occurs as a result of this commitment problem.37 Even if the short-term economic gains placate the disgruntled citizens in the short term, there is no guarantee that their underlying grievances would not surface again. Yet, if atomized incorporation can successfully depoliticize citizen discontent – by changing the citizens’ cognition about the nature of the problem – it could be a more effective tool for political resilience. 37

Acemoglu and Robinson (2009), for instance, assume that without institutional changes elites would not be able to demonstrate credible commitment to their promises of future redistribution.

3.2 Toward Atomized Incorporation

59

Atomized incorporation could affect citizens’ cognition. First, atomization itself has been used by governments and political leaders to dissipate threats from disgruntled groups, particularly in the face of popular mobilization. The “divide-and-rule” technique has been documented to be effective in creating divergent interests within a social group and weakening their collective capacity.38 Atomization could slow the formation of what Freire (1968) calls “critical consciousness,” the cognitive process that enables marginalized citizens to critically understand and analyze their social position.39 Without an organizational space to offer a political framing of citizens’ everyday grievances and connect them with one another, atomized citizens could find it difficult to generalize their own struggles. When citizens are isolated and lack information on the nature and extent of mass grievances in the society, they might not only be discouraged from initiating collective action for fear of its failure,40 but they are also denied the breadth of perspective that would allow them to perceive the political nature of their problems. Atomized incorporation could enhance the effectiveness and sustainability of atomization by creating positive inducements for disgruntled citizens to focus on isolated and thus largely apolitical framing. I call this a particularistic framing, as I explain more in detail in Chapter 4. The balance of carrots and sticks is important for isolating those actors with the most radical interests, by rewarding those with moderate claims for sticking with a particularistic framing. At the same time, when citizens are exposed to, and learn to use, a particularistic framing to describe their struggles and contention, they could be discouraged from finding an alternative framing that broadens the scope of their understanding. As Iyengar (1994) shows, when citizens are exposed to a particularistic framing of social issues, they are less likely to see the issues’ systemic nature and hold politicians accountable. Thus, atomized incorporation could change both citizens’ behavior and cognition by incentivizing them to make apolitical demands and, subsequently, discouraging politicization of their discontent. There is an ongoing debate about the extent to which Chinese workers politicize their grievances in the current political context. On the one hand, some evidence suggests that atomized incorporation has successfully deradicalized protest behavior in China. The Chinese regime has created informal rules around what is accepted and what is restricted in 38 39 40

Koopmans, 2004b; Acemoglu et al., 2004; Vanhuysse, 2006; Chen, 2018. Freire, 1974. Kuran, 1991; Lohmann, 1994b.

60

Atomized Incorporation

terms of citizens’ contentious behavior. Scholars argue that Chinese citizens understand these informal rules and internalize them.41 Li (2018) argues that the informal rules have contributed to the Chinese regime’s political stability, because they not only create a safe space for citizens to air grievances but also make protesters self-censor their behavior. These informal rules seem to have shaped citizens’ preferences for a particularistic framing. Because contenders believe in the effectiveness of atomized action, they perceive broad mobilization as futile. Contenders are trained to frame their grievances in a narrow way, which Chen (2018) calls “particularization.” Protesters often display their commitment to a particularistic framing because they believe in its efficacy. For instance, in her study of laid-off workers, Lee (2007a) finds that workers believed in the efficacy of atomized contention: A worker is quoted arguing, “It is no use coordinating with retirees from other factories, because some firms are more generous or stronger financially, and their workers get more subsidies” (p. 236). In this case, workers genuinely perceived broad mobilization to be counterproductive. In this environment, citizens seem unlikely to be exposed to a broad framing that attributes their struggles to the political status quo, and – even if they come in contact with one – unlikely to be persuaded by it. Because there is no sign that migrant workers use political framing during collective action, some scholars suggest that migrant workers are not politically conscious. On the other hand, microlevel studies of migrant workers’ contention have found signs of politicized grievances on the ground. Scholars suggest that migrant workers have increasingly displayed class consciousness during labor disputes, using language that demonstrates their understanding of the subordinate class position.42 This means that workers might not be solely focused on their job-specific concerns. Chan (2010b), for instance, finds that workers join protests for various reasons, and wage demands are not always central. Chan observes, “curiosity, prolonged discontent, and a sense of unfairness seemed to be more important factors” (p. 93). While migrant workers use apolitical framing and demand job-specific gains, this does not necessarily mean that they lack politicized grievances.

41

42

Perry (2008) argues that the tradition of protests about socioeconomic security dates back to pre-Modern China. The author suggests that the rise in popular protests is not necessarily a political threat to the regime, because they reflect citizens’ “rules consciousness.” See Li (2010a), Perry (2010), and Lorentzen and Scoggins (2015) for the scholarly debate on the “rights” and “rules” consciousness. Chan and Pun, 2009; Chan, 2012a.

3.2 Toward Atomized Incorporation

61

Since the Honda protest in Nanhai, studies have found that the new type of “offensive” and “interest-based” protests – in which workers demand wages and benefits beyond the legally guaranteed minimum – might signal that the seemingly apolitical protests of migrant workers could evolve into something more political.43 This means that atomized incorporation also entails risks and limitations for the regime. Atomized incorporation could fail to depoliticize labor grievances for a few reasons. First of all, atomized channels might not be sufficient to contain workers’ discontent within the firm if workers can access other sources of information. If workers perceive that similar issues are experienced elsewhere, it makes it easier for them to see a generalized pattern among the individual hardships. To maintain the framing that workers’ struggles are due to job-specific contexts, the regime also needs to restrict workers’ access to channels that might give them information about the prevalence of the problem. In this regard, the prevalence of atomized contention could be counterproductive at some point. If atomized protests are too common, it not only lowers the coordination costs among different groups of workers44 but also reveals information about the pervasiveness of the issue. As I demonstrate in Chapter 4, workers’ direct experiences as migrants affect their attribution of blame and politicize their struggles. Atomized incorporation could also be a politically risk strategy if those interested in a radical change appropriate the tolerated space of contention. Studies on the letters and visits (xinfang) system have shown that, despite the system’s original intent to discourage collective action, contenders appropriated the institution to engage in – often disruptive – collective action.45 Marginalized citizens with limited resources create opportunities for contention by appropriating existing institutions for their goal. If the regime does not detect and punish those workers, activists, or organizations with anti-regime identities, they could appropriate the space for atomized contention and develop it into a platform for anti-regime movement. In South Korea, for instance, the Labor Management Council under the Chun military regime was originally created to absorb labor discontent at the firm level in a depoliticized way but was increasingly used as an organizational resource for the anti-regime labor movement. Despite the Council’s weak capacity to represent labor 43 44 45

For instance, see Chen and Tang (2013); yet, my analyses of interest-based protests in Chapters 6 and 7 demonstrate otherwise. Kuran, 1991; Lohmann, 1994a. Chen and Xu, 2012, Chapter 4.

62

Atomized Incorporation

demands independently, it played an important role during labor mobilization, as those workers with a pro-democracy agenda took control of its organizational resources. Thus the regime’s capacity to monitor and punish defectors remains the most critical factor. Studies and recent observations show that the regime has not always succeeded at isolating workers proactively. Fu (2018), for example, argues that workers’ contention seems atomized in appearance, but many labor organizations have been coaching them behind the veil, potentially contributing to a collective consciousness. Some labor NGOs evolved into “movement-oriented” organizations that encourage collective bargaining to resolve labor disputes,46 and some of them supported workers’ collective action. Many of these NGOs, not surprisingly, were shut down by the government during large-scale crackdowns in 2015 and 2017, but this points to the risk of atomized incorporation.47 In Chapters 4–7, I examine these potential benefits and limitations of atomized incorporation in contemporary China. 3.3 ATOMIZED INCORPORATION IN POST-REFORM CHINA

In this section, I explain the elements of atomized incorporation in the regime’s approach to migrant labor and describe how and why the state– labor relations have evolved away from labor coercion. The elements demonstrate that, instead of suppressing labor conflicts altogether, the regime has been encouraging firm-level resolution of labor conflicts. The strategy to an extent parallels “enterprise unionism” in Japan. Enterprise unionism was a strategy to absorb growing labor unrest and discontent at the firm level by allowing the establishment of enterprise-level unions, while undermining their horizontal linkages with industrial or regional unions. The military regime in South Korea also obliged firms to establish the Labor Management Council to discuss firm-level grievances as labor unrest grew. Rather than resolving labor conflicts by bringing them into the political arena, these authoritarian regimes found a different way to contain labor demands and suppress workers’ collective capacity to challenge the regime.

46 47

Chen and Yang, 2017; Franceschini and Lin, 2019. These attempts at external mobilization and what it implies in the context of atomized incorporation are explored in Chapter 8.

3.3 Atomized Incorporation in Post-reform China

63

3.3.1 Migrant Labor Enters Labor Policymaking Migrant labor was largely excluded from labor policymaking until the early 2000s. Despite the widespread reports of labor abuses and exploitation of migrant labor in the 1990s, for the most part the central government did not engage. The central directives on migrant labor were mostly concerned with controlling and limiting the “blind flow” (mangliu) of migration. Migrant workers were not officially acknowledged as part of the working class, and the two major labor-related organizations – ACTFU and the National Bureau of Labor and Social Insurance – did not consider migrant labor to be part of their constituency. Beginning in the early 2000s, however, the central government began to issue policy documents that specifically addressed how to protect migrant workers, marking what some scholars have called a “paradigm shift” or “a turning point” in Chinese labor policy.48 This is when migrant labor started to have a central role in the Chinese state’s labor policymaking.49 In their first document of 2003, the State Council urged local governments to abolish any hukou-based discrimination in the labor market, to improve living and working conditions of migrants, and to protect migrants’ legal rights.50 In the same year, the central government also abolished the “regulation on detention and repatriation,” which was mainly used to force unregistered and unemployed migrants to return to their place of origin. Next the central government produced a series of policy documents on labor rights and public welfare for migrants, such as education for migrant children, vocational training, and wage arrears. The number of times that the term “peasant workers (nongmingong)” appeared in the title of People’s Daily shows the rising importance of migrant workers during the mid-2000s (see Figure 3.1). In 2006, the State Council issued “Several opinions of the State Council on resolving problems of rural migrant labor,” the most comprehensive policy document to date on migrant labor. This document addressed a wide range of issues concerning migrant labor, including low wages, wage arrears, discrimination, labor rights, job training, social protection, education, land rights, and hukou reform. The document reflects a shift in the attitude of the central government toward migrant labor, in its characterization of migrant labor as a major component of industrial 48 49 50

Froissart, 2005; Cui, 2007; Chou and Wen, 2007. Ngok and Cheng, 2010. China State Council, 2003.

Atomized Incorporation

0

50

Number 100

150

200

64

1995

2000

2005 Year

2010

2015

FIGURE 3.1 The number of mentions of “peasant worker” in the title of People’s Daily

labor. This was the first time that the central government acknowledged migrant workers’ status as “worker (gongren)” rather than as peasant workers who have temporarily left their farmland. It also showed the central government’s willingness to officially recognize migrant workers’ contributions to China’s industrialization, rather than portraying them as unruly floaters that needed to be controlled. The National Bureau of Labor and Social Protection – which had been dubbed the “Labor Bureau of Urban Workers” for its strong urban bias – also started to address formally the job market issues of migrant workers.51 These changes marked the point when the interests of migrant labor achieved meaningful status in the state’s labor policymaking agenda. 3.3.2 Introduction of Pro-Labor Laws The central government has enacted various labor-related laws to protect labor rights and improve the bargaining position of low-wage workers. The Minimum Wage Regulations were enacted in 2004, greatly 51

Ngok, 2011, p. 254.

3.3 Atomized Incorporation in Post-reform China

65

strengthening the implementation of the Minimum Wage Law.52 In 2007, the Chinese regime enacted three major labor laws. First, the Labor Contract Law (laodong hetongfa) confers a considerable degree of labor protection, obliging firms to sign a labor contract and pay benefits. Unlike the Labor Law of 1994 – which failed to address labor rights of migrant workers – the Labor Contract Law is known to target the improvement of migrant workers’ welfare benefits and working conditions.53 Second, the Employment Promotion Law (jiuye cujinfa) guarantees equal employment and labor rights for rural and urban workers.54 Finally, the Labor Dispute Mediation and Arbitration Law specifies the procedures for labor dispute resolution, making it easier for marginalized workers to seek redress via institutional channels. There is evidence from various sources that these laws were not empty initiatives, although their implementation has been far from ideal. For instance, survey data show that firms thought that the implementation of the Labor Contract Law was strict. According to the People’s Manufacturing Firm Survey in 2009, the majority of firm managers (95.9%) believed that the Labor Contract Law was strictly enforced.55 Data from multiple surveys, conducted between 2009 and 2013, also reflect firms’ perceptions that the new labor laws have contributed to growing labor costs and posed a serious business challenge.56 Ngok (2008) writes that, compared with the Labor Law of 1994, the new Labor Contract Law prescribes the punishment more clearly, not only for the employers who violate the law but also for the government officials who are responsible for enforcement. One of the most visible outcomes of these pro-labor laws has been rising labor costs, particularly for labor-intensive sectors in export-oriented regions. Studies find that as a result of stronger labor protection, laborintensive firms have become more likely to default on bank loans and decrease investment.57 Observational evidence shows that more than 52

53 54

55 56 57

See Du and Pan (2009) and Huang et al. (2014). Although the Minimum Wage Law was enacted in 1994, only a small number of cities had participated in the minimum wage scheme; cities were reluctant to increase the minimum wage and enforce that increase. The 2004 regulations substantially increased the geographical coverage and the level of local minimum wage. Kahn and Barboza, 2007; Ngok, 2008. Article 31 of the Employment Promotion Law stipulates: “Rural workers who enter the cities for employment shall enjoy the same labor rights as urban workers. No discriminating restrictions may be placed on rural workers in the cities.” Gallagher et al., 2015. Chen and Yang (2010); CMAHK, 2008, 2010, 2011, 2013. Lu et al., 2015; Yu and Liao, 2017.

66

Atomized Incorporation

1,000 shoe and accessory factories went bankrupt in the PRD region as a result of improved labor protection.58 With increasing local minimum wages and decreasing labor flexibility, firms in labor-intensive sectors have also reduced employment.59 In the YRD region, labor-intensive firms reported that their labor costs increased by 10–20 percent as a result of the new labor legislation, while skill-intensive firms did not report such a substantive change.60 These findings suggest that firm managers’ reports of the labor laws’ effects are not merely biased perceptions. These new laws and policies have generated some positive effects for migrant workers. Most notably, the Labor Contract Law has led to the formalization of low-wage workers’ status, particularly for migrants, by increasing the number of migrant workers who hold a signed labor contract.61 Because of migrant workers’ unstable residential status, employers typically did not provide them with a written contract. This led to abuses of labor rights, because workers could not use official channels to resolve labor disputes if they could not prove the employment relationship. Access to signed labor contracts meant that migrant workers became more likely to take part in the social insurance scheme and less likely to suffer wage arrears.62 Some studies also find that the new minimum wage regulations increased migrant workers’ wages in export-oriented manufacturing firms in PRD.63 3.3.3 Legal Mobilization These findings are not evidence that the central and local governments have actively invested in top-down monitoring and enforcement of labor laws. Studies have found that the labor inspection system in China is fragmented and decentralized, hampering the establishment of an effective inspection model. While the labor inspection division was upgraded to a bureau in 2008, the bureau does not have substantive control over 58 59 60 61

62 63

Chung, 2008. Huang, 2012; Huang et al., 2014. “The effect of the new labor contract law on labor costs in different types of firms (xinlaodonghetongfa dui butongqiue leixing gongzichengbende yingxiang)” 2008. According the China Urban Labor Survey, the percentage of migrants with a labor contract increased from 36.7 percent in 2005 to 59.8 percent in 2010. During the same period, the share of urban local workers with a contract increased from 73.1 percent to 80.2 percent. The survey data also reveal that the percentage of migrants and urban local workers with a contract did not increase between 2001 and 2005, suggesting that the increase was due to the new labor legislation. For a detailed analysis, see Gallagher et al. (2015). Li and Freeman, 2015. Sun and Shu, 2011; Ye et al., 2015.

3.3 Atomized Incorporation in Post-reform China

67

the labor inspection branches at the local level and lacks the resources and manpower to initiate frequent top-down inspections.64 As a result, a majority of labor inspections were initiated by workers’ complaints and petitions.65 In other words, the enforcement of labor laws has relied on “fire alarms” triggered by workers’ reports of violations, rather than “police patrols” by labor inspection bureaus.66 This “fire alarm” mechanism of enforcement relies heavily on worker mobilization. The central government has launched large scale campaigns to disseminate legal information and encourage workers to use “law as a weapon,” as part of its bigger project to establish rule of law as a mode of governance.67 The rhetoric of “legal rights” has been central to the regime’s discourse on migrant labor issues. In particular since labor laws and regulations were passed in the early 2000s, the regime has actively promoted the protection of migrant workers’ legal rights and urged workers to acquire legal knowledge. In line with the central directive, the ACFTU has provided legal training and legal aid for migrant workers, attributing workers’ workplace experiences to their lack of legal knowledge. As a result of legal mobilization, workers’ legal knowledge and rights consciousness have increased. In interviews, many firm managers argued that one of the most important changes from the new labor laws has been workers’ increasing consciousness about their legal rights, and the tendency to defend them. For instance, a firm manager of a machine part factory in Nantong said:68 The biggest change after the new labor law, of course, would be the substantive increase in labor costs, particularly because of the social insurance payment. Now you need to pay the social security funds even for outside workers. An equally important change, however, is in workers’ attitudes. They are much more conscious of their own legal rights and even try to use their legal knowledge to get economic compensation. In the past, when they wanted to quit, they just quit. Now they are waiting for a chance to get economic compensation, taking advantage of the protective legal clauses.

While migrant workers lag behind urban workers in legal knowledge due to their more limited schooling, studies find that media coverage helps 64 65 66 67 68

Zhuang and Ngok, 2014. Ibid., p. 575. McCubbins and Schwartz, 1984; Gallagher, 2017. Gallagher, 2005; Gallagher, 2006; Wang, 2015. Interview 12-2012.

Atomized Incorporation

68

them to learn about labor laws.69 The regime has encouraged the media to report stories of migrant workers who use the law to resolve their labor issues. Wage arrears – one of the most prevalent violations of labor rights for migrant workers – have received considerable attention from the media, after then-Premier Wen Jiabao initiated a campaign against it in 2003. Many newspapers have designated sections on labor issues and have carried stories of migrant workers who succeeded in using legal channels to recover unpaid wages. TV shows that deal with legal issues – such as Legal Report (jintian shuofa) and Law Online (fazhi zaixian) – also frequently report on stories of migrant workers who used the law to resolve their grievances. As Stockmann and Gallagher (2011) argue, these stories mostly deal with successful cases that were decided in workers’ favor, in order to enhance citizens’ trust in the legal system. The central government’s active mobilization of migrant workers as bottom-up enforcers of labor laws has given workers a new tool to address grievances and correct the behavior of firms and local government officials. Migrant workers, who had been considered outsiders by firms and local bureaucracies, have acquired a political lever to enhance the implementation of labor laws. Gallagher (2017), therefore, suggests that legal mobilization should be classed as a form of political participation.70 Although this is not to suggest that migrant workers are always successful in their bid for legal rights, legal mobilization has created a stronger incentive for firms to abide by the labor laws. Even with only limited topdown monitoring, more firms comply with labor laws because they are concerned about workers reporting violations. For instance, when asked why the firm had abided by labor laws even when there was no top-down inspection during the previous year, one manager of an apparel factory in Shanghai said, “Workers know everything these days. They can just call 12333 [the local labor bureau hotline]. If workers file a complaint, the labor bureau has to come out for an inspection. It is better to avoid that risk.”71 A panel assembly factory in Shenzhen echoes this shift, suggesting that “workers have become more likely to file a complaint for a violation of labor rights. If the labor bureau receives a complaint, they must come out for an inspection. Even one case could tarnish our reputation.”72 As will be discussed in Chapter 7, however, the extent to 69 70 71 72

Gallagher, 2017, p. 121. Ibid., pp. 115–118. Interview 07-2012. Interview 05-2012.

3.3 Atomized Incorporation in Post-reform China

69

which such bottom-up mechanisms work is heavily dependent on the firm’s characteristics and the local economic context. Legal mobilization, while generating positive outcomes for migrants, is still atomizing. Since the labor laws only protect individual workers’ rights, promoting the use of legal channels fractures state–labor relations into individual interactions. As scholars have argued, the Chinese regime’s reliance on rule of law to deal with labor conflicts means that what could be collective cases are broken into individual ones.73 An emphasis on individual labor contracts as the basis for claim-making prevents workers from advancing their claims collectively. The number of workers per labor dispute accepted by the formal legal channels, for example, decreased consistently, while the number of union organizations with a legal aid center has increased since the early 2000s.74 Trade unions have largely functioned as mediators of the atomized state–labor interactions. When workers face a labor dispute, trade unions intervene to provide legal knowledge and guide workers to utilize legal channels, to minimize the chance of collective action. Non-government labor organizations are also given incentives to encourage workers to resolve their grievances via legal channels instead of via collective action. Existing research on the role of NGOs in China confirms this argument: a survey of twenty-eight labor NGOs in China, for example, finds that twenty-four out of twenty-eight list “legal services” as their main activity.75 3.3.4 Tolerance of Atomized Protests Another feature of atomized incorporation has been the regime’s growing leniency toward instances of labor protest that take place outside the institutional channels of claim-making. The Chinese regime’s leniency toward labor protests is most visible in two areas. First, the central government has been cautious about resorting to police repression to manage labor protests. This partly reflects the central government’s changing strategy toward mass incidents in general. For instance, the Ministry of Public Security issued “Public Security Bureaus’ Handling of Public Order Incidents” in 2000, articulating three principles for handling mass incidents: (1) preventing conflicts from intensifying; (2) cautious use of police and coercive measures; and (3) cautious use of police weapons and equipment. 73 74 75

Friedman and Lee, 2010; Friedman, 2014; Gallagher, 2017; Chen and Gallagher, 2018. Friedman, 2014, pp. 49–50. Xu, 2013.

70

Atomized Incorporation

The head of the Public Security Bureau of Guangdong Province reemphasized the state’s stance toward police repression in 2011, describing the principle as “three cautious use” (san ge shen yong): cautious use of police force, cautious use of weapons, and cautious use of compulsory measures. Evidence suggests that individual labor protests in particular have been treated with increasing leniency since the mid-2000s. A growing number of Chinese policymakers and cadres have characterized labor protests as “economic” protests that are the inevitable by-products of marketization. During the National People’s Congress in 2011, for instance, Zeng Qinghong – a national legislator and the general manager of the Guangzhou Automobile Group – submitted a proposal to legalize labor strikes.76 Chen Wei-guang – the former secretary of Guangzhou Trade Union – also remarked in 2013 that “many strikes are reasonable and the workers are just trying to get their rights.”77 The regime recommends police intervention during labor protests if there is violent and extreme behavior that might disrupt public order. The Futian district government in Shenzhen, for instance, lays out the role of its public security bureau during workers’ collective action as follows: After large-scale labor mass incidents take place, the public security bureau should immediately send personnel to maintain order; if workers adopt extreme measures such as collective petitioning, street demonstrations, road blockage or blockage of government office gates, the public security bureau should take measures according to the law and prevent the incident from escalating.

The likelihood of police intervention during collective action, therefore, is highly dependent on the tactics the protesters use. Protests in the form of within-firm labor strikes are substantially less likely to attract police intervention because they are not considered to be as disruptive. Most cases of police repression occur when the incidents “seriously threaten the public order,” as prescribed in the Emergency Response Law enacted in 2007. Chen Wei-guang’s interview after the 2011 strike wave in Guangzhou – in which a large number of automobile factories experienced labor strikes demanding higher wages – further illustrates the standard for defining “acceptable” labor protests:78

76 77 78

China Labour Bulletin, 2011. Friedman, 2014, p. 67. Lüthje et al., 2010, pp. 4–5.

3.3 Atomized Incorporation in Post-reform China

71

[W]e believe that the demands of the [striking] workers are justified. But we hope that such economic disputes do not develop into political incidents and will not disrupt social order. This is our bottom line. For this reason, we ask workers not to use illegal means in the struggle for their rights and interests. We have to say that our strikes have been very orderly, there were no walkouts from the factories to the streets, no destruction of machinery, no playing of politics. Everything remained in the framework of disputes within factories. [. . . ] In all of the more than 60 conflicts in Guangzhou this summer, negotiated settlements were achieved. [W]e can proudly say that in Guangzhou no striking worker was dismissed and not one worker was arrested by the police, although the strikes included tens of thousands of workers.

The quote illustrates that as long as labor protests occur within individual firms, demand only economic – and not political – gains, and avoid utilization of violent or disruptive tactics, even those protests that involve thousands of workers could proceed without any police repression. The statistical evidence that supports this is in Chapter 5. Scholars and labor activists have noted that the official media is more open to reporting labor protests.79 For workers, the media is a major tool for gaining third-party support and enhancing the chance of success. The growing media coverage and the state’s pro-labor attitudes, Elfstrom and Kuruvilla (2014) argue, could have led to “cognitive liberation” as workers perceive a growing number of political openings. The discussion implies that the socioeconomic changes discussed in Chapter 2 also had an indirect effect on increasing labor assertiveness by creating a new incentive for the regime to appear pro-labor. 3.3.5 Deterrence of Organized Labor The Chinese regime’s pro-labor attitudes, however, do not imply that the regime has become less likely to use repression overall. As the logic of atomized incorporation predicts, along with the pro-labor policies, the regime has maintained and in some areas intensified selective repression of individuals and organizations that could unite these atomized claim-makers. Cases of labor protests at SOEs in the 1990s and early 2000s showed that the central and local authorities would frequently arrest protest leaders and activists on the charge of “subverting the state power.” When workers at the Ferro-Alloy factory in Liaoning Province protested against corruption and unpaid benefits in 2002, two protest

79

Chan, 2001; China Labour Bulletin, 2012.

72

Atomized Incorporation

leaders received sentences of seven and four years on charges of subversion. Cai (2006) argues that selective punishment of protest leaders during the SOE reform helped the regime to suppress further collective action and to limit the scope of protests.80 Labor protests by migrant workers have also risked repression of protest organizers and labor activists, but few have faced the charges of serious political crimes, such as subversion of the state power, that were levelled during former SOE workers’ protests. This might be because protests by laid-off SOE workers were more likely to target the state and thus were regarded as more politically threatening than those of migrant workers.81 Arrests of protest leaders and labor activists in private industries have usually been made on charges of “picking quarrels and provoking troubles” and “gathering a crowd to disturb a social order.” State repression has been harsher for labor activists and NGOs than for leaders of firm-level protests. Protest leaders might have the resources to mobilize workers within a firm but can rarely maintain their mobilization capacity beyond the factory. By contrast, labor activists and NGOs can assist a large number of workers from different factories and from different regions, so they have the potential to initiate a broader labor movement. Involving labor activists in protests could transform a onetime and within-factory labor protest into a more persistent, cross-firm alliance. Therefore, selective repression of these individuals and organizations has been a critical element of atomized incorporation. While the regime has become more willing to acknowledge individual labor rights, it has to maintain high levels of selective repression of demands that have the potential to consolidate collective labor rights and give them a political voice. Internalizing the costs of participating in collective action, labor activists have self-censored and avoided taking part in workers’ protests. During my fieldwork, I encountered NGO staff advising workers against initiating strikes. In one case, when workers on a strike sought advice from the NGO, one member of staff refused to help, saying “If you are already on strike, we cannot help you. You should have come to us before

80 81

Cai, 2006, pp. 90–100. Even though SOE workers mostly demanded economic gains such as unpaid wages and benefits, their main target was the state because the workers were protesting against the state’s abandonment of the social contract under socialism. The economic demands were in many cases related to corruption at the state-owned company, inevitably leading to protests against party cadres.

3.3 Atomized Incorporation in Post-reform China

73

going on a strike.” Scholars and observers agree that the harsh political environment has limited the scope of labor NGOs.82 When labor activists from several different NGOs decided to advise striking workers at the Yue Yuen factory protest in 2014 – which involved 300,000 to 500,000 workers at a shoemaking factory in Dongguan – it was considered as an unusually bold move. Not surprisingly, the labor activists who got involved were intimidated and arrested. Lin Dong – a labor activist from the Chunfeng NGO – was stopped by the local police on the way to the strike scene and was eventually arrested for “picking quarrels and provoking troubles,” after the strike had ended.83 Some labor NGOs managed to tap into the area of collective bargaining and transitioned themselves to “movement-oriented NGOs”: They have not survived the state repression either, as I detail in Chapter 8. Labor NGOs and activists have operated in a gray zone, testing the limits of how much labor activism the regime will tolerate. Labor NGOs keep a distance from workers’ protests and instead look for other ways to assist workers. Some labor NGOs trained workers in strategies for collective bargaining, believing that it might be seen as less politically dangerous and could contribute to a substantial improvement in workers’ welfare. Han Dongfang, the founder of Hong-Kong–based labor rights organization China Labor Bulletin, for example, explained his 2005 decision to pursue collective bargaining as the organizational goal in the following way: Ten years ago was a very different time for labor in China when it was still regarded as one of the most sensitive issues in the country. [. . . ] I believe that this sense of political sensitivity was heavily affected by the impact of the Solidarity movement in Poland in the 1980s, which was the beginning of the end of the Communist regimes in Eastern Europe. [W]e decided to be brave and make a strategic decision to change the direction of our work: instead of trying to put an end to the Chinese Communist regime, the next stage in China’s labor movement should focus on introducing collective bargaining into workplaces[. . . ] In this way, we would be able to depoliticize an overly politicized labor issue, transforming it into what it was originally: a socioeconomic issue. We decided, instead of seeing the Communist regime as an enemy, we would see them as a potential partner who was still fearful and therefore hostile towards the labor movement, a partner who needed time to build up the confidence to let go of its unnecessary fears.84

82 83 84

Friedman and Lee, 2010; Xu, 2013. Roberts, 2014; Hui, 2014. Han, 2015.

74

Atomized Incorporation

His remarks convey his conscious decision to pursue a politically safer agenda in order to survive in a hostile political environment. His emphasis on “depoliticization” of labor issues parallels what the regime has been trying to achieve with atomized incorporation: Rather than repressing workers’ discontent, the regime allows limited channels of contention but still depoliticizes the claims by restricting them to economic issues and maintaining those workers who would make claims in an atomized state. The efforts by labor NGOs to promote collective bargaining as a main strategy for labor protection led to successful cases, where workers were able to exert bottom-up pressure on improving labor conditions at the firm.85 The Guangdong Provincial People’s Congress, after a long delay due to opposition from business communities, finally passed the new regulations on collective contracts in 2014, which address the rights of employers and employees to conduct collective bargaining. Despite the NGOs’ efforts to walk a fine line, scholars and observers have noticed the Chinese regime’s increased tendency to crack down on labor activists in the Xi era. Government tactics to intimidate labor activists have ranged from direct repression, such as arrests, to indirect tactics such as closing NGO offices by pressuring the landlords.86 Labor activists have been constantly surveilled by security officials, had their phones tapped, and been subject to arbitrary tax investigations.87 Following the labor dispute at Lide shoe factory in Guangzhou in 2014 – during which the staff at the Panyu Migrants’ Center provided advice to workers on the strategies and tactics of collective bargaining – the local authorities started a massive crackdown on labor organizations and arrested five labor activists, including the Center’s founder. While the Center did not take part directly in workers’ collective action, their role in training and organizing workers behind the scenes might have been perceived by state authorities as crossing the line. As documented in Fu (2017), some of these NGOs have been coaching workers in the background of their seemingly atomized action, thus undermining the regime’s efforts to atomize workers. The increased crackdown on these NGOs in the Xi era signals that the regime might be aware of these efforts. The regime’s repression of labor organizations reflects the fact that incorporation of migrant labor inevitably creates a stronger need for labor atomization on the regime side, because it increases the number 85 86 87

Kuruvilla and Zhang, 2016; Froissart, 2017. Xu, 2013; Harney, 2015. Xu, 2013.

3.3 Atomized Incorporation in Post-reform China

75

of workers engaged in popular contention. Because of this greater potential for mobilization, the regime has been particularly hostile to “external actors” getting involved in labor disputes. For instance, foreign influence is one of the main reasons cited in the arrests of labor activists. An internal report by the Guangdong Politics and Law Committee of the CCP in 2009 reveals the special attention that the local government had been paying to foreign-funded NGOs. The report classifies those individuals or organizations that receive foreign support as one of the four types of professional citizen agents (zhiye gongmin dailiren).88 The report says the foreignfunded agents “use rights protection as their bait to intervene in people’s internal conflict and become the enemy and the forefront of destructing local governance and taking away popular support.” Local police monitor foreign-funded organizations and ensure that they do not engage in politically sensitive activities. Many prominent labor NGOs decided to stop receiving foreign funds but still did not survive the repression by the state. Following the arrest of the founder of the Panyu Migrant Center, the People’s Daily ran an editorial accusing him of receiving funding and training from “overseas organizations hostile to China.”89 Student activists who were involved in recent labor disputes – most notably the JASIC incident – faced immediate repression by the state authorities as well. In 2018, a small group of student activists from prestigious universities in China supported workers’ protests at JASIC Technology International – a privately owned company that exports welding machines – by forming the JASIC Workers Solidarity Group. This group helped workers to demand better labor conditions and the establishment of an independent trade union at the firm. It was reported that riot police raided the apartment used by the members of the group and detained fifty people.90 This was followed by another crackdown on student activists in five cities later that year, during which at least ten students disappeared.91 These students characterized themselves as Marxists and Maoists and appealed to the central government on the grounds that their actions were in line with the communist ideology. The crackdown shows clearly that the Xi leadership has been ramping up its efforts to atomize labor. In Chapter 8, I return to this topic to discuss the changing political landscape in the Xi era and how the theoretical framework 88 89 90 91

The term refers to individuals or organizations that legally represent citizens who directly harm social stability. Conner, 2016. Hernandez, 2018. Haas, 2018.

76

Atomized Incorporation

of atomized incorporation can be applied to the new developments in state–labor relations. 3.4 POLITICAL ECONOMY OF ATOMIZED INCORPORATION

Before moving onto empirical investigations of the political implications of apolitical labor protests, I ask why the regime adopted atomized incorporation in the way it did. Containing labor grievances at the firm level – and thereby directly exposing employers to growing demands from labor – is not the most obvious strategy if the regime’s priority is an attractive business environment for investors. The distributional consequences of atomized incorporation suggest that firms, not the state, bear most of the costs. This raises the question of why the Chinese state, particularly at the local level, might want to adopt pro-labor attitudes at the expense of business interests. Some suggest that conflicting interests at the local level are why local authorities have been unwilling to implement pro-labor policies. The argument implies that atomized incorporation is only a strategy on paper, but as I will explain, it is inaccurate to suggest that local authorities have ignored pro-labor policies altogether in order to help industry. While there is variation between regions and industries, local governments – particularly those that are assumed to rely heavily on export-oriented production – have made some efforts to implement pro-labor policies. Furthermore, not all local initiatives to atomize incorporation were passive responses to the central government’s directives. In the literature on democratization, some authors show that even business interests could start to favor labor protection when labor resistance is growing. In El Salvador and South Africa, for example, sustained labor insurgency was argued to have reduced economic elites’ demand for labor repressive institutions, as businesses preferred a stable supply of labor.92 According to the studies, frequent labor unrest could undermine firms’ economic incentives, and subsequently the interest of the political elites in maintaining a coercive labor institution. In these cases, the inclusionary political system came about because business groups wanted to eradicate the institution that coerced labor. The argument predicts that the Chinese regime’s pro-labor change could have been driven by requests from the business community, once the costs of labor unrest were perceived to exceed the benefits of labor coercion. 92

Wintrobe, 1998; Wood, 2000.

3.4 Political Economy of Atomized Incorporation

77

There is no evidence that in China, the government’s changing attitudes toward migrant labor were the result of pressure from economic actors. On the contrary, there was intense opposition from various business communities when the regime started to pursue pro-labor reforms.93 For instance, while drafting the Labor Contract Law in 2006, the central government launched a public consultation. Both the European Union Chamber of Commerce in China (EUCCC) and the American Chamber of Commerce in Shanghai (ACCS) submitted recommendations that opposed the labor law reform. ACCS argued that the draft of the labor contract law might “negatively impact the PRC’s competitiveness and appeal” and “reduce employment opportunities for PRC workers.”94 EUCCC likewise wrote that the labor contract law would increase labor costs and lead to a decline in private investment. A representative of an organization for multinationals in Shanghai threatened that firms could withdraw their investment. In the end, the final version of the Labor Contract Law was more pro-business than the original draft. In this section, I show that the state’s declining enthusiasm for laborintensive and export-oriented production has been one of the major drivers for the pro-labor policies. For the central government, the declining efficiency of the growth model that was based on labor-intensive exports prompted discussions about a more sustainable path of development. At the local level, the rising shortage of labor and land as a result of rapid industrial expansion has created new financial incentives for the local authorities to replace low value-added industries with those that are higher value-added. These changing attitudes toward labor-intensive sectors have had important implications for labor policies. 3.4.1 Declining Efficiency of the Input-Driven Economy In China, as in other input-based economies, the continued expansion of labor-intensive industrialization has led to the depletion of resources. In discussing historical examples such as the Soviet Union, economists have argued that an input-driven economy without technical innovation faces inherent limits to growth, because the supply of land and labor is finite.95 China, too, went from having abundant supplies of labor, land, and water to a situation of relative resource scarcity that made the traditional model of development more expensive to maintain and reduced its benefits. 93 94 95

Hui and Chan, 2014. Global Labor Strategies, 2007. Krugman, 1994; Easterly and Fischer, 1995.

78

Atomized Incorporation

Facing declining efficiency in labor-intensive sectors, the regime concluded that the old development model would not produce long-term growth. While official documents had discussed the unsustainability of the old development model as early as in the mid-1990s,96 the need to transform the development model became central to the regime’s economic goals when the Hu–Wen administration took office in 2003. Policy documents and government reports have referred to the differences between the “extensive” and “intensive” form of development, reflecting economists’ view that extensive growth by expansion of factor inputs, instead of productivity growth, cannot be sustained. The traditional labor-intensive industries that propelled the country’s growth during the 1980s and 1990s were declining in terms of competitiveness and contribution to total economic output.97 Another source of fragility was that the economy had been becoming increasingly reliant on exports since the reform era, rendering it vulnerable to external shocks. 3.4.2 Industrial Upgrading and Industry Relocation The Hu–Wen administration proposed “scientific” and “sustainable” development as a new model of growth. Two pillars of this new model were “indigenous innovation” (zizhu chuangxin) – which emphasizes domestic innovation to boost high-technology development – and “industrial upgrading” (chanye zhuanxing shengji), which aims to restructure the industrial composition, especially of foreign capital investment and export processing. While these two policy goals sometimes clashed with each other, both of them reflect the regime’s desire to replace traditional low value-added industries with technologically advanced ones.98 In 2005, the State Council issued “Regulations on Temporary Measures to Promote Industrial Upgrading” and urged local governments not to “blindly” accept investment. According to the document, those industries that are classified as a “restriction” type should only be selectively accepted, while the “elimination” type should be eradicated. 96

97 98

The regime’s ninth economic plan (1996–2000), for example, states that there should be a fundamental change in the economy’s growth model because “the country’s economy is already large and increases investment every year, but there exist many problems of low economic efficiency.” Wen et al., 2009. Chen and Ku, 2014.

3.4 Political Economy of Atomized Incorporation

79

Local governments in developed industrial areas have actively promoted industrial upgrading. Since the mid-2000s, the Guangdong provincial government has been pushing to replace low-end industries with high value-added ones. In its five-year plan for economic and social development between 2006 and 2010, the Guangdong province identified industrial upgrading as a major goal. In 2008, the province announced a more concrete plan, “Decisions on promoting industrial and labor relocation,” and introduced an industrial restructuring plan – commonly referred to as “double relocation” (shuangzhuanyi) – that aims to relocate both low value-added industries and low-skilled workers out of the PRD to less developed regions in the province. Wang Yang – who took office as the party secretary of Guangdong in 2008 – stated that it was time to “empty the cage and change the birds” (tenglonghuanniao). It is important to note that the provincial government’s initiative to relocate labor-intensive industries was not a passive response to the national government’s policy directives. The central–local relations in China have been analyzed as a principal–agent problem, because the goals of local government officials often diverge from those of the central government.99 Local leaders exercise discretion in enforcing central directives if the directives do not align with local leaders’ interests, demonstrating “minimum compliance.”100 The “double relocation” policy of the Guangdong Province, however, was a locally initiated experiment, although it eventually received support from the central government. The new initiative also started after the financial crisis when the economy was slowing down, signaling the provincial government’s commitment to industrial upgrading. Various sources suggest that the provincial government and other city governments in Guangdong saw the financial crisis as a good opportunity to move forward with industrial upgrading. Local governments had tried to relocate or eradicate low-efficiency firms in the past, but Wang Yang argued that the financial crisis “lowered the costs of relocation” and thus was the right time to act.101 The then-Party Secretary of Foshan mirrored this view, stating that before the crisis the government had to conduct “ideological work” (sixiang gongzuo) for those firms to leave; yet, as a result of the crisis, “the firms were already out of the market and with other regions welcoming their relocation, we can promote 99 100 101

Oi, 1999; O’Brien and Li, 1999; Huang, 1999; Whiting, 2004. Oi, 1999; O’Brien and Li, 1999. “There won’t be hollowing out in PRD (zhusanjiao buhui chuxian kongxinhua)” 2009.

Atomized Incorporation

80

industrial upgrading even more.”102 Even in cities like Dongguan where the financial crisis created both a “bankruptcy wave” (daobichao) and an “unemployment wave” shiyechao, the municipal government displayed a commitment to industrial upgrading. As reflected in Guangdong’s “double relocation” initiative, one of the key strategies of industrial upgrading has been geographical relocation of industries both at the national and regional levels. While emphasizing the need to accelerate high-technology development, the central government also has been against ejecting labor-intensive industries from China entirely. The unequal development between coastal and inland provinces during the last three decades means that underdeveloped areas of the central and western regions can still benefit from labor-intensive production. The “Go West” policy encourages industries in eastern provinces to relocate to inland provinces where labor and other production costs are cheaper. Firms in eastern provinces that produce low value-added products have been offered financial incentives to relocate the labor-intensive parts of production to inland regions, while keeping their headquarters in the east.103 The central government also produced a “List of Restricted Commodities in Process Trade” in 2007, laying out the types of products for which production is restricted in coastal regions but permitted in inland provinces. These are mostly labor-intensive products such as apparel and textiles. While the central government has aimed for a more balanced development between the inland and coastal provinces, provincial governments have largely focused on industrial reallocation within provinces. The “Double Relocation” strategy of Guangdong prioritizes relocating labor-intensive industries out of the PRD region to northern cities within Guangdong, rather than to other provinces. Similarly, Jiangsu Province in the YRD region also promoted industrial relocation from the more developed Southern Jiangsu to the less industrialized Northern Jiangsu. 3.4.3 Changing State–Industry Relations at the Local Level Three factors have driven provincial and city governments in industrialized coastal regions to pursue industrial upgrading and relocation of industries: reduced profitability of labor-intensive sectors due to the labor shortage and high turnover rates; increasing scarcity of resources such as land, and the social costs of mass migration to the cities. 102 103

Ma, 2009. Zhu and Pickles, 2013.

3.4 Political Economy of Atomized Incorporation

81

3.4.3.1 Declining Tax Contribution First, the rising labor shortage and higher turnover rates in low-skilled sectors have greatly reduced the sectors’ profit margins, which in turn decreased their contribution to the regional economy and, more importantly, to local tax revenues. The real wages of migrant workers, which had stagnated for more than a decade, have increased by ten to twenty percent annually since 2000.104 Therefore, labor-intensive industries that had been already operating under tight profit margins suffered further declines in profitability. According to the Federations of Hong Kong Industries, for example, the profit margins of HK-owned factories in the PRD, which mostly produce low-end products such as apparel and toys, fell from 18 percent in 2002 to 10 percent in 2007, mainly due to rising labor costs.105 In 2004, more than 240 enterprises in the Bao’an district of Shenzhen went bankrupt because they could not afford to pay the legal minimum wage to workers.106 Exports of low-skilled products, which enabled the explosive growth of the Shenzhen economy in the 1980s and 1990s, now only make up a small fraction of the local GDP and tax revenues.107 Comparison of profits and value-added taxes of different industries in the PRD further reveals the declining contribution of labor-intensive sectors to the economy (see Figures 3.2 and 3.3). While the profits and value-added taxes of electrical equipment and machinery, which is a high-skilled sector, have increased considerably, those of clothing and footwear – a low-skilled sector that was among the top three industries in terms of total output in the 1980s – have stagnated or grown only slightly.108 Local governments perceive industrial upgrading to be one of the most effective ways to increase local tax revenues. The Shenzhen Local Tax Bureau, for example, attributes the rapid growth of local tax revenues – which on average grew by 21.86 percent per year between 1995 and 2004 – to the development of high-tech industries and tertiary industries. The report demonstrates that the tax revenues of high-tech industries had increased more rapidly than those of the traditional manufacturing

104 105 106 107 108

Cai and Du, 2011; Li et al., 2012. FHKI, 2010. Ngok, 2011. Shenzhen Government Research Team, 2005; Chu and Zhang, 2010. For instance, food processing and the textile industry were among top three industries in terms of total output in Guangdong in 1985.

Atomized Incorporation

200

400

600

800

1000

1200

82

1998

2000

2002

2004

2006

2008

year Electric equipment and machinery

3.2 Profits of select industries in PRD

0

50

100

150

FIGURE

Clothing and footwear

1998

2000

2002

2004

2006

2008

year Electric machinery and equipment FIGURE

Clothing and footwear

3.3 Value-added taxes of select industries in PRD

3.4 Political Economy of Atomized Incorporation

83

sector, making up 55.13 percent of the manufacturing sectors’ tax contribution in 2004. Three tertiary industries – services, real estate, and finance – were the major contributors to business tax revenue in 2004.109 A report released by the Guangdong Local Tax Bureau confirms this: The push for industrial upgrading translated into more tax revenues, resulting in 17.5 percent growth of tax income in the PRD region. The Guangdong government revenues are now very dependent on the tertiary sector, which contributes 58.4 percent of the total tax revenues. More than half of the manufacturing sector’s taxes in 2015 came from advanced and high-technology manufacturing.110 In short, these illustrations demonstrate that local governments in developed coastal regions no longer perceive labor-intensive sectors to be a major contributor to their financial revenues and don’t face any incentives to cater to the sectors’ needs. 3.4.3.2 Insufficient Land Second, previously abundant resources such as land and electricity have become relatively scarce, prompting local governments to relocate unprofitable industries. If not for the resource scarcity, governments could wait for market forces to drive out inefficient industries, while seeking more investment from high-skilled sectors. But because resources are limited, the opportunity costs are perceived to be high if the low value-added industries continue to consume resources while making low returns. The land shortage has been identified as a major obstacle to economic development in PRD and YRD cities. During a 2012 planning meeting for industrial upgrading in Shenzhen, for example, the head of Shenzhen’s Economy, Trade, and Information Commission stated that Shenzhen needs to free up land in order to promote industrial upgrading. Dongguan had less than 400,000 mu – roughly 267 square kilometers – available in 2009, which could be used up in a couple of years.111 “Ten Main Projects for Industrial Upgrading in Shenzhen (2011–2015)” designates as a main goal “to get rid of low-end industries that are uncertified, unsafe, polluting, and energy consuming and to clear 5 million square meters by 2013 and 7.5 million square meters by 2015 for industrial development.”

109 110 111

Xiang et al., 2005. Wei et al., 2016. Che and Kong, 2008.

84

Atomized Incorporation

The Shenzhen government sums up their rationale in the following way: During the last 30 years, while Shenzhen transitioned itself from a small fishing village to a modern megacity, labor-intensive and extensive industries played an important role in the early period of economic growth. Advanced industries have been growing; yet, the labor-intensive industries that were attracted in the early period still make up a considerable part of the economy, leading to the problem of “high input, high energy use, high pollution, low benefits” and lack of innovative or competitive ability.112

3.4.3.3 Social Costs of Migration Finally, labor-intensive industries have generated social costs related to the large number of workers migrating into the cities. The city governments in coastal regions where migrant labor became the backbone of the local low-wage labor force acknowledged the benefit of migrant labor as a cheap but quiet and disposable labor force. Migrant labor’s contribution to the growing profits of export-oriented industries during those periods has translated into a surge in government revenues, since fiscal decentralization in the 1980s allowed local governments to retain taxes generated by business activities in their jurisdiction.113 More importantly, the hukou system legitimized the state’s denial of public goods and welfare services that support the labor force to migrant labor. For migrant labor, city governments did not have to pay for social security, children’s education, care for the elderly and unemployed, or housing.114 As Wolpe (1972) notes in his analysis of Apartheid, the institutional exclusion made home villages function as a “reserve” of surplus labor and enabled employers and the local governments to “pay the worker below the cost of his reproduction” (p. 11). For instance, the remarks by the Shenzhen Labour Bureau in 1992 capture the perceived benefits of accepting migrant workers at that time:115 The major component of the labor service market is the temporary workers. Their characteristic is that they are active and flexible in the market. They are assiduous, hardworking, easily manageable, and economically productive. What’s more they do not transfer their hukou (actually they are not permitted to do so). If 112 113 114

115

Shenzhen Government Research Team, 2013, p. 213. Oi, 1992; Oi, 1999; Montinola et al., 1995. While many of the subsidies and benefits that local governments distributed to urban residents largely disappeared in the 1990s, migrant workers are still excluded from important public good provisions such as public education, social assistance (minimum livelihood guarantee scheme, or dibao), and subsidized housing. Re-quoted from Pun (2005, pp. 46–47).

3.4 Political Economy of Atomized Incorporation

85

there is work, they come; if not, they go. This lessens the burden of enterprises, solves the problem of labor use, and at the same time does not result in urban overpopulation.

However, following the rapid industrialization and development of laborintensive factories, major cities in coastal areas have experienced a population explosion. As the previously tight restrictions on migration were relaxed in the post-reform era, the number of migrant workers in industrial cities surged despite the remaining institutional barriers. The residential population in Shenzhen, for example, increased from 3.8 million to more than 8.3 million between 1997 and 2010, mostly due to the influx of migrant workers.116 From local governments’ perspectives, migrant workers can be a significant source of financial and social burden due to the localities’ responsibility for public good provision. Local governments also tend to perceive migrant workers as low-quality residents who often commit crimes and disrupt social order.117 When migrant workers contributed to the local economy by generating profits in laborintensive sectors, local government officials helped enterprises to recruit workers from the countryside. Yet, now that labor-intensive sectors’ contribution has shrunk, local governments have increasingly been unwilling to pay the costs. In 2007, labor-intensive industries in the PRD – such as home appliances, leather, and clothing manufacturing – utilized a quarter of the province’s labor force, but only created 8.5 percent of the total value-added output.118 Industrial relocation is an attractive strategy for the city governments in developed coastal regions, therefore, partly because it could reduce the flow of migrant workers into the city. This point contrasts with the existing argument that the local governments are using pro-labor policies to attract migrant workers and combat labor shortage. The “double relocation” strategy in Guangdong clearly shows the province’s willingness to use industrial upgrading as a tool to reduce population congestion, especially by relocating low-skilled migrants. The 2008 policy document explicitly states that double relocation will “create more room for development, reduce the population in the PRD, increase the quality of labor, and relieve the burden of social management.” Likewise, the Zhejiang Provincial government claims that one major benefit of their policy efforts to “replace people with machines” (jiqihuanren) has been the reduced 116 117 118

Shenzhen Government Research Team, 2013, p. 212. Zhang, 2001. “Creating a Modern Industrial System: PRD Prepares to Move Labor-intensive Sectors” 2007.

86

Atomized Incorporation

number of migrant workers. A report by the Zhejiang Provincial Bureau of Statistics states that the policy improved the quality of job seekers in the province, relieved the labor shortage in many enterprises, and improved labor productivity.119 The broader strategy for dealing with labor shortage, therefore, has not necessarily been to raise the attractiveness of the cities for migrant workers but to relocate the very industries that lack low-skilled workers. At the same time, the coastal cities have made efforts to attract high-skilled workers to the high-tech industries.120 3.4.3.4 A City Official’s Perspective These reasons together illustrate that even local governments with shortterm interests in economic development have reasons to discourage investment in labor-intensive sectors and to encourage them to relocate. In an interview, a vice mayor of a coastal city who had been overseeing business activities in the region, expressed the city’s enthusiasm for industrial upgrading and relocation, mainly for financial reasons.121 When asked how the government views industrial upgrading, the vice mayor emphasized that it was a pressing need, singling out tax revenues as the most important factor. According to him, the city has too many “backward” (louhou) industries, which he defines as those “that cannot pay taxes, cannot feed their workers, pollute the environment, and have no potential.” These labor-intensive industries “make little or no tax contribution but waste scarce resources.” More specifically, he states that the way in which the city attracts good investors has changed and that the timely elimination of low-end industries is paramount to make room for more profitable industries to settle down: From the perspective of the local government, tax payment surely is the most important reason why the city wants to attract high-technology and profitable firms. There are other reasons too. High-technology firms can also promote a good image of the city and bring in other good firms, which rent land, buy raw materials, and facilitate the local economy. It is far better to welcome one good firm than to retain 10,000 firms in the traditional industries such as leather, textile, or hardware (wujin). But, unlike in the 80s when we provided tax exemptions or free land to attract investment, we cannot directly provide what high-technology firms want. As a 119 120

121

Zhejiang Bureau of Statistics, 2013. For example, many cities introduced a point-based hukou conversion program under which only the individuals with highly competitive labor market attributes can accumulate enough points to receive the local hukou. Interview 07-2017.

3.4 Political Economy of Atomized Incorporation

87

local government that needs to provide various public goods, we can no longer offer large tax cuts. High-technology firms also want different things. They want a good market. They are interested in what kinds of firms are located in the area and what kinds of markets they can enter. The city, therefore, can only provide indirect services to make the market attractive. The most important task for the city is to make resources available. Especially, land management is critical. Only by strictly controlling land usage, can the city provide enough land to high value-added firms. The problem with those backward firms is that they waste land. Therefore, the city must eliminate them. The earlier, the better.

The vice mayor’s remarks clearly show that the city is not interested in retaining industries that are labor-intensive and technologically backward. The financial incentives combined with the city’s efforts to promote long-term growth have left the city indifferent toward labor-intensive firms that cannot survive in the new policy environment. Moreover, the city has actively sought to eliminate unprofitable firms to increase economic efficiency, especially the efficiency of land use. This complements and reinforces the findings in the literature that cities with strong financial capacity are more likely to implement policies that are unfriendly to some businesses.122 3.4.4 Investment Discouragement and Labor Policy Local governments’ approaches to labor have been tightly linked to their industrial policies. The incentives for industrial restructuring gave local governments in industrialized cities some motivation to enforce pro-labor policies. As much as the state’s coercive labor policies helped localities to attract investment in labor-intensive sectors, its pro-labor policies have managed to discourage investment in unprofitable sectors. Interestingly, the same policy tools that had been used to encourage investment were also wielded to discourage it. The pro-investment policies that in the early stage of industrialization ensured cheap labor, easy access to resources, and low-cost land were re-shaped to create disincentives for investment. In the case of labor policy, having turned a blind eye to or even encouraged coercive labor practices in the past, the local authorities in industrial cities now utilize pro-labor policies to increase the production and operating costs of labor-intensive industries to push unprofitable businesses out of the market sooner. For example, when asked how the city achieves the goal of industry relocation and elimination, the vice mayor quoted 122

Kamp et al., 2017.

88

Atomized Incorporation

above cited two major policy tools: labor law and environmental law. He stated: It is easy [to eliminate backward firms]. You just to need to enforce labor law or environmental law. For example, you enforce the five insurances and one fund (wuxianyijin) law. When the labor costs go up, you naturally eliminate those unprofitable firms that cannot pay for the labor costs. Or, one day, the environmental bureau can shut down a factory for violating environmental laws and by the time the factory reopens, the workers are gone, and the factory needs to move out or close.

The enforcement of pro-labor policies is an effective policy tool for city governments in regions where there are labor-intensive sectors for which labor costs are a relatively large part of total production costs. Enforcing or adopting pro-labor policies can disproportionately increase costs for labor-intensive firms, while having a marginal effect on high-technology firms. Some local governments have been very transparent about this strategy. Announcing the city’s decision to increase the legal minimum wage in 2005, for example, the Shenzhen Bureau of Development and Reform stated that the decision was the first step toward forcing out the labor-intensive industries that also use large amounts of energy and space.123 The city estimated that 5,800 factories, 4.14 percent of the total number of enterprises in the city, would go bankrupt as a result of the minimum wage increase.124 Responding to firms’ concerns about the negative impact on firm profits, the head of the Labor and Social Security Bureau stated that “if a firm cannot even pay the minimum wage to workers, the firm will inevitably be eliminated. Shenzhen needs [a] high-skilled workforce and we are in the process of decreasing the ratio of low-skilled workers.”125 After the introduction of the Labor Contract Law, the Longgang district in Shenzhen reported that 46 enterprises relocated to another region and 156 enterprises closed down.126 The central government has promoted the “lawful” elimination of unprofitable and low-end firms by enforcing tightened regulations that increase production costs. Neither the central government nor local governments want to appear to be directly intervening in the market and forcefully removing certain firms, and this “regulatory discouragement” provides important cover. The localities still want to attract investors 123 124 125 126

“Shenzhen wants to drive away labor-intensive firms (Shenzhen yao bizou laodongmijiqiye)” 2005. Ngok, 2011, p. 295. Liu, 2005. Longgang District Government Labor Bureau, 2008.

3.4 Political Economy of Atomized Incorporation

89

in high-skilled sectors, so it is important that the low-end firms appear to be choosing to leave the market rather than being forced out. After the announcement of the double relocation policy in Guangdong, for instance, the vice mayor of the Dongguan municipal government tried to reassure stakeholders that the role of government in industrial relocation was only supplementary, arguing that the government does not seek to remove enterprises forcefully or replace market forces. Instead, he argued, “the government is supporting the role of the market so that, during the course of industrial upgrading, the market can play its role better.”127 Firm managers in industrialized cities also perceive pro-labor policies to be a state strategy to deter investment in labor-intensive sectors. In Shenzhen, the city government rigorously implemented the legal minimum wage regulation between 2004 and 2005. The primary motive, according to a manager of an electronics firm, could have been to clean out the most advanced district, the Futian district, to create more space for office buildings.128 Implementation of the series of new labor regulations, a manager of a fan factory in Dongguan stated, “is a strategy of the city government to put pressure on factories like ours. The government is telling us either to move out or upgrade.”129 Cities such as Zhongshan were more explicit about the criteria that the city government would use to eliminate firms. When the Zhuhai government announced its plan to clean out industries for industrial upgrading in 2013, the policy document listed the following conditions:130 1. Undesirable enterprises (those that meet any of the following criteria): (1) uncertified or illegal enterprises; (2) those that do not meet safety regulations; (3) those that violate environmental regulations; (4) those that do not meet labor-related laws and regulations on social insurance, and create large-scale labor conflicts; (5) those that are low-technology, simple processing, and ill-managed. 2. Low-end enterprises (those that meet any of the following criteria): (1) those that do not fit into Zhuhai’s new direction for development; (2) those that have labor productivity that is lower than 30 percent of the industry-wide average or than 50 percent of the entire city’s industrial average; 127 128 129 130

Wang, 2008. Interview 06-2013. Interview 02-2014. Gao et al., 2017.

Atomized Incorporation

90

(3) enterprises with a gross output per land use that is lower than 30 percent of the area’s industry average or than 50 percent of the city-wide average; (4) enterprises whose tax payment per land use is lower than 30 percent of the industry average or than 50 percent of the city-wide average; (5) enterprises with energy consumption per industrial output that is higher than the industry average.

Local officials are urged to eliminate firms that do not meet safety, environmental, and labor laws and regulations. The policy document states that by enforcing such laws strictly, the governments can legally push out undesirable enterprises. The criteria reflect the city government’s message that, in order to operate in the region, the firm should at least be profitable and have sufficient resources to cover the legally required safety, environmental, and labor costs. The document also shows that the local government uses economic efficiency as a standard for elimination. The “low-end” category suggests that once a firm passes the first test, meeting relevant laws and regulations, there is a second test that measures their efficiency per unit of land in terms of gross output, tax obligations and energy consumption, as well as labor productivity. The less efficient firms are given a probational period for internal innovation. The criteria reveal how the city government is strategically calculating the yield of land in terms of economic growth and tax contribution. The document states that when these enterprises do not properly upgrade their industrial structure in a timely manner, they should also be included in the elimination category. Some firms successfully upgrade and eventually generate more profits. A Taiwanese jewelry factory in Guangzhou, for example, used to be one of the low-skilled industries that the district government did not want to retain. The district government put the firm on probation and gave it a period to upgrade, because the enterprise still made some profits and met all the legal requirements. The local government suggested that the enterprise reform its industrial structure to be more service-oriented. The factory trained their workers to become jewelry-making entertainers so customers could come to the factory and design their own jewelry while watching the workers perform. The factory owner argued that the transition was a “win-win” for both the factory and the government: The factory could survive the intense competition of the low-end jewelry industry, while the local government could receive more taxes from the factory.131 131

Interview 07-2013.

3.4 Political Economy of Atomized Incorporation

91

The new policy environment explains city governments’ greater commitment to enforcing labor laws and regulations as well as more widespread concern among factory managers about not being able to meet the enforced standards. The owner of a small clothing factory in Guangzhou, for instance, explained that local government officials had become much stricter about the social insurance and overtime payment. In the past, she could just pay one yuan extra per hour for overtime and pass the district’s labor inspection, which she said had happened on a regular basis. Now, the labor inspectors require her factory to pay the legally mandated level to all of her workers. She expressed concerns about the factory’s survival, saying that “the local government obviously does not want my factory in the district any longer.” When there was a labor strike in 2010, she did not even want the local government to know of the incident because it would reflect badly on the factory. The labor bureau, she said, would definitely not intervene unless there was violence involved. More importantly, the local government “did not have any economic incentive” to help her factory. The owner was worried that the labor bureau and other bureaus might visit her factory later to “find faults” because her factory experienced a labor protest.132 These observations about stricter enforcement of labor standards should be interpreted with caution, however. The shifting financial incentives of local governments mean that many local governments have applied labor laws instrumentally to assist local economic growth. Many city governments announce a target for the number of low-end enterprises to be removed from the region to increase the area’s economic efficiency within a set period of time. As a result, the strictness of the implementation of labor laws and regulations waxes and wanes, and fails to protect labor rights in a more fundamental way. This might explain, for example, why some local inspection bureaus are directed to deal with high-tech firms differently from low value-added firms. A district government in Shenzhen suggests that the local labor inspection should rely on color ratings to differentiate firms of varying quality. For those firms that are low value-added and labor intensive (red color), the labor inspection bureau should crack down harder to promote industrial upgrading. The inspection bureau is urged to strictly monitor, manage, and punish the firms. For those “relatively big, high-technology, profitable, and new industry” firms, the labor inspection bureau should actively help to maintain peaceful labor relations to support rapid growth. 132

Interview 05-2012.

92

Atomized Incorporation

Even when the firm has a history of violating labor laws, according to the report, the district labor bureau should help them to deal with the matter internally. My interviews with firms in different regions and industries provide preliminary evidence that there is wide variation in the local governments’ efforts to implement pro-labor policies, due to the diverging incentives. A cosmetics factory in Guangdong – which pays the most tax of all the factories in their district – revealed that the local government had been less strict with the firm. “One of the perks of being the biggest tax contributor,” the manager said, “is that the local government gives a lot of leeway in terms of regulations, for instance, in regards to environment, labor, and safety. Of course, we try to comply with the regulations, but any factory that operates here could be [found to be] violating one or two if the government decides that it is a problem.”133 A manager of a wig factory in Shenzhen stated that, even within the same industry, the extent to which firms abide by labor laws and regulations differs greatly depending on the firm’s location:134 Here [in Shenzhen] you need to abide by the labor laws. Especially as a foreigninvested firm, you carry a lot of risk by not abiding by the local law. We adopt a five-day work week and pay workers double the usual wage for working on weekends. In Henan Province, however, wig factories only take two days off per month and do not even provide overtime compensation. Such practices are acceptable because the local governments in Henan still rely on labor-intensive industries such as wig factories for local economic development. If they enforce those laws strictly, most of the firms in their localities will close down.

While further research is needed to determine whether the level of labor law implementation differs systematically across regions,135 some firm managers in inland provinces speculate that they might have wider latitude for illegal labor practices. A firm manager and the head of the HR department at a power equipment company, which has twenty-four factories in Henan, suggested that the firm did not feel that labor law implementation had become particularly strict. Even as a large-scale company, the company encountered no pushback for paying the five insurances only to managerial-level employees and excluding production

133 134 135

Interview 07-2017. Interview 07-2013. Gallagher et al. (2015) examines regional variations in how workers perceive the implementation of the Labor Contract, but there is no study that analyzes how the actual level of implementation differs.

3.5 Conclusion

93

workers. The company, according to the firm manager, “has not particularly received pressure for industrial upgrading,” even if most of their factories would be classified as labor-intensive.136 The variations might explain why migrant workers in coastal cities and more advanced cities are more likely to have a signed labor contract and social insurance.137 3.5 CONCLUSION

The political and economic environment in China started to change in the early 2000s. In this chapter, I analyzed how the Chinese regime responded with a new strategy to accommodate and control labor. “Atomized incorporation” describes an approach that accommodates labor demands but restricts the scope of claim-making to individual workplaces. Atomization is preserved, but tactics switch from coercive control of labor to limited incorporation. The strategy differs from state corporatism in Latin America, since the regime does not permit any form of organized labor to mediate state–labor relations. Instead, the regime has adopted formal and informal channels of grievance resolution in which atomized workers can express discontent at the firm level, reflecting only the voices of individuals or small groups of workers. In the context of these changes, I discussed theoretical expectations about the conditions under which atomized incorporation can depoliticize labor discontent and ensure the regime’s long-term political stability. In Chapters 4–7, I present in-depth empirical analyses for these theoretical expectations. The distributional consequences of atomized incorporation suggest that we need to understand the changing state-capital relations at the central and local levels, to put into context why the regime has introduced and enforced pro-labor policies that could harm business profits. The empirical evidence suggests that atomized incorporation was possible because of the state’s declining interest in retaining labor-intensive industries, and its pursuit of industries with higher returns to land, labor, and capital. Particularly in export-oriented regions, local authorities have pursued a new industrial strategy of replacing traditional labor-intensive industries with high-technology ones. These efforts toward industrial upgrading have influenced the ways in which local authorities enforce pro-labor policies. In some areas, pro-labor laws have been used to discourage labor-intensive, low-margin investment, to make room for 136 137

Interview 06-2017. Gallagher et al., 2015; Gao et al., 2017.

94

Atomized Incorporation

companies that will pay more in taxes. Implementation of pro-labor policies in these regions, therefore, is not necessarily a reflection of a different model of local governance but rather should be understood as a continuation of local authorities’ focus on accumulation. At the same time, atomized incorporation provides flexibility to local governments, who can adjust their responsiveness to workers’ atomized claim-making based on the local economic conditions. The material in this chapter reinforces the conclusion of Chapter 2 that rapid economic growth that depends on labor coercion is not sustainable. In addition to the long-term political changes – increasing labor discontent and structural empowerment of workers – the pursuit of laborintensive growth reduces economic efficiency, making it a less attractive strategy for the regime. When economic development is based on inputdriven industry, the success of that approach ultimately leads to shortages of the inputs it relies on, pushing prices of inputs up and profitability down. In the long run, labor coercion evolves from being a source of economic and political stability, to becoming a threat to both. History shows that many authoritarian regimes faced strong pressure for democratization at this critical juncture. Without political or organizational linkages with labor, the regime becomes increasingly reliant on atomization – that requires investment in monitoring, preemption, and selective repression – to prevent a labor movement from emerging. This shows that the political environment for migrant workers’ collective action has changed substantially over the last two decades. The changes have generated both positive and negative conditions for collective action. Workers’ bargaining power in the labor market has increased, the regime has adopted some pro-labor positions, and workers have become more educated about their rights and interests. However, to maintain atomized incorporation, the regime cracks down on any attempts to organize labor independently and suppresses workers’ collective capacity to raise grievances outside the boundaries of each factory. Increased labor mobility also undermines the formation and strengthening of social connections within workplaces to function as a glue for collective action. As a result, despite their growing market power, workers still face acute collective action problems that stand in the way of organizing more substantial protests inside and outside the factory gate.

4 Politicization of Labor Discontent and Blame Attribution

Despite the increasing discontent among migrant workers, scholars have disagreed as to whether their grievances would ever pose a threat to oneparty rule in China. On the one hand, earlier surveys show that migrant workers seem generally satisfied with the economic reforms and appear to believe that success is a matter of individual effort. Their focus on jobspecific economic demands at protests supports a conclusion that there is little sign of politicized labor activism. On the other hand, there has been an emergence of critical consciousness, particularly among young migrant workers. Scholars argue that many of these younger workers display class consciousness during collective action and that their demands for material benefits might reflect a more diffused sense of dissatisfaction about their marginalized status. There is a long line of inquiry about whether economic grievances determine political attitudes and behavior. There is ample evidence, for instance, that electoral outcomes in democracies depend on voters’ evaluation of the nation’s economic performance.1 In authoritarian states, political economy theories see democratization as an outcome of a redistributive battle and posit that the poor would oppose the status quo when economic inequality is high.2 In China, studies have emphasized the importance of “performance legitimacy” in explaining regime resilience, with the assumption that the remarkable growth of the Chinese economy has been a central pillar of popular support. Yet, often objective economic conditions do not map perfectly onto the patterns of political support. There has been substantial debate in 1 2

Monroe, 1979; Lau and Sears, 1981. Rueschemeyer et al., 1992; Boix, 2003; Acemoglu and Robinson, 2009.

95

96

Politicization of Labor Discontent and Blame Attribution

democratic contexts about the conditions under which economic circumstances affect individuals’ voting behavior and policy preferences.3 Chinese society has become increasingly unequal in the post-reform era, but it is not clear whether the inequality has been politicized in a way that makes the central government a target. Not only do citizens disagree with one another about the extent of inequality, many also find inequality to be an acceptable byproduct of economic growth.4 The effect of economic conditions on political attitudes could be even more nuanced in a country like China, given that political leaders can use a range of tactics to depoliticize citizens’ everyday struggles. Politicians are known to devote considerable resources to distancing themselves from policy failures.5 In authoritarian regimes, leaders have the advantage of controlling the flow of information and framing the extent and nature of difficulties that citizens experience. Depoliticization in this regard is ultimately a strategy for blame avoidance. Leaders strive to maintain political support by fostering a perception that citizens’ struggles do not originate in politics. In this chapter, I examine atomized incorporation as a strategy of blame avoidance. I study the patterns of politicized grievances among migrant workers and ask whether atomized incorporation has successfully shifted blame away from the central government. As discussed in Chapter 2, migrant workers have begun to express their grievances about low wages, poor working conditions, limited job choices, and discrimination in the cities. Although the central government has a clear responsibility for these hardships, it is not clear whether migrant workers hold the central government accountable. The logic of atomized incorporation predicts that the central government can avoid blame by framing migrant labor issues as an economic conflict between employers and employees. In this chapter, I ask under which conditions migrant workers politicize their grievances and hold the central government responsible vis-à-vis other actors. Using my survey data and interviews, I investigate the determinants of blame attribution and present two major findings: (1) Atomized incorporation has successfully depoliticized labor protests because in the current environment, those who protest are less likely to blame the central government. (2) The strategy, however, has not successfully depoliticized labor discontent more broadly. A substantial number of migrant workers 3 4 5

Monroe, 1979; Lau and Sears, 1981; Mansfield and Mutz, 2009; Rho and Tomz, 2017; Inglehart and Norris, 2017. Whyte, 2010. Weaver, 1986; Ellis, 1994; Hood, 2010.

4.1 Debate about Migrant Labor Discontent

97

now perceive a generalized pattern of migrant labor struggles and have started to form grievances about the broader system. The high mobility of migrants becomes an alternative source of information and generates social grievances about limited upward mobility, income inequality, and social fairness. It is these social grievances, rather than particular grievances about job-specific wages, that change the direction of blame attribution and politicize labor discontent. 4.1 DEBATE ABOUT MIGRANT LABOR DISCONTENT

Migrant workers currently experience hardships that are largely due to the regime’s institutionalized discrimination against rural and migrant workers. The hukou system reduces migrant workers’ socioeconomic status by limiting their income, restricting access to good jobs, creating negative prejudice, and discouraging and impeding integration into urban areas. One result of the regime’s uneven investment in public goods is the large urban–rural gap in educational attainment, which is one of the most important determinants of upward mobility in post-reform China.6 While institutionalized discrimination has weakened over time, its legacies still limit migrant workers’ life prospects and social inclusion.7 Although the origins of this marginalization of the rural and migrant population are structural, it has been argued that migrant workers support the political status quo. Studies find that migrants express greater optimism and more acceptance of social inequality than urban residents.8 The China Justice Survey from 2004 found that only 0.49 percent of migrant respondents expected that their life would become much worse within five years.9 Migrant workers’ positive assessment of their upward mobility might be explained by their standard of comparison. Those who remember dire economic conditions in rural areas might appreciate the outcomes of economic reform, even if the benefits have been unevenly distributed. Wright (2010) suggests that workers in the private sector would not have “a pressing material reason to push for regime change,” as long as they believe that the regime is going to deliver a basic safety net and upward mobility (p. 163). The literature on the “new generation” of migrants, however, challenges this characterization of migrant labor, noting that the younger cohort is substantively different. As explained in Chapter 2, the younger 6 7 8 9

Qian and Smyth, 2008; Yeung, 2014. Zhan, 2011. Feng, 2010; Whyte, 2010. Feng, 2010, p. 229.

98

Politicization of Labor Discontent and Blame Attribution

cohort has a higher standard of living than the previous generation and expresses higher social aspirations.10 The generation of migrants born in the 1980s and 1990s has weaker connections to the village and more ambitious career goals than the preceding generation, but it is still difficult for these migrants to exit the dagong life, and they resent that they are treated as inferior by urban society. Despite the improvement in wages and working conditions since the mid-2000s, many in the younger cohort have not experienced substantial upward mobility that will lift them out of the dagong life. Even as their wages grow, they are still perceived as low-quality workers who take the jobs that are shunned by urban residents; the difficulty of integration into urban society exacerbates their frustrations. Given these divergent observations, scholars have disagreed about the extent to which migrant labor discontent has been, or could be, politicized. Some find that migrant workers generally are content with their migrant life and have weak political consciousness. Others, however, argue that it is necessary to reexamine this thesis because the younger cohort is more prone to resist the status quo.11 Microlevel investigations of collective action suggest that migrant workers have developed substantial class consciousness and collective identity,12 which earlier studies find to be a precursor to anti-regime attitudes.13 4.2 POLITICIZATION, ATOMIZATION, AND BLAME ATTRIBUTION

4.2.1 Internal Attribution of Blame among Marginalized Communities The fact that many migrant workers do not politicize their personal struggles might not be surprising, given our understanding of marginalized populations. Various strands of literature suggest that marginalized citizens often do not blame the system that enables the marginalization. They do not identify that there is an authority that establishes and controls the structure of marginalization. Even when they have grievances about their everyday struggles, their perceptions would not necessarily converge because every individual experiences marginalization in their own specific context. For instance, Piven and Cloward (1979) argue: 10 11 12 13

Pun and Huilin, 2010; Chan and Pun, 2010; PRC Bureau of Statistics, 2011; Béja, 2011; Liu et al., 2012; Pun and Chan, 2013; Fan and Chen, 2013; Cheng, 2014. Nang and Pun, 2009; Pun and Huilin, 2010. Chan, 2009; Chan and Siu, 2012. Koo, 2001.

4.2 Politicization, Atomization, and Blame Attribution

99

[P]eople experience deprivation and oppression within a concrete setting, not as the end product of large and abstract processes, and it is the concrete experience that molds their discontent into specific grievances against specific targets. [...] In other words, it is the daily experience of people that shapes their grievances, establishes the measure of their demands, and points out the targets of their anger.14

The authors’ insights predict that migrant workers do not necessarily perceive their specific grievances about wages, living conditions, or career options as having the same culprit. In Pedagogy of the Oppressed, Freire (1968) suggests that people’s “perception of themselves as oppressed is impaired by their submersion in the reality of oppression” (p. 45). The study calls for a pedagogical process that enables people to “first critically recognize its causes, so that through transforming action they can create a new situation” (p. 29). In the author’s words, it is difficult for marginalized citizens to develop “critical consciousness” because of the way the broader system affects their cognition. Similarly, in What is to be done, Lenin argues that political knowledge and education should be brought to the workers “only from without, that is, only from outside the economic struggle, from outside the sphere of relations between workers and employers.” Without the external input, he postulates, workers would not become political agents simply by battling economic challenges. Applying these insights to contemporary trade unions, Ahlquist and Levi (2013) suggest if they are to engage in broader political and social movements, they need to overcome what Lenin calls “economism,” that is, ordinary workers’ “focus on the narrow economic interests bound up in the job” (p. 3). Unions first need to enlarge “the community of fate.” The role of modern unions might not be confined to providing the type of political education that Lenin argued for, but by disseminating information and providing opportunities for political socialization, union leadership could shape workers’ beliefs about the interdependence of their interests in the larger society. From a worker perspective, causal beliefs about the origins of their plight matter. For workers to pursue a collective solution, they need to recognize the structural origins of individual issues and be willing to hold the relevant agents accountable. If workers do not recognize the role of the broader system, however, they might not be inclined to politicize their everyday struggles. 14

Piven and Cloward, 1979, p. 20.

100

Politicization of Labor Discontent and Blame Attribution

Marginalized citizens often attribute blame to themselves. A migrant worker from Sichuan Province working in Shanghai said that she was stuck in the dagong life because of her limited education. “I am most regretful of my decision to drop out of high school and come out to work so early,” she said.15 She did not think education would help her at all because the kind of job available to her seemed to be the same regardless of whether she completed high school or not. Yet, now that she had worked for more than a year in the factory, she realized that she would never get a good job because she was low-educated. From her perspective, it was she who should carry the blame for the low wages and unsatisfactory jobs, not the firm or the state. This internal attribution of blame has been argued to be a critical impediment to political activism. When citizens blame external actors, they are more likely to act on their grievances and engage in behavior aimed at political change. For instance, when citizens attribute a bad economic situation to policy failures, they are more likely to vote against the incumbent government.16 It also matters where citizens direct their blame. If they blame a specific politician or institution, they are also more likely to participate in collective action.17 In authoritarian regimes, even citizens with low capacity to mobilize participate in anti-regime protests when they have strong economic grievances that they attribute to the regime.18 Therefore, both the extent and direction of blame attribution are factors in channeling economic grievances into political activism. 4.2.1.1 Blame Avoidance Strategy: Framing and Agency This could explain political leaders’ constant efforts to influence citizens’ cognition in ways that help them to avoid blame. According to Weaver (1986), the desire for blame avoidance is so strong among politicians that they will adopt blame avoidance strategies even if it means that they miss out on receiving credit for policy successes. One major strategy is to change and control the “framing” of the issues.19 Politicians, for instance, could argue that an economic downturn is the consequence of a global crisis, rather than the government’s policy failure. Framing can be effective because citizens normally lack access to full information to make a clear judgement about causality. In some contexts, politicians 15 16 17 18 19

Interview 06-2017. Lau and Sears, 1981; Hurwitz and Peffley, 1987; Tilley and Hobolt, 2015; Arceneaux, 2016. Javeline, 2003. Thomson, 2018. Brändström and Kuipers, 2003.

4.2 Politicization, Atomization, and Blame Attribution

101

intentionally withhold some information from citizens to avoid blame.20 Insufficient information makes it difficult for citizens to attribute blame based on the objective circumstances.21 An authoritarian regime has an advantage over its democratic counterparts in controlling the flow of information. In China, the central government has relied heavily on a strategy of blame avoidance to maintain its political legitimacy amid rising social tension. Studies of media censorship have shown that the central government carefully manages public opinion by monitoring and shaping how social and political issues are framed.22 During the SOE restructuring, for instance, Cai (2002) shows that the state media attributed laid-off workers’ unemployment to their “old” mentality.23 During the SARS crisis, the central government did not fully disclose information for fear of being blamed for its failure to rapidly control the spread of the disease.24 Another strategy to avoid blame is to find an alternative agent to absorb the blame. Ellis (1994) calls these the “lightning rods.” He finds that US presidents have delegated policy initiatives to policy advisors or experts so that the public would perceive these individuals as the immediate agents and attribute policy mistakes to them. Hood (2010) develops this concept and discusses the “agency” strategy as a key tool of blame avoidance.25 By letting individual advisors or agencies to take charge, political leaders are able to find scapegoats for policy failures. By pointing to other actors as being responsible for social issues, they are able to diffuse their responsibility.26 Even though the leadership has the ultimate authority to make policy decisions, it can pass the blame to the agents who appear to be closer to the design and implementation of the policies. These two strategies – framing and agency – might not be mutually exclusive, since the leadership could frame the agency without actual delegation. The agency strategy has also been widely used in China. In the postreform era, the central government granted more autonomy over policy implementation to local governments. Local government officials are

20 21 22 23 24 25 26

Page and Brody, 1972. Brown, 2015. Lü, 2014; Stockmann and Gallagher, 2011; King et al., 2013; Lorentzen, 2014; Roberts, 2018. Cai, 2002, pp. 88–89. Baekkeskov and Rubin, 2017. In addition to lightning rods, Hood (2010) also introduces other agency strategies such as reorganization of political institutions and blaming market conditions. McGraw, 1990.

102

Politicization of Labor Discontent and Blame Attribution

perceived to have direct authority over many policy outcomes. The delegation of responsibility has helped the central government to maintain political legitimacy, since most citizen grievances are directed at local governments.27 Despite the greater prominence of social problems, the central government is able to maintain relatively high levels of public support, while local governments are at the front line of absorbing citizens’ grievances. Cai (2008) shows that the central government’s delegation of responsibility for dealing with popular protests has been an important mechanism for helping the central government to avoid blame. Lü (2014) demonstrates that as a result of biased media reporting, Chinese citizens attribute policy benefits to the central government and policy failures to the local government. 4.2.1.2 Atomized Incorporation as a Blame Avoidance Strategy The Chinese regime’s framing of migrant labor issues in the 1980s and 1990s facilitated migrant workers’ internal attribution of blame. In the media and official discourses, migrants were depicted as “low quality,” “unruly,” “low-class,” and “backward.”28 The “floating population” was often blamed for all sorts of social problems in urban areas such as crime, public health concerns, and congestion. In the Beijing Evening Daily, one official said, “The contrast between city and countryside is too difficult for peasants in the cities to adjust to; and this huge contrast of material wealth produces an imbalance in their minds... leading to the path of reaping without sowing.”29 The comment attributes migrant workers’ difficulties with integrating into the urban areas and limited opportunities to their “inherent flaws.” The official rhetoric was so powerful that not only urban residents but also migrants themselves seem to have internalized it. Feng (2010) shows, for instance, that a substantial percentage of migrant workers believe that urban biases are fair, although they were less likely to believe this than urban or rural residents were.30 The same survey shows that while the majority of migrants (78.17 percent) perceive that the national income inequality is too large, most of them (74.43 percent) also believe

27 28 29 30

Cai, 2008; Lü, 2014; Tong and Lei, 2013. Zhang, 2001, p. 142. Beijing Evening Daily, March 24, 1996. Requoted from Zhang (ibid., p. 143). Only 15.15 percent of the migrant respondents in the 2004 China Social Justice Survey displayed a strong objection to the claim that those with an urban hukou should receive more opportunity (p. 231). The percentage was higher than among urban residents (12.44) or rural residents (9.99), but it is still low given that migrants are the ones that actually experience the hukou-based discrimination in the cities.

4.2 Politicization, Atomization, and Blame Attribution

103

in hard work, reflecting the argument in Whyte (2010) that Chinese citizens’ perception of income inequality has not necessarily led to political grievances, because citizens believe in upward mobility. In another survey from 2005, Wong (2011) finds that a substantial percentage of migrants themselves agree with the common prejudice against them: About 36.6 percent agreed with the claim that migrants steal jobs from locals and 26.0 percent agreed that migrants have lower “quality” than locals. Local governments and firms have relied on these common stereotypes to discriminate between migrant workers and locals. There is a caveat that surveys conducted in the early and mid-2000s might not reflect the changing demographics of migrant workers. It is plausible that, as the existing studies postulate, the younger cohort of migrants display a higher level of dissatisfaction and are less willing to accept the rhetoric that hard work pays off. More recent literature suggests that migrant workers’ resistance in the workplace originates in their negative emotional reactions to the experience of marginalization.31 Their shared experience, Pun and Huilin (2010) find, “precipitated anger, frustration, and resentment conducive to the emergence of the workers’ consciousness.”32 The central government has acknowledged that the new generation is different and, perhaps also attributing growing labor activism to generational change, emphasized society’s obligation to treat them better. In 2010, for instance, amid the strike wave in auto plants, then-Premier Wen Jiabao urged that the Chinese society needs to take better care of young migrant workers “as if they are our own children.”33 4.2.1.3 Particularistic and Generalized Framing Atomized incorporation can be understood as a type of blame avoidance strategy for the central government; it steers workers away from forming a causal belief that their everyday struggles are a product of the broader system for which the central government is ultimately responsible. In other words, atomized incorporation promotes a particularistic framing of workers’ plight, which could be politically safer because it places firms as the agents that create and maintain the system. A feature of atomization is that it reduces workers’ exposure to a generalized framing – the type of framing that generalizes workers’ individual experiences. When denied access to independent labor unions that would

31 32 33

Pun and Huilin, 2010; Chan and Pun, 2010; Wong, 2011; Pun and Chan, 2013. Pun and Huilin, 2010, p. 496. “Wen Prime Minister: Treat migrant workers as if you are treating your own children (wenzongli: xiangduidai zijide haiziyiyang duidai mingong)” 2010.

104

Politicization of Labor Discontent and Blame Attribution

educate workers about their place in the economic and political structure, workers might find it difficult to connect everyday struggles to the larger system. In the case of South Korea under military rule, studies find that voluntary social groups were critical in forming labor’s rebellious identity. For instance, Koo (2001) demonstrates that small groups formed by church and college students fostered workers’ collective consciousness and agency. Similar to Chinese migrant workers’ experience of dagong identity, workers in South Korea suffered from the social prejudice against industrial workers and internalized the degrading image. Negative self-perception led to internal attribution of blame and political passivity. The night schools (yahak), organized by church activists and college students, provided an opportunity for industrial workers to establish positive self-identity and develop active agency. Their training also contributed to the emergence of labor’s rebellious political identity by generalizing, and thus politicizing, the everyday struggles of workers. In contemporary China, there were underground organizations and activists – some directly influenced by the Korean activism – that made similar efforts to provide pedagogical training to migrant workers, but most have not survived the waves of crackdowns in the Xi era.34 Atomized incorporation further incentivizes workers to adhere to particularistic framing by increasing the positive returns to using such framing. Atomization alone might not be effective in the long run, as workers’ grievances about working conditions and career perspectives become more pervasive; as documented in the studies of pro-democracy movements, the labor force is an attractive target for mobilization by insurgent groups because of its large size and capacity to disrupt production.35 Even if workers might not form social grievances, they could be exposed to generalized framing of their struggles and swayed by it. When disgruntled workers are able to receive some material benefits by utilizing a particularistic framing, however, such positive inducements could lower the chance that workers focus their attention on such generalized framing. Since the early 2000s, by officially acknowledging the contribution of migrant workers to China’s economic growth and praising their role as the backbone of industrial labor, the central government began to display sympathy to their plight. The state has presented itself as a protector of migrant workers’ rights and interests. The state media’s reports on migrant labor issues since the mid-2000s present sympathetic accounts of 34 35

See Chapter 8 for a more in-depth discussion. Wood, 2000; Koo, 2001.

4.3 Patterns of Blame Attribution

105

migrant workers who experience labor disputes in their workplaces. The stories emphasize that migrant workers’ struggles originate from firms’ failures to abide by labor laws and end with a positive spin about how the state’s legal system can protect migrant workers. Stockmann and Gallagher (2011) describe the framing of these stories as “bad apples but happy endings” (p. 445). They send a message that poor workers are struggling at the hands of a few bad employers, but with the help of the state they are able to achieve justice. 4.3 PATTERNS OF BLAME ATTRIBUTION

Atomized incorporation since the early 2000s has framed workers’ struggles as a product of firm-specific policies, deflecting blame from the central government. This means that a majority of workers are conditioned to rely on internal attribution and are not likely to blame the central government. Yet, few studies have documented migrant workers’ blame attribution; while there are surveys of migrant workers, few have investigated their political attitudes, particularly about the central government. Using my survey data and qualitative evidence, I can assess empirically the extent and direction of blame attribution among different workers. As introduced in Chapter 1, the survey dataset was collected in Shenzhen – one of the largest manufacturing hubs – based on a random sampling of manufacturing firms during 2013. At each sampled firm, survey enumerators recruited four workers, two rank-and-file workers and two line supervisors. In total, the dataset includes 618 respondents from 155 factories, 307 line supervisors, and 311 general production workers. The survey oversampled line supervisors from each factory because they are known to play a central role in enabling or impeding atomized protests. Line supervisors have been touted as the new generation of labor activists in China due to their prominent role during large-scale labor strikes, including those at IBM, Pepsi, and Shanmukang Technology.36 As I detail in Chapter 5, even when they are not visible as protest leaders, without their tacit consent it is easy for potential protesters to be demobilized. In other words, junior-level supervisors are usually the gate-keepers of atomized protests. The sampling did not target migrant workers, but because they make up most of the workforce in factories, the majority of the respondents were migrants. Among general production workers, 98.7 percent had a 36

Harney and Ruwitch, 2014.

Politicization of Labor Discontent and Blame Attribution

0.7 0.65 0.55

0.6

Percentage of respondents

0.75

0.8

106

Central FIGURE

Local

Firm

Self

4.1 Reported perception of blame attribution: Who is to blame?

nonlocal hukou; among line supervisors, more had a local hukou but still the majority (89.9 percent) were migrants.37 4.3.1 Individual Heterogeneity To what extent do migrant workers blame the central government vis-ávis other actors? In the survey, workers were asked to rate their agreement with four separate statements: “If a migrant worker’s life is difficult, it is the responsibility of [the central government / the local government / the firm / the worker himself].” The answers were coded binary, 1 indicating their agreement with the statement and 0 disagreement, generating four variables: blame-central, blame-local, blame-firm, and blame-self. The data presented in Figure 4.1 show that respondents on average attribute most blame to the workers themselves and least blame to the central government. Internal attribution, in other words, is more common than external attribution. However, there was also a substantial difference between line supervisors and rank-and-file workers (Table 4.1). For line supervisors, internal attribution dominated: the majority of them (83.4 percent) indicated that 37

For a more detailed explanation of the sampling scheme and an overview of the dataset, please see Appendix 1.

4.3 Patterns of Blame Attribution TABLE

107

4.1 Perceptions of upward mobility, inequality, and fairness

My socioeconomic status will not get better (%) My life has no exit (%) Compared to locals, my income is too low (%) This society is unfair (%)

Supervisors

Rank-and-file

19.4

38.3 24.8 41.5 44.1

Source: Author’s survey.

the migrant worker himself should be blamed for his own plight. Contrary to the line supervisors, rank-and-file workers were almost equally likely to hold the firm and individual workers responsible. About 72.8 percent of respondents attributed blame to the firm, and 72.3 percent attributed blame to the individual. Similar to line supervisors, rank-andfile workers were also least likely to blame the central government. This confirms past studies on hierarchical trust that found that the central government enjoys relatively high levels of political trust and support.38 The data show that while rank-and-file workers are more prone to use external attribution than line supervisors, their target is less likely to be the state and more likely to be the firm. Why do line supervisors favor internal attribution? Unlike other managerial-level employees, line supervisors usually have a production background and achieved their position by internal promotion. They defied the slim odds and experienced upward mobility in their career, which is rare for unskilled migrant workers. The high returns to their individual efforts, therefore, are likely to have influenced their attributional beliefs. My interviews with low-level managers support this interpretation. Many expressed pride in having progressed upwards in the organization, in spite of the low odds of promotion and their relatively low levels of education. One line leader at a textile factory, having mistaken the author for a job seeker, stated: “Even if you are a migrant worker, you really should not be ashamed of low education. If you continue to develop your skills like me, you can also be in my position.”39 For migrant workers who were promoted, their belief in individual merit as a path to success could have contributed to their perception that individuals should take responsibility for their own plight. 38 39

Lü, 2014; Tang, 2016. Interview, 2012-07.

Politicization of Labor Discontent and Blame Attribution

0.5

Percentage of respondents 0.6 0.8 0.7

0.9

108

General production worker Blame−central FIGURE

Blame−local

Line supervisor Blame−firm

Blame−self

4.2 Reported perception of blame attribution by job position

Position in the workplace hierarchy (job position) has a substantial and statistically significant effect on blame attribution, even after controlling for other demographic variables (Figure 4.2). Table 4.2 shows the average marginal effect of demographic variables on the four dependent variables. The results confirm that line supervisors were less likely to blame the central and local government, and more likely to blame individuals. The biggest difference between the two groups lay in their assessment of the central government’s responsibility. Controlling for job position, those with higher income had a stronger tendency to blame the central and local government than firms or workers. Level of education did not have a substantial effect, but higher income – which could capture one’s skill level – predicted a stronger tendency to blame the central and local government.40 These patterns suggest that co-opting line supervisors might be important for the regime to maintain political resilience but not sufficient to ensure the overall political loyalty of migrant labor. The overwhelming 40

Since the models control for education, firm tenure, and length of working experience, wage differences are likely to capture skill level. Due to the difficulty of measuring skill level in survey data, the literature uses either education or income level as a proxy. Migrant workers do not vary greatly in educational attainment, so their skills could be better captured by wage level.

4.3 Patterns of Blame Attribution TABLE

109

4.2 Multivariate analysis of blame attribution Blame-central

Age Male Education Rural Wage (log) Supervisor Firm tenure Migration years Total N

∗∗

Blame-local

Blame-firm ∗∗∗

Blame-self

–0.014 (0.006) 0.056 (0.051) 0.005 (0.021) 0.008 (0.059) 0.308∗∗∗ (0.118) –0.256∗∗∗ (0.079) 0.009 (0.018) 0.022∗∗∗ (0.008)

–0.008 (0.005) 0.050 (0.053) 0.014 (0.020) –0.024 (0.057) 0.256∗∗∗ (0.111) –0.187∗∗∗ (0.065) –0.004 (0.015) 0.014∗ (0.008)

–0.017 (0.005) –0.045 (0.051) 0.002 (0.016) –0.044 (0.052) –0.138 (0.091) –0.011 (0.061) –0.033∗∗ (0.013) 0.033∗∗∗ (0.007)

0.007∗∗∗ (0.004) –0.027 (0.032) 0.016 (0.013) 0.063 (0.044) –0.194∗∗ (0.061) 0.133∗∗ (0.061) –0.003 (0.013) 0.009 (0.006)

602

602

601

602

Note: Average marginal effects were calculated based on logic regressions with firm-level random effects. Robust standard errors are presented in the parentheses. ∗∗∗ p < 0.01, ∗∗ p < 0.05, ∗ p < 0.10.

preference of line supervisors for internal attribution suggests that they might be less interested in converting the atomized protests into something more political compared to rank-and-file workers. Yet, a large percentage of line supervisors – more than 50 percent – still blame the central government. Given the capacity of line supervisors to mobilize atomized protests, it is still important for the central government to discourage them from using their position in the workplace for politicized activism. As I detail in Chapter 5, line supervisors have not posed a serious risk to the regime because they are not interested in atomized protests. Due to their superior position and firm-specific embeddedness, line supervisors are substantially less willing to join a protest in the first place. The results also reveal interesting trends in the effect of age and migration years. The two variables have opposing effects on the direction of blame attribution. As predicted by the literature on the new generation of migrants, older workers were more likely to attribute blame internally, while younger workers were more likely to attribute blame to external factors. In particular, younger workers were more likely to blame the central government and firms than older workers. The result confirms

110

Politicization of Labor Discontent and Blame Attribution

the intuition in the extant literature about the generational gap in political attitudes, although age might have a more continuous effect than what the literature suggests. The effect of age also supports the qualitative evidence that older workers seem to believe in the value of hard work more than younger workers. One firm manager stated, “older workers are grateful for their opportunities to work and willing to experience hardship (chiku). However, younger workers lack patience and believe that they deserve a better life without making real efforts.”41 Interestingly, the firm manager’s remarks reflect the government’s rhetoric about migrant workers in the past: For migrant workers to be considered worthy of a better life, they need to be willing to undergo hardship. My results suggest that the younger cohort is less willing to believe that rhetoric. When younger workers blame external actors, it is directed not only at firms but also at the central government. Migration experience also induced external attribution but had an opposite effect to age. Those with more years of experience as a migrant were substantially more likely to blame the central government, the local government, and firms. These two findings about age and migration experience signal that migrant workers’ grievances about the central government could grow in the future. While migrant workers may not direct a lot of blame on the central government, as the younger generations with higher social aspirations mature, they could become increasingly likely to attribute responsibility to the central government. 4.3.2 Social Grievances and Blame Attribution Why does migration experience matter? One potential explanation is that the experience of migration exposes workers to information that challenges the framing the regime uses to atomize migrant workers’ struggles. Stockmann and Gallagher (2011) argue that one reason why the regime’s framing works is because citizens lack access to sources of information that would challenge it. As identified in Chapter 3, atomized incorporation might fail to depoliticize labor discontent if workers are exposed to alternative sources of information that help them to see the commonalities in isolated cases of labor discontent. Migrant workers have high levels of labor mobility and change jobs often. Their experience of different jobs and geographic locations could be an alternative source of information that alerts them to the prevalence and systemic nature of discrimination against, and mistreatment of, migrant labor. 41

Interview 03-2014.

4.3 Patterns of Blame Attribution

111

To probe the determinants of blame attribution further, I analyze the effect of economic and social grievances. For conceptual simplicity, I call workers’ grievances about job conditions that dominate atomized contention “economic grievances” and those that go beyond firm-specific material conditions “social grievances.” Social grievances concern workers’ socioeconomic status in the society, including their perception of upward mobility, social fairness, and income inequality between urban and rural workers. This is not to suggest that economic and social grievances can be defined as a binary; the grievances about economic conditions that acknowledge their structural origins are both economic and social. In this analysis, however, I discern workers’ particularistic discontent about job conditions and their generalized discontent about their standing in the society. My hypothesis is that social grievances are more easily politicized than economic grievances since they concern the broader social system, not a specific job. This hypothesis argues against the claim that low wages and bad working conditions at a given workplace would directly induce politicized discontent of workers. My contention is that these grievances need to be channeled through workers’ social lens. I measured workers’ perception of upward mobility by giving them three options for what they expected would happen to their socioeconomic status in the next ten years: will get better, will stay the same, will get worse. The majority (71.1 percent) of workers reported that it would get better. However, as predicted, there was a substantial difference between line supervisors and rank-and-file workers. About 80.6 percent of line supervisors expected upward mobility, compared to 61.7 percent of rank-and-file workers. This means that a substantial percentage (38.3 percent) of rank-and-file workers did not expect their socioeconomic status to improve even in ten years. The survey asked additional questions about social grievances, but only to rank-and-file workers. Respondents were presented with three statements and asked if they agreed or disagreed with each one. The first statement, “As a migrant worker, my life does not have an exit,” was presented as an alternative measure for grievances about a lack of upward mobility. As discussed in Chapter 2, this is the language that migrant workers commonly used to describe their frustration. The second statement, “Compared with local people, my income is too low,” measured the perception of income inequality between local workers and migrants. Finally, the third statement, “Our society is very unfair,” was presented to gauge perception of social unfairness. Based on these three variables, I created a social grievances measure that was an average of the

112

Politicization of Labor Discontent and Blame Attribution

three answers, ranging from 0 (for those who did not agree with any of the statements) to 1 (for those who agreed with all of the statements). Out of the 311 respondents, 15.1 percent agreed with all of the statements, while 40.5 percent selected none. The data displayed in Table 4.1 show that grievances about income inequality between migrants and locals, and about social unfairness, are more widespread than those about limited upward mobility. While the respondents’ discontent about being trapped in the dagong life is still quite substantial – about a quarter of the respondents answered that there was no exit – respondents’ concerns over inequality and fairness were more prevalent. The findings suggest that a significant portion of migrant workers have developed social grievances but not all workers perceive the problems in the social structure to be serious. This leads to another question: What would drive a worker to form grievances about the overall social structure? Would migration experience induce stronger social grievances? To answer this question, I regressed workers’ demographic variables on social grievances. The first three columns displayed in Table 4.3 present the average marginal effect of demographic variables on no mobility. The first model includes both line supervisors and rank-and-file workers. The results show that older, female, less-educated, and low-income respondents were more likely to perceive limited upward mobility. The results might reflect the effect of preferred characteristics in the labor market. Since the labor market in China favors younger, male, skilled, and educated workers, individuals outside these categories might not expect their socioeconomic status to improve in the future. Rank-and-file workers and those with a longer firm tenure also reported that they perceived limited upward mobility. Except for monthly wage, all the other factors had a stronger and more statistically significant effect on rank-and-file workers. The preferred characteristics seem to affect rank-and-file workers more, perhaps because line supervisors were promoted because of these characteristics. The effect of age and gender, however, changed direction when I examined the determinants of social grievances. The final two columns in Table 4.3 show that younger and male workers were more likely to report social grievances than older and female workers. The results stay the same if I examine their effects on the perception that there is no exit, an alternative measure of grievances about upward mobility.42 The 42

A one year increase in age decreased workers’ perception of no exit by 1.0 percentage points, while male workers were 10.6 percentage points more likely to indicate perception of no exit.

4.3 Patterns of Blame Attribution TABLE

113

4.3 Multivariate analysis of social grievances

Age Male Education Rural Wage (log) Supervisor Firm tenure Migration

No mobility

No mobility (supervisors)

No mobility (rank-andSocial file) grievances

Social grievances

0.010∗∗∗ (0.004) –0.111∗∗∗ (0.039) –0.041∗∗ (0.016) –0.038 (0.049) –0.245∗∗∗ (0.089) –0.156∗∗∗ (.055) 0.019∗ (0.010) 0.009 (0.006)

–0.003 (0.005) –0.016 (0.055) –0.016 (0.017) –0.063 (0.051) –0.293∗∗∗ (0.111)

0.011∗ (0.006) –0.180∗∗∗ (0.048) –0.049∗∗ (0.024) –0.042 (0.077) –0.293∗ (0.164)

–0.008∗∗∗ (0.003) 0.066∗∗ (0.027) –0.019∗ (0.011) 0.082∗∗ (0.041) –0.098 (0.085)

–0.008∗∗∗ (0.003) 0.065∗∗ (0.027) –0.016 (0.011) 0.099∗∗ (0.042) –0.112 (0.085)

–0.004 (0.010) 0.011 (0.007)

0.066∗∗∗ (0.024) 0.016 (0.010)

–0.004 (0.009) 0.019∗∗∗ (0.005)

597

296

301

304

0.002 (0.009) 0.013∗∗ (0.006) 0.020∗∗∗ (0.007) 302

Jobs Total N

Note: The reported average marginal effects are based on logic regressions with firmlevel random effects in the first three columns. The final two columns report marginal effects based on OLS regressions. Robust standard errors are presented in parentheses. ∗∗∗ p < 0.01, ∗∗ p < 0.05, ∗ p < 0.10.

divergence between the effect of age and gender on these two variables – no mobility and social grievances – suggests that they might not measure the same underlying dynamic. Those who believe in the potential to improve their socioeconomic status might not necessarily believe that there is an exit from a life as a migrant worker. Younger and male workers, while more likely to express the belief that their economic situation would improve in the future, reported more social grievances. These seemingly contradictory results might come from these groups’ different standards of upward mobility. Migrant workers’ wages and working conditions have improved continuously since the mid-2000s, which explains why a substantial portion of migrant workers do report optimism about enjoying a better economic situation in the future. Yet, many workers also do not seem to consider a small wage increment as meaningful upward mobility. “As a migrant,” one worker said, “your

114

Politicization of Labor Discontent and Blame Attribution

wages might go up but you can never exit this life. You might earn a few hundred yuan more as an experienced worker, but you can never have a good job, a job that is not tiring, hard, or boring but still pays you well.”43 In fact, about 17.4 percent of those who answered that their socioeconomic status would get better in the future indicated that there was no exit out of the migrant life. These respondents were also predominantly young and male. Younger workers, after controlling for the length of migration, were much more aggrieved about social inequality and unfairness. They might believe that – with many years of work ahead of them – their economic conditions would get better, but still be frustrated about being stuck in the dagong life. Likewise, male workers may believe that their life would get better due to their superior market position as male workers, but still form strong grievances about limited upward mobility, because compared to female workers they have greater aspirations for economic success. My interviews corroborate the findings. While male workers often discuss their responsibility as the main bread winner in the household and the necessity of increasing their earnings, many female workers see their migrant work as a temporary job between family duties. Finally, the positive effect of migration years shows that migration generalizes individual experiences. While the years of migration had a positive effect in all models, the effect of migration years was larger and more statistically significant for social grievances than for no mobility. Ten years of migration increased workers’ perception of social grievances by one standard deviation. The finding is further supported by qualitative evidence. In interviews, experienced migrant workers described how their perception of society and their own lives had changed substantially. “When I just started out to dagong,” one worker from Hunan at an electronics factory in Shenzhen stated, “I had so many hopes and dreams. I saw older sisters in my village who went out to migrate and came back wearing nice clothes. I really envied them. That’s why I came out to dagong without completing high school.” Her first job six years ago was at an apparel factory in Guangzhou where she worked with her relative. She said she enjoyed her life in the big city, because she had much more freedom away from her family and could buy things with the money she earned. She liked her financial independence. The city life that she was experiencing for the first time also gave her new sources of excitement.

43

Interview, 06-2013.

4.3 Patterns of Blame Attribution

115

Yet, as she continued to work as a migrant worker and changed jobs many times, she realized her life might not improve substantially. I was not able to save much money. At first, I was very bad with my finances and did not even bother to save. My first job in the apparel factory did not pay a lot because it was based on piece rates and I was not quick enough. My relative also wanted to move somewhere else because she had worked in Guangzhou for many years. We heard from a fellow villager that there was a factory in Shenzhen that paid more. We came here and started to work at an electronics factory. It seemed better because the factory was bigger and had a nice canteen. The new job indeed paid more. I was happy. But, it did not seem enough. I still could not save much money. I wanted to open my own shop after going back to the village so I wanted to find a better job. [Then, how many jobs did you have after that?] I changed my job many many times. I cannot even count. I also worked in Shanghai, but came back to Shenzhen last year. What I realized during the years is a job is a job. No matter where I work, I am not saving enough money. During the years, I really thought that I should go back to my village after earning enough money, because in the city I am always an outsider. But, now that I have worked so many years but do not have much money, I am not even sure. If I go back without money, people will look down on me.

This worker might not represent migrant workers as a whole. She was much more entrepreneurial and showed a stronger desire to establish her own business than other workers, particularly compared to many female workers. Yet, her story shows how the experience of migration and job hopping could contribute to growing social grievances, because at some point, workers see the pattern behind their experiences in various firms and regional locations. The final column in Table 4.3 confirms this: When I introduced the number of jobs prior to the current one into the model, it revealed a substantial influence over workers’ perception of social grievances. Both the length of migration and the number of jobs induced a stronger perception of social grievances. An interview with a protest participant at Yue Yuen factory in 2014 also highlights that years of experience working as a migrant could induce frustrations about social fairness and justice, as workers realize that even after so many years of hard work, they are not substantially better off:44 For many workers, this is a simple and honest feeling. We were once proud to be Yue Yuen workers. We just want a decent job and a dignified life. We love the Communist Party and our homeland, and we hope our country can develop better. But when we workers cannot benefit from the sustained economic development, I think it’s not right.

44

“An in-depth discussion with a veteran worker at Yue Yuen” 2014.

116

Politicization of Labor Discontent and Blame Attribution

If we go home, we don’t have capital for investment nor do we have professional skills, so it’s hard, because we just don’t know what we should do. Things we learnt in the factory cannot help at all, and there are no factories like Yue Yuen at home. What can we do when we are old? Many factories don’t recruit workers over 35-years-old so we can’t work at the factory for ever.

The findings show that migration experience is an important source of information that exposes workers to generalized patterns. Migrant workers are very mobile and experience a large number of firms and regions. After working in many jobs and living in several regions without any substantive improvement in their social or economic standing, they might be less inclined to believe that their life struggles are unique to them or to a specific workplace. This type of information could be akin to what Ferree and Miller (1985) call “consensus information,” which leads to the perception that “other people have had the same experience and that one’s reactions are not idiosyncratic.” They argue that while individuals often tend to blame themselves, “the experience of the problem in a context of consensus” increases the chance of external attribution (p. 45). This could explain why workers with more years of migration become less susceptible to attributing blame internally. Similarly, job hopping could induce the perception that migrant workers’ plight is present in all firms and is not necessarily a reflection of a few “bad apples.” Estimating the relationship between social grievances and blame attribution confirms that social grievances politicize labor discontent and encourage external attribution of blame, particularly to the central and local government, but mostly among rank-and-file workers. I regressed social grievances on the extent of blame attribution, controlling for demographic variables. I also included job-specific grievances about wages as an independent variable to compare the effects with those of social grievances. The results highlight interesting differences between line supervisors and rank-and-file workers. As shown in Figures 4.3 and 4.4, the effect of wage grievances is most substantial on attribution of blame to firms, although its extent differed between line supervisors and rank-andfile workers. For rank-and-file workers, wage grievances had a very large and statistically significant effect on blame attribution to firms but did not yield a statistically significant effect on attribution to other actors. For line supervisors, wage grievances reduced internal attribution and increased external attribution to firms and local governments but did not have a statistically significant effect on attribution to the central government. This suggests that wage grievances – the most common type of economic grievances about job-specific conditions – are not easily converted into

4.3 Patterns of Blame Attribution

117

Central government

Local government

Firm

Self

Pay increase

No mobility

Pay increase

No mobility

−0.2

0

0.2

0.4 −0.2

0

0.2

0.4

FIGURE 4.3 Effect of economic and social grievances on blame attribution: Line supervisors

Central government

Local government

Firm

Self

Pay increase

No mobility

Pay increase

No mobility

−0.2

0

0.2

0.4 −0.2

0

0.2

0.4

FIGURE 4.4 Effect of economic and social grievances on blame attribution: Rankand-file workers

118

Politicization of Labor Discontent and Blame Attribution

a political issue that can be used to target the central government, but that they have a large and substantive effect on workers’ attitudes toward firms. The effect of social grievances was starkly different between the two groups of respondents. The variable no mobility did not show a substantial effect on line supervisors’ attribution of blame to the central and local government but induced a positive effect for rank-and-file workers. The result shows that social grievances are more important than economic grievances in politicizing labor discontent among rank-and-file workers – the majority of migrant workers who have not experienced upward mobility in the workplace. The variable also did not encourage respondents’ attribution of blame to firms, implying that workers do not consider their limited upward mobility to be the responsibility of firms. The substantive size of the politicization effect was also large: no mobility increased rank-and-file respondents’ attribution of blame to the central government and the local government by 8.8 and 10.7 percentage points, respectively. I replicated the analysis with the variable social grievances (see Figure 4.5). The conclusions are similar and the effect size was even larger. Among rank-and-file workers, social grievances contributed to a substantial increase in their attribution of blame to the central government. When the variable social grievances changed from its minimum value (0) to the maximum (1), it increased the percentage of the respondents who blamed the central government by 39.5 percentage points. The effect size on blame attribution to the local government was slightly smaller but still substantial, at 24.1 percentage points. As with no mobility, social grievances did not induce a substantial change in workers’ attribution of blame to firms; the variable, however, discouraged internal attribution more strongly. Overall, these results indicate that the politicization effect is stronger when we consider social grievances and confine our analysis to rank-and-file workers. Economic grievances, on the other hand, did not have a clear effect on attribution of blame to the governments but changed the extent to which workers held firms responsible. The presence of economic grievances, therefore, does not challenge the logic of atomized incorporation, since workers’ discontent is still constrained within the firm boundary. Social grievances, however, appear to be an important determinant of political grievances against the central government, at least among rank-and-file workers. The findings question the logic of atomized incorporation as a tool of material co-optation.

4.3 Patterns of Blame Attribution

119

Central government

Local government

Firm

Self

Pay increase

Social grievances

Pay increase

Social grievances

−0.5

0

0.5

−0.5

0

0.5

FIGURE 4.5 Effect of economic and social grievances on blame attribution: Alternative independent variable

Even when the regime is able to mitigate economic grievances by rewarding short-term economic gains to protesters, it might not be effective in decreasing the political grievances of the majority of workers. It could, however, be useful in discouraging line leaders with mobilizational capacity within workplaces from politicizing their discontent and targeting the central government. 4.3.3 Effect of Protest Behavior and Information The previous section showed that social grievances are relatively widespread among rank-and-file workers and induce politicization of labor discontent. As workers become more aggrieved about the social structure, they also become more likely to attribute blame to external actors, in particular the central government. An important mechanism by which ordinary workers’ discontent is politicized is the accumulation of experience in different workplaces. By experiencing different jobs and locations, workers are exposed to the patterns behind their individual grievances and the broader social structure.

120

Politicization of Labor Discontent and Blame Attribution

Do these findings mean that atomized incorporation has largely failed to depoliticize labor struggles? The next question that I analyze is whether a worker’s direct or indirect experience of collective action affects the politicization of labor discontent. If atomized protests help the regime to depoliticize labor conflicts, then we would expect protest participants to be more likely to blame firms rather than the central government. If atomized incorporation creates protesters who blame the central government, it poses a significant political risk to the regime: While protests appear apolitical, the underlying political grievances could induce politicized contention, using the tolerated space for collective action. Indirect experience of protests – such as hearing about protests in other firms – can have both negative and positive effects on the regime’s efforts to depoliticize labor discontent. On the one hand, if workers learn that workers in other firms achieved success via atomized protests, they could internalize the regime’s rhetoric that labor grievances can be resolved by targeting the firm. On the other hand, information about other protests could contribute to politicization of labor discontent because it reveals the extent of labor grievances.45 When workers hear about other labor protests and conclude that labor grievances are widespread, they might be more likely to attribute labor discontent to the central government. Direct and indirect experiences of a labor protest were measured by past protest behavior (protest behavior) and exposure to information about other protests (protest information). Respondents were asked whether they had participated in any collective action. About 8.4 percent of workers indicated that they had protested at least once in the past.46 We asked workers, “Have you heard about other firms’ labor strikes?” About 45.2 percent of the respondents answered positively, demonstrating a high level of indirect experience of labor protests. Considering the frequency of labor contention in Shenzhen, the result is not surprising.

45

46

Studies on pro-democracy protests emphasize the role of information that reveals the extent of underlying grievances (Kuran, 1991; Lohmann, 1994a). My analysis focuses on a different causal mechanism. While the literature stresses the effect of information on citizens’ perceptions of the likelihood of success – since widespread grievances usually lead to higher turnout – I examine whether the information induces a cognitive process that politicizes the discontent. This percentage is much higher than in the data from national representative surveys where less than 1 percent of migrant workers had participated in a labor dispute (Cai and Meiyan, 2012, p. 18). The difference might be due to a combination of sectoral and regional factors. Collective action by migrants is substantially more prevalent in manufacturing sectors and in Shenzhen.

4.3 Patterns of Blame Attribution

121

Central government

Local government

Firm

Self

Protest behavior

Protest information

Protest behavior

Protest information

−0.4

−0.2

0

0.2

0.4 −0.4

−0.2

0

0.2

0.4

FIGURE 4.6 The effect of protest behavior and information on blame attribution Note: Average marginal effects based on logit regressions with firm-level random effects are reported. The standard errors are clustered at the firm level.

Those who had heard about another labor strike were also more likely to report having participated in a protest themselves. While only 4.5 percent of those who had not heard about a labor strike reported having joined a protest, 13.4 percent of those with an indirect experience did so.47 I ran logistics regressions on blame attribution with the two independent variables, protest behavior and protest information.48 The average marginal effects are displayed in Figure 4.6. The results demonstrate that the two variables had a different effect on blame attribution. As hypothesized, those who had participated in a labor protest were less likely to blame the central and local government – by 20.7 and 26.7 percentage points respectively – but more likely to blame firms by 20.1 percentage points. While it did not affect the extent of internal attribution, it affected who workers attributed blame in a way predicted by the logic of atomized incorporation. 47 48

The effect of indirect experience on participation was also statistically significant, after controlling for demographic variables. The models included basic demographic variables as control variables. The results did not substantially change even when I included social and economic grievances as additional control variables.

122

Politicization of Labor Discontent and Blame Attribution

These findings should be interpreted with caution as the analyses do not reveal the causal direction. On the one hand, the experience of atomized contention itself could make workers believe that firms are the direct authority to resolve their labor grievances, particularly given its high success rates.49 On the other hand, another explanation might be that protesters differ in their attributive beliefs ex ante, and only those who hold firms responsible would have participated in a protest. In Chapters 5–7, I show that the latter explanation might be a better fit for the patterns of atomized protests, since those with political grievances often do not see atomized protests as a viable option to resolve their discontent. Exposure to protest information, on the other hand, promoted external attribution, particularly to the central and local government. The variable increased attribution of blame to the central government by 7.2 percentage points, although the effect was not statistically significant, and to the local government by 14.8 percentage points. Protest information also reduced the number of respondents reporting internal attribution by 5.4 percentage points. In general, the patterns of the effect resemble those of migration experience, which also functions as a channel of information, but the extent to which workers generalize the patterns of grievances seems to be more at the regional level than the national level. Given that the number of labor protests is highest in a few cities in Guangdong Province, workers who learn about the protests in the region might not immediately hold the central government accountable. In other words, workers’ exposure to other protests seems to pose a bigger threat to local governments than to the central government at the moment. Yet, the analyses still indicate that the frequency of atomized protests could matter for the regime’s maintenance of political stability. 4.4 CONCLUSION

In this chapter, I examined the conditions in which migrant workers report social grievances and attribute blame to different external actors. While not all migrant workers have high levels of social grievances, social frustration is higher among those who experience more years of migration. I argue that migrant workers’ job mobility leads experienced workers to perceive a generalized pattern of labor issues and to begin

49

In Chapter 5, I also show that workers who had participated in a labor protest are substantially more willing to protest again and perceive the chance of success to be high.

4.4 Conclusion

123

to attribute their individual circumstances to external factors, particularly the central and local government. Among rank-and-file workers, the extent of social grievances is a stronger predictor of blame attribution than job-specific grievances, raising the question of how successful the regime can be in depoliticizing labor discontent by delivering short-term material gains. The findings have important implications for the success of atomized incorporation. Given that social grievances will continue to grow, as younger workers who have higher social aspirations and desire for a better economic future accumulate migration experience, is the regime able to avoid blame by allowing atomized protests? My analyses show that the regime has enjoyed partial success. Those who have participated in a protest are more likely to blame firms than to blame the central or local governments; the prevalence of atomized protests has not induced substantial grievances about the central government. Yet, the findings I present in this chapter also suggest that atomized protests might not fully reflect the extent of grievances about the broader social structure and the central government. To better understand what grievances are being expressed in atomized protests, we need a more in-depth understanding of how workers are mobilized to engage in collective action within workplaces. There could be structural factors that shape individual decisions and collective dynamics in the workplace. In Chapters 5–7, I build on my findings and investigate the structural factors underlying the dynamics of collective action and firm-level patterns of collective action.

5 Workplace Mobilization and Collective Action

The regime’s strategy of atomized incorporation creates a more lenient political environment for workplace-based protests in China, and changes in the labor market have increased workers’ structural power. Yet, these changes have not directly translated into workers’ collective action. Given the widespread dissatisfaction with wages and working conditions, it appears that not all underlying grievances are being expressed as collective action. While the number of labor disputes and protests has increased, many factories that offer low wages and subpar working conditions still do not experience labor protests. Studies have pointed to the regime’s suppression of labor organizing as a critical barrier to workers’ collective action. Without an autonomous organization, workers do not have the resources to lead, plan, organize, and mobilize participants, so protests – even when they do occur – are usually short-lived. The emphasis on the regime’s repression of autonomous labor organization, however, has largely ignored how this restriction interacts at the firm and individual levels. Macrolevel variables are important for explaining temporal changes, but they should be linked to microlevel explanations to understand how and why collective action emerges at different types of firms. To understand the political implications of atomized protests, we must also understand the firm-level dynamics of collective action. The mechanism by which individual perceptions and beliefs translate into firm-level action merits further study. In this chapter, I study how heterogeneous workers overcome collective action problems to understand why collective action might still be difficult to initiate, even in a relatively favorable environment. Moving 124

5.1 Barriers to Collective Action for Unorganized Workers

125

away from the conventional focus on the free-rider problem, I argue that the biggest impediment to collective action among unorganized workers is a shortage of first movers – the first mover problem. The heterogeneity of workers in their personal resources and sense of grievance helps them to overcome the first mover problem, but only when those who have the resources to mobilize them are interested in collective action. Using qualitative interview evidence and my survey data, I analyze workers’ protest intention and the determinants of their protest intention, with a focus on understanding different workers’ capacity to mobilize within a firm. 5.1 BARRIERS TO COLLECTIVE ACTION FOR UNORGANIZED WORKERS

5.1.1 Free-Riding Problem What is the most acute collective action problem for unorganized workers? The traditional literature emphasizes that collective action often fails because of individual incentives to free-ride. In his book The Rebel’s Dilemma, Lichbach (1998) says of collective action by political dissidents: Rational actors will not participate in collective action, even if they support the cause, because they can free-ride. Potential dissidents realize that while each individual needs to pay the private costs of participation in collective action, even nonparticipants will benefit. If the results can be enjoyed without participation, a rational individual will choose not to be seen protesting. The author illustrates the dilemma with an example of a college student deciding whether to join a university-wide protest: Jane Smith, an undergraduate at the University of Colorado, strongly supports the principle of diversity in employment. [. . . ] Jane thus advocates a vigorous affirmative action program, one with quotas in various job categories for certain protected classes. The Cause, a campus civil rights organization, holds similar beliefs and advocates a similar program. Moreover, it has just announced that it plans to picket the Administration next Monday morning, starting at 9:00 A.M, and has called for volunteers.

Lichbach shows that Jane faces a situation analogous to the Prisoner’s Dilemma. Whether other people join the demonstration or not, she is always better off by choosing to stay at home. If other people join, Jane can still enjoy the benefits of diversity without joining the protest. If other people do not join, no one can access the benefits of diversity anyway, so she does not need to pay the costs of participation. Therefore, under situations parallel to Jane’s, Lichbach concludes, “rational people will never rebel” (p. 7). The rebel’s dilemma predicts that migrant workers, even if they want to achieve better wages and working conditions, would not

126

Workplace Mobilization and Collective Action

participate in a protest because they believe that they can still benefit if others protest. Scholars have pointed out that factors such as selective incentives or group pressure should be present for the individuals to overcome the free-riding problem. 5.1.1.1 First-Mover Problem The rebel’s dilemma, however, does not address the most serious impediment to collective action in a context where there is no organization or central leadership. In Jane’s example, the rebel’s dilemma presumes that The Cause – the campus civil rights organization – makes the first move by announcing that there will be a protest. Jane decides whether to join the protest or not under the assumption that the probability of everyone else joining the protest is fixed – that is, it does not vary conditional on Jane’s participation. Under this assumption, Jane would never be a first mover because the central organization already exists. What happens if there is no organization? For collective action to occur, somebody has to initiate it – but initiating a protest might be high cost with low expected return. The initiator has to plan, choose the date and time, find a venue, secure permits, mobilize participants, and announce the protest. She might be uncertain whether other people will join. If no one joins, the initiator’s efforts will be wasted. Oliver et al. (1985) and Marwell and Oliver (1993) label the lack of first movers the “accelerative collective dilemma.” They posit that some types of public goods have an accelerating production function, where the marginal return of a unit contribution of resources increases as more resources are invested. For these types of public goods, initial contributors pay high costs of collective action with only low levels of expected return. Collective action for decelerating public goods, on the other hand, is more vulnerable to free-riding, since initial contributions reduce the next person’s willingness to participate. Most public goods have an S-shaped general third-order production function where the marginal contribution of a unit added to the provision of the public good increases up to a certain point and then decreases. Collective protests by unorganized workers are a public good with an S-shaped production function and a long accelerating range. For instance, the first thirty workers to strike in a hundred-employee firm might not induce much response from management, but when the next thirty people join the strike, management might start to worry. Moving from thirty to sixty protesters could quickly shift the management’s response from minimal to large. If workers know that the first thirty protesters might

5.1 Barriers to Collective Action for Unorganized Workers

127

make little difference, no one would try to initiate the strike, even though once the first thirty workers join, other people might become more willing to join. Therefore, as Oliver et al. (1985) assert, “the usual outcome of the accelerative collective dilemma is that nobody rides free because nobody contributes and there is no ride” (p. 547). 5.1.2 Solution Prior research on contentious collective action has suggested that group heterogeneity in interest level could mitigate the first-mover problem. The previous discussion of the accelerative production function assumed that each worker attaches the same value to their labor protest. However, the assumption is unrealistic. If workers’ preferences for collective action are heterogeneous, those with exceptionally high interests in the outcome can take the role of first movers. Tilly (1977), for example, introduced the term “zealots” to denote those “who, compared to other groups, set an extremely high value on some collective good in terms of the resources required to achieve that good” (p. 88). These individuals are so interested in acquiring the goods that they will do anything in their ability to produce the goods, regardless of other people’s contributions. In the case of labor protests, some workers could have extraordinary levels of interest in either producing collective action (an occurrence of collective action) or producing the collective goods (getting the pay raise). While no one would dispute that highly motivated individuals might be more willing to initiate a collective protest than others, extreme interest alone is not sufficient. Studies have argued that protest turnout by individuals with extreme interests might not be conducive to mobilizing others to join.1 Turnout by individuals with extreme interests, compared to that of more moderate individuals, is not a strong signal that grievances are widespread and collective action is likely. Other scholars suggest that instead of those with extreme interests, groups should rely on resourceful individuals to overcome the initial stage of the accelerative production curve.2 In Olson’s language, some organizations could be “privileged,” when some members of the group are resourceful enough to take full responsibility for providing the public good.3 While the study is mostly concerned with direct production of collective action – a resourceful individual could contribute additional units 1 2 3

Lohmann, 1994a, 1994b, 2000. Oliver et al., 1985; Marwell and Oliver, 1993. Olson, 1965, pp. 49–50.

128

Workplace Mobilization and Collective Action

of resources to enhance the chance of securing the collective goods – the logic could be applied to cases such as labor protests in which every member can only contribute one unit, her own participation. Some workers though can contribute more units if, by their participation, they induce enough other workers to join to the point that the threshold for collective action is reached. In the face of the first mover problem, an organization benefits greatly from individuals with mobilizational resources. 5.2 PROTEST INITIATORS IN CHINESE FACTORIES

Chinese citizens have long been characterized as unorganized. The description applies to many social groups, not just migrant workers. Since the regime has discouraged social organization outside the state institutions, social groups have a weak organizational basis for engaging in political actions that are not sanctioned by the state.4 Without a formal organization, protesters rely on their personal networks and communal ties for mobilization, and protest leadership becomes a critical condition for effective mobilization.5 Studies of labor protests of laid-off SOE workers in China have also highlighted that protest leaders play an important role in defining the protest strategy and mobilizing other workers, enhancing the chance of success.6 The leaders were often depicted as cadres with substantial political training due to their experience during the Mao era. Cai (2006) shows that most protest organizers were “the elite among workers who wield more influence than the rank and file” (p. 107). These protest leaders had an advantage over their followers because they were experienced in dealing with the state bureaucracy. They knew how to frame their protest to maximize the perceived legitimacy of their claims in the eyes of the state. According to the research by Chen (2008), protest leaders were often “former Cultural Revolution activists” who were able to command the right Maoist rhetoric to frame their protests (p. 89). Protest leadership among migrant workers, on the other hand, has received only limited attention. Migrant workers’ protests have commonly been characterized as spontaneous wildcat strikes with not much planning behind them.7 Since collective action has been perceived as a last resort of migrant workers, arising from anger and desperation, there has been limited analysis of their strategic agency. More recent studies and observations, however, suggest that we have not understood the role of protest leaders in achieving collective mobilization. Leung 4 5 6 7

Zhou, 1993. Li and O’Brien, 2008. Cai, 2002, 2006; Chen, 2008. Lee, 2002; Chen and Xu, 2012.

5.2 Protest Initiators in Chinese Factories

129

(2015) finds that, even without independent labor organization or representative trade unions, workers’ protests have become more organized and planned. It is especially difficult for the offensive protests to emerge without strong leadership, because compared to defensive protests, the demands are much more complicated and it is harder to secure workers’ general support.8 This suggests that protest leadership could be a critical factor for interest-based collective action, when workers’ demands go beyond the legal minimum. Since workers cannot rely on the law as the standard for justice, it is much more difficult to generate consensus about what constitutes “reasonable” treatment. This means that mobilization for interest-based goals often requires more sophisticated organizational tactics, and hence stronger leadership. 5.2.1 Low-Level Managers and Mobilizational Resources Who initiates and leads protests in Chinese factories? Case studies and media reports have paid attention to the role of low-level managers, who in many cases initiated and organized collective action. Chris Chan, for example, notes in his cases that “line supervisors were the front-line organizers and leaders” and average production workers were “generally enthusiastic followers.”9 Leung (ibid.) similarly described the majority of strike leaders he encountered as “division foremen, line supervisors, or skilled workers” (p. 126). Despite their instrumental role in labor mobilization, we have a limited understanding of why low-level managers become protest leaders, and how they differ from rank-and-file workers in important characteristics such as preferences for protest, the extent and nature of grievances, and beliefs about government. Their active engagement in collective action is interesting, given that their relationships with rank-and-file workers are not always cooperative. Line supervisors are the lowest-level managers, so they absorb discontent and grievances at the front line; they are also expected to meet production targets and boost productivity, which potentially brings them into conflict with the workers they supervise. Rank-and-file workers, particularly those that are unskilled, often complain about their line managers having a negative attitude toward them and putting undue pressure on them to increase the speed or improve the quality of production. Yet, evidence suggests that the role of line supervisors in initiating collective action is crucial, because of their capacity to mobilize and 8 9

Leung, 2015, Chapter 4. Chan and Siu, 2010, p. 91.

130

Workplace Mobilization and Collective Action

de-mobilize protesters. Compared to rank-and-file workers, line supervisors are better placed to gain trust and respect from fellow workers. One worker activist argues that a typical strike initiator (bagong daitouzhe) is well connected with other people and trusted by them.10 Line supervisors have usually been at the firm for longer than rank-and-file workers and have developed a stronger social network at the firm. They are usually promoted from ordinary production roles, which suggests that they performed well and are considered to be competent. While there are conflicts between line supervisors and ordinary workers, their relationships are still generally more amicable than those between production workers and higher-level management. Rank-and-file workers identify more with the low-level managers who have a production background. Line managers also often express their solidarity with rank-and-file workers over higher-level management, although they should appear to follow directives from management. An interview by Leung (2015) with a strike participant at an ornament factory illustrates this point11 : Usually only workers with good skill and good personality can gain the support and respect of other workers, and perform that role of [strike] leader. It is not easy to find a rank-and-file worker with such quality. But without such a leader workers would [find it] difficult to develop mutual trust for taking the action. In this Super Ornament case, the position of supervisor was crucial.

After a labor strike at a bulb factory in Foshan, one of the protest participants explained that line managers, while not coordinating the strike, knew about the strike and did not tell the higher-level managers: “The fact that higher-level managers were not even aware of this plan itself shows that there was some kind of tacit consent of line managers. Sometimes line managers know that a few workers are trying to create troubles but they pretend that they do not know of it. For one, the strike also benefits line managers. Also, as in this case, if those workers are the ones that are close to the line manager, those that have stayed with him for some time, then line managers also understand the workers. It is not always that these low-level managers are on the firm’s side.”12 My survey also supports this view. When general production workers were asked, “Do you agree with the following statement: I share the same interests with [my firm manager/ my line supervisor]?” they displayed a much stronger 10 11 12

Hao, 2016. Leung, 2015, p. 127. Interview 04-2012.

5.2 Protest Initiators in Chinese Factories TABLE

131

5.1 I share the same interests with [. . . ] Firm manager

Line supervisor

Highly agree Somewhat agree Somewhat disagree Highly disagree

22 (5.2%) 172 (32.3%) 100 (55.5%) 16 (7.1%)

66 (21.3%) 213 (68.7%) 27 (8.7%) 4 (1.3%)

Total N

310

310

Note: The table only displays answers from ordinary production workers.

sense of proximity to their line supervisor (see Table 5.1). While 90 percent of general production workers indicated that they share the same interests as their line supervisor, only 62.6 percent of them did so for the firm manager. Line supervisors also appear to have positive relationships with the workers on the line. When asked whether they share the same interests as their subordinates on the line (xiashu gongren), 97.8 percent of line supervisors said yes. Although a majority of line supervisors – 76.0 percent – also stated that they share the same interests as the firm manager, their trust for rank-and-file workers was substantially higher than for the firm manager. In other words, both rank-and-file workers and line supervisors are more likely to consider each other to be allies, than they would consider themselves to be allies of higher-level management. Ordinary workers having more trust in line supervisors than in higherlevel firm managers could play an important role during mobilization, because line supervisors can control information. For example, a labor strike might be sparked by a new piece of information. Many firms in the labor-intensive manufacturing sector lack information transparency, so workers often rely only on their social network for information. The firm could claim that they have had several months of lower income due to a declining global market. If the message from line supervisors differs from management’s version, workers are more likely to believe line supervisors, and management could lose control over the narrative. This point is supported by the survey data. General production workers display much higher trust in information about the firm that is delivered by line supervisors (88.4 percent) than by the firm manager (54.5 percent). Case studies show that line supervisors are the group most likely to have the knowledge and organizational skills to make labor protests successful. The perception that line supervisors could enhance the likelihood

132

Workplace Mobilization and Collective Action

of success helps them to mobilize other workers more easily. For instance, a worker who participated in a strike at a battery factory acknowledged that the line supervisors greatly enhanced the chances of collective action succeeding because they knew how to allocate people and use the organization of production as a resource: [T]his time line supervisors took part in planning and how to make the strike successful. They appointed different batches of workers to block places, and they arranged shifts to guarantee the blockage was maintained around the clock. The strike was efficiently organized, similar to the organization of production (emphasis added). [. . . ] Women workers in the assembly department got very excited when they noticed the success of their action—earlier in May their strikes were unsuccessful. Some commented, “Hey, when line supervisors join us, it really makes a difference!”13

The quote illustrates that line supervisors have better organizing skills for protest activity, because they can apply their experience as a production line manager to collective action. They also have access to inside information that could be important for maximizing workers’ bargaining power. Line supervisors can use their knowledge of production and shipping schedules to optimize the timing of a strike. Line supervisors know that timely production and shipping are paramount, especially for exportoriented firms, and are in a position to exploit that knowledge. One strike organizer commented, “We have to grasp the opportunity. Only at times when the order is rush and workers are difficult to recruit [are we] able to win the bargain.”14 Even if line supervisors do not have a good relationship with every worker on the line, they could exercise authority over ordinary workers and exert social pressure. Ordinary workers depend on their line supervisors for benefits such as overtime shifts and promotions, so for them it can be hard to resist line supervisors’ orders. A dialogue between a worker activist (A) and a senior worker at Yue Yuen factory (B) acknowledges the authority of line supervisors even over new workers:15 B: At Yue Yuen, it has often occurred that the Taiwanese [the firm manager] wants to get rid of a line manager, and then that manager allies with other managers from the same locality to lead their subordinates to strike. The Taiwanese then cannot deal with the situation himself and in the end has to ask the line manager to resolve the issue. A: Do those strikes initiated by line managers take place frequently? B: Yes. I have heard many of them. 13 14 15

Globalization Monitor (2009, pp. 163–4), re-quoted from Pringle (2011, p. 103). Leung, 2015, p. 79. “An in-depth discussion with a veteran worker at Yue Yuen” 2014.

5.2 Protest Initiators in Chinese Factories

133

A: Would new workers also join the strikes? B: If you are a new worker, and your manager leads you to strike but you do not join, can you keep working on that line? The manager will curse you.

Even when line supervisors are not protest leaders, their support for collective action can be crucial. Line supervisors are well placed to mobilize collective action, but also to demobilize it. They can spot “troublesome” workers who might create labor unrest and are often asked to report them to the firm management. My interviews show that firm managers know that line supervisors are important mediators between firm management and frontline workers and pay special attention to managing this group. An HR manager of a large-scale packaging factory in Zhongshan said the firm considered line supervisors to be the “core” personnel for stabilizing labor relations. While line supervisors can be an effective channel for communicating the firm’s stance on labor issues, if the firm loses control over line supervisors, the line supervisors themselves can become a major source of instability. This HR manager describes the firm’s challenge in hiring the right kind of people to be line supervisors: “Firms need to be careful when hiring line managers. You do not want them to be too dumb, but you also do not want them to be too smart.”16 The role of line supervisors is critical for explaining how collective action still takes place when the regime actively suppresses external mobilization. When people outside the factory cannot easily mobilize disgruntled workers and organize a protest, workers rely on pre-existing organizational structures: the organization of production. It might be inaccurate to assume that workers lack any “mobilizing structures,” which McAdam (1982) defines as “collective vehicles, informal as well as formal, through which people mobilize and engage in collective action” (p. 9). When asked about their participation in collective protests, few workers describe the decision as an individual one. Most workers refer to their unit of production as the subject of participation, such as my “line” (xian), “department” (bumen), or “factory” (chang), reflecting that the organization of production is the basis of mobilization. This observation parallels the findings in the industrial relations literature that work organization is a major factor that explains the incidence of unorganized strikes.17 When workers cannot rely on activists or organizations that operate outside production to mobilize workers, those who exercise the most influence over the organization of production have the capacity to enable, or disable, collective action. 16 17

Interview 09-2013. Zetka Jr (1992).

134

Workplace Mobilization and Collective Action

How production is organized is also important for understanding social cohesion among workers, due to their long and tiring work days. Other types of social networks, such as dormitory or hometown networks, play a role during collective action. This has been documented by many scholars,18 yet I find that the social network formed during production is still the central basis of mobilization for collective action. Since units for collective action are usually formed out of units of production, it is difficult to produce collective action without mobilization within the unit. At the same time, workers do spend most of their time working and thus “working together” could become a more substantial nexus for creating cohesion than “living together.” The mix of workers in day shifts and night shifts in many factories also means that some workers rarely see each other, even if they live in the same space. Social networks also seem to play a role when they overlap with the organization of production. One factory manager of an electronics factory in Guangzhou argued that hometown networks that carried over into production could contribute to labor conflicts: “Some workers say they want to leave because the line leader provides preferential treatment to those from the same province. For instance, a Hunan leader would only allow those from Hunan to work overtime hours. As a result, you would have lines filled with Hunan people. In that case, you should be careful. When the line leader does not like something, he can pressure the workers to create troubles.”19 In one factory that makes LED screens in Dongguan, one worker complained that the production lines are organized around regional factions such as Hunan factions and Sichuan factions, and, as a Yunnan person, she could not win line leaders’ approval.20 Although not all factories experience such extreme cases of regional factions, hometown networks are likely to carry over to production due to workers’ preferences for working with those from the same province. In the survey, about 22.2 percent of rank-and-file workers reported that they are working with a line leader from the same province. This was also strongly predicted by firm tenure, meaning that those who had worked in the firm for a long time were more likely to be in a line led by someone from the same province.21 This suggests that line leaders might be in a better position to receive senior workers’ support not only because 18 19 20 21

Smith and Pun, 2006; Pun and Smith, 2007. Interview 04-2012. Interview 03-2012. One additional year of firm tenure increased this chance by 2.4 percentage points. The result is based on logit regressions controlling for the worker’s age, gender, hukou status, income, education, years of migration, firm ownership, and firm size. The effect is statistically significant at the 5 percent level.

5.3 Understanding Heterogeneous Protest Intention

135

of their longer firm tenure but also because of other social networks that eventually overlap with production units. 5.3 UNDERSTANDING HETEROGENEOUS PROTEST INTENTION

Social cohesion becomes even more critical for explaining the dynamics of generating collective action when factories experience high levels of labor turnover. It is difficult to develop trust when the majority of rankand-file workers leave within a year of employment. New workers lack resources for mobilizing and thus cannot initiate collective action easily. This implies that structural empowerment would not necessarily translate into a higher probability of overcoming barriers to collective action, as the newly acquired structural power could weaken workers’ social networks. This section uses survey data to understand workers’ preferences for collective action and its determinants. The central questions concern how those with mobilizational resources and those without might differ in terms of protest intention and the conditions under which they would support a protest. 5.3.1 Overview: Protest Intention Although they have been dubbed the next generation of “labor activists,” the survey data demonstrate that low-level managers are much less willing to engage in a collective labor protest than rank-and-file workers. When asked if they are willing to strike at the current firm, “Imagine that some workers at your firm are talking about staging a strike for a pay increase. Are you willing to strike and demand a pay increase?” 51.2 percent of respondents answered that they were unwilling to strike, while 27.5 percent were uncertain and 21.4 percent were willing (see Table 5.2). Of the 608 respondents who gave a valid answer to this question, less than a quarter showed a preference for collective action. In the survey, we used “pay increase” as the major goal of the strike because this is the most frequent interest-based demand expressed in collective action. Looking at workers’ willingness to strike according to job position, line supervisors were far less likely to say yes: only 11.5 percent of line supervisors, compared to 31.3 percent of ordinary production workers, were willing. The data establish that, in ordinary circumstances, line supervisors are more reluctant than rank-and-file workers to engage in collective action. Likewise, workers with a longer firm tenure are on average less willing to strike than new workers. Willingness drops by 2.1 percentage points for each additional year at the firm, after controlling for demographic

136

Workplace Mobilization and Collective Action TABLE

5.2 Are you willing to protest for a pay increase? All

Supervisor

Rank-and-file

Willing Not sure Unwilling

21.4% 27.5% 51.2%

11.5% 26.0% 62.5%

31.3% 29.0% 39.8%

Total N

608

304

304

variables. This suggests that those with mobilizational resources are more reluctant to protest, which might explain why labor protests occur less frequently than the individual survey data would suggest. While more than 20 percent of workers support the idea of protests, survey data from various years show that only 5–10 percent of workers in the PRD area actually experience one.22 In the survey dataset, only 12 out of 155 firms (7.7%) had workers reporting an incidence of collective action during the previous year, revealing the substantial gap between protest intention and protest action. 5.3.2 Studying Collective Action with Survey Data This suggests that an average worker’s preferences for collective action might not inform us about the firm-level incidence and patterns of collective action. For this reason, if we use survey data to gain insights about which firms are more prone to experience collective action, but treat every individual equally, it might not be effective. It is more critical to understand the distribution of protest intention across groups with different resources for workplace mobilization. The protest intention and perceptions of those who have mobilizational resources – potential protest initiators – would tell us much more about whether collective action might occur at a given firm. The survey dataset, by oversampling line supervisors, offers a unique insight into the protest intention of those who can lead collective action, and those who might follow. At an individual level, studies find that attitudes toward collective action and other related perceptions have substantial predictive power for actual protest behavior. The framework of “reasoned action” suggests that social actions do not necessarily result from spontaneous reactions to the situation but often 22

Interview 02-2013. A research team at Sun Yat-sen University had fielded a yearly survey with migrant workers in the Guangzhou train station until 2014. According to the research team, roughly 5–10 percent of workers report an occurrence of a labor protest at their firm every year. In the 2011 dataset, 10.1 percent of respondents (113 out of 1118) indicated that their firm had an incidence of collective action in the previous year.

5.3 Understanding Heterogeneous Protest Intention

137

reflect one’s underlying attitudes and beliefs that are more stable.23 In the case of protests, studies such as Finkel and Muller (1998) likewise find that individual perceptions concerning collective action ex ante – such as dissatisfaction with the status quo, perceived group efficacy, and perceived personal importance – are important predictors of their protest behavior in a later period.24 Survey data still need to be interpreted with caution because conditions can change, and workers’ perception of collective action could shift before and during collective action. Yet, the literature shows – and I show in this and subsequent chapters – there could be substantial consistency in individual perceptions and preferences across different time periods, which affect actual behavior.25 When the underlying perceptions and beliefs are structurally determined, survey data could provide useful insights about protest behavior. In the context of group-level incidence, one has to be cautious about treating individual responses equally because there are group-level dynamics at play. However, the individual perceptions and protest intention of those with mobilizational resources can be highly informative about the baseline distribution of protest “likelihood” across firms. 5.3.3 Determinants of Preferences for Collective Action Are there structural reasons why line supervisors and those with a longer tenure at the firm – who could function as social glue in the workplace – are more reluctant to participate in collective action? In this section, I analyze determinants of willingness to protest, and how they might differ among heterogeneous workers. I focus on the conditions under which workers with mobilizational resources might be willing to initiate a protest. Based on the existing scholarship, I identify five determinants of willingness to protest26 : (1) workplace grievances, (2) perception of alternative job opportunities, (3) perceived chance of success of collective 23 24 25

26

Ajzen and Fishbein, 1980; Ajzen, 1991. Finkel and Muller, 1998; Kelloway et al., 2007. This is not to suggest that workers do not react to events that might change their intention and perceptions. Most protests are responses to “trigger” events that influence the workers’ interests in collective action. However, workers also respond differently to these trigger events. Studies on protest behavior have repeatedly emphasized the importance of two core factors: individual dissatisfaction with the status quo and the sense of personal and group efficacy. While scholars have given these factors different names, a large volume of both theoretical and empirical research highlights their importance. Klandermans (1992), for instance, argues that the sense of injustice coupled with the instrumentality of participation – the perception that collective action is likely to lessen the injustice – are critical.

Workplace Mobilization and Collective Action

138 TABLE

5.3 Sources of preferences for collective action (% agree)

Factors

Rank-and-file

Line leaders

Job-specific grievances I deserve a pay raise I will not get promoted in three years

84.1 23.5

74.6 41.5

Turnover intention I can get a better job than the current one I plan to leave the firm within one year

76.1 24.4

71.7 3.6

Perceived chance of success All of the workers will participate The firm will give a pay raise

53.4 66.6

43.0 52.1

Perceived costs of participation Some workers will be arrested Some workers will be fired

62.4 80.7

63.4 76.6

49.5

52.8

6.2

10.6

Personal efficacy My nonparticipation will greatly diminish workers’ strength Past participation I have participated in a labor protest Source: Author’s survey.

action, (4) perceived costs of participation, and (5) personal efficacy. I describe these factors and compare the differences in how they manifest between line supervisors and rank-and-file workers (Table 5.3). 5.3.3.1 Workplace Grievances While grievances alone do not translate directly into collective action, it is still necessary to understand where grievances come from to understand the nature and consequences of workers’ collective action. Collective action is most commonly focused on wage-related goals. Other issues, such as social insurance, are sometimes the target of collective demands, but the majority of labor disputes still concern wages.27 According to the Shenzhen Bureau of Human Resources and Social Security, wage-related issues made up 46.80 percent of the total number of letters and visits accepted by the Bureau in 2010, followed by labor contracts (30.69%)

27

Opp (1998) highlights that individual discontent with the current distribution of public goods and their personal efficacy were strongly related to political protests in East Germany. Cai and Meiyan, 2012, p. 19.

5.3 Understanding Heterogeneous Protest Intention

139

and social insurance (6.31%). Wages also seem to be the main factor that workers consider in choosing a job: In the 2010 survey carried out by the Shenzhen Federation of Trade Unions, more than 80 percent of migrant workers indicated that it is wages that they care most about when selecting a job. Line leaders and senior workers have higher wages, so they could have fewer incentives to protest because they are relatively satisfied. Due to the higher job status and seniority, these workers might enjoy a privileged position in the firm and be therefore unwilling to incur the costs of staging collective action. The survey data suggest that line leaders receive a substantially higher salary than general production workers. On average, line leaders’ monthly salary was 1.64 times that of ordinary workers employed in the same firm, and the multiple ranged from 1.04 to 3.03. Senior workers on average also received more than new workers – an additional 60 yuan of the monthly salary per year of tenure – although the difference is not large. Not surprisingly, the survey data suggest that regardless of role and firm tenure, wage grievances are widespread. When asked, “Do you think your firm should give you a pay rise?” about 84.1 percent of ordinary production workers and 74.6 percent of line supervisors said yes. When asked about the ideal monthly wage, the respondents reported a substantial gap between the current wage and the wage that they thought they deserved: General production workers sought a 21.1 percent increase, while line supervisors wanted a 14.5 percent increase. Line supervisors were less likely to want a pay rise and also reported a substantially smaller gap between the actual and ideal wage. Yet, the data still show that wage-related economic grievances might be too widely held to explain protest behavior. Although wage demands are often at the forefront of workers’ demands during collective action, it might not mean that wages are the grievance that is driving collective action. Studies have shown that the stated goals of collective action do not always reflect the real motives.28 In discussing the labor strikes in France in the nineteenth and twentieth centuries, Shorter and Tilly (1974) argue that although workers’ collective demands concentrated on wage issues, “[m]ore important than the content of a grievance was the fact that it had been advanced at all,” because wage demands often helped the workers bring together broader and more diffuse sources of discontent (p. 68). In their interviews with me, workers 28

Gouldner, 1954; Golden, 1996.

140

Workplace Mobilization and Collective Action

often did not conceive of their wage demands in narrow terms of monthly pay, but in a broader context of chances of promotion, upward mobility, and future economic opportunities. A worker’s statement that their wages are too low might need to be interpreted in the context of how workers define “good” wages. When talking about wage grievances, workers often mention the lack of upward mobility within the firm. As discussed in Chapter 4, upward mobility for production workers is limited. Their wages either do not increase or increase only marginally even if they stay with the firm for many years. There is a very small reward for experienced workers unless they get promoted, and the chances of promotion are low. During the Honda strike, one of the protest leaders stated, “we do not only want a one-time increase of wages. What is important is to establish a system to protect workers’ interests in the long run,” pointing to the unfairness of the existing promotion system: It was taking fifteen years to get promoted to the next major level.29 At the Yue Yuen shoe factory, workers also received only small compensation for their tenure. After the first year of employment, the seniority bonus only increased by 10 yuan, roughly 1.4 dollars per year.30 One worker who had been at an electronics factory for five years lamented, “How come my salary and a new worker’s salary are almost the same when I have worked here for five years?”31 The survey data also confirm that rank-and-file workers do not get promoted easily and perceive the chances of promotion to be low. Out of 310 general production workers, only 7.1 percent of workers reported that they had been promoted during the previous year. When I restrict the sample to workers who had been at the firm more than one year (n = 155), the percentage increases only marginally to 7.7. When asked if they expect to be promoted if they work in the current firm for an additional three years, slightly more than half of workers (56.6%) said yes. As few rank-and-file workers stay at the same firm for three years, this suggests that most workers have a low expectation of getting promoted. The expectation is higher for line supervisors: About 75.4 percent of line supervisors expected a promotion in the current firm. This difference also contributes to line supervisors’ weaker inclination to protest. Line leaders who have already defied the odds and been promoted have more to lose by participating in collective action. They 29 30 31

“Honda Strike’s Deep Secrets: Revealing Pay Slips’ Blood and Tears (bentian bagongmen shendu jiemi: baoguang yuangong gongzibiao ningju xuelei)” 2010. “An in-depth discussion with a veteran worker at Yue Yuen” 2014. Interview 04-2012.

5.3 Understanding Heterogeneous Protest Intention

141

TABLE 5.4 Demographic characteristics of workers by job position

Age Male High school Rural Income Firm tenure Migration years

Supervisors

Rank-and-file

31.9 80% 85% 68% 4715 4.26 9.39

26.3 53% 52% 88% 2887 1.75 5.52

Source: Author’s survey.

fear being demoted or dismissed, and losing their hard-won position.32 The training for line leaders is generally firm-specific and not transferable between employers, and line leaders who lose their job often have to start as an ordinary production worker in a different firm. This raises the perceived costs of protest for line leaders, who enjoy a privileged but precarious status in the factory. 5.3.3.2 Turnover Intention and Exit Behavior In a tight labor market, one might assume that it would be easy to find a new job. Migrant workers in China are prone to job hopping, so their reluctance to engage in protest could originate from their turnover intention. Using Hirschman’s terms, workers do not necessarily need to raise their “voice” when they can choose to “exit.”33 As a firm manager at a small ceramics factory in Foshan puts it, many firms “are less concerned about striking workers, but more about having no workers at all!”34 Line leaders, as shown in Table 5.4, have characteristics that are much more valued in the labor market than rank-and-file workers: They are more likely to be male, educated, and hold an urban hukou. This might explain why line supervisors are less enthusiastic about collective action, as they can leave the firm for another job opportunity. Although the manufacturing sector in China has high levels of labor turnover in general, not all workers are prone to job hopping. In the survey, line supervisors were less likely to report that they have a better 32 33 34

Globalization Monitor, 2009. Hirschman, 1970. Interview 03-2012.

Workplace Mobilization and Collective Action

142

Age

Male

Education

Rural

Income

Supervisor

Tenure

Migration years −0.2

0

0.2

0.4

FIGURE 5.1 Predicted effect on perception of better outside option Note: Average marginal effects are based on a logit regression analysis with firmlevel random effects. Robust standard errors clustered at the firm level.

job option (see Figure 5.1). While a majority of workers – 73.9 percent – indicated that they could find alternative employment easily, line supervisors and senior workers – after controlling for income and education – were less likely to perceive such an opportunity. Being a line supervisor and having an additional year of firm tenure reduced the perception by 11 and 2.1 percentage points, respectively. Both groups of workers also displayed a much weaker intention to exit, due to their firm-specific embeddedness. When asked how long they intended to stay at the current firm, 24.5 percent of rank-and-file workers indicated that they would leave within a year, while only 3.4 percent of supervisors said they would. Controlling for job position, firm tenure also substantially reduced workers’ intention to leave a job: One additional year of work at a firm decreased the turnover intention by 8.6 percentage points. While turnover intention does not equate to turnover behavior, it can be a strong predictor of employment attrition.35

35

Steel and Ovalle, 1984.

5.3 Understanding Heterogeneous Protest Intention

143

The long tenure of line supervisors and senior workers could reflect their relative satisfaction with the current working and living environment. When a worker activist asked a protest participant at the Yue Yuen factory what made workers stay despite the low wages, the worker pointed to the seniority of the workforce: “Many workers are middle aged, and they started working when they were in their teens and twenties. They find it difficult to leave the place.”36 During interviews, senior workers often said that they are now so accustomed to the people, location, and work at the current place that their decision to exit is not just a matter of economic compensation. This means that the same qualities that enable firm-based mobilization, such as a strong social network, also reduce workers’ propensity to stage a protest. The discussion so far highlights that in the current environment, where job hopping is widespread and the regime undermines external mobilization, those who have the potential to mobilize others for atomized protests are not those who are most aggrieved about their current situation. Workers accumulate the resources that would make them effective at mobilization when they build social capital in the workplace, but since dissatisfied workers pursue upward mobility by job-hopping,37 they do not establish a strong social network within the firm. As demonstrated in Chapter 4, the experience of job hopping – with little upward mobility – induces social and political grievances that are critical for politicizing labor discontent. This means that due to structural constraints, those with high levels of social grievances cannot easily tap into the tolerated space for contention, even if they want to. Those who stay at one firm for a long time and build up the social network, such as line supervisors, are relatively more satisfied with their job and have higher stakes at the current firm, which lowers their protest intention. The mismatch between mobilizational resources and grievances, therefore, not only lowers the chances of collective action but also means that those who are able to initiate a labor protest are not necessarily those most likely to form anti-regime identities. 5.3.3.3 Perceived Chance of Success The conventional view of migrant workers’ collective action has emphasized its spontaneity, which highlights workers’ emotional reaction to exploitation.38 Collective action has been described as workers’ last 36 37 38

“An in-depth discussion with a veteran worker at Yue Yuen” 2014. Daming and Xiaoyun, 2014. Chan and Senser, 1997.

144

Workplace Mobilization and Collective Action

resort after exhausting all alternatives to address their grievances and express anger.39 This argument reflects the social movement theory on “expressive benefits,” which postulates that some protest participants get psychological benefits from participation itself.40 For these actors, “its importance [...] is in expressing support for the cause, regardless of whether it produces the desired visible consequences.”41 Alternatively, workers could use collective action to “let off steam” or take a break from work. Chan and Hui (2012)’s case study of a labor protest in Shenzhen illustrates that “curiosity, prolonged discontent and a sense of unfairness seemed to be more important factors than the wage demand” (p. 93). On the other hand, studies of popular protests in China have emphasized citizens’ strategic reasoning. Contenders weigh the costs and benefits of their contention strategies and decide based on the probability of success.42 Citizens’ reliance on troublemaking tactics has been attributed to their expectation that disruptive tactics would yield better results.43 Contenders are ready to take any measures that enhance their chances of success,44 but this also implies that potential contenders are discouraged from initiating actions if they don’t think they will succeed. Cai (2006) shows that laid-off SOE workers were more likely to protest when they believed in the effectiveness of the group action. This argument reinforces the conclusions in the existing studies that the decision to protest substantially hinges on the perception of group efficacy – the expectation that a groups’ demands will be met through their collective action.45 Migrant workers might vary in their perceptions of the likely efficacy of group action. For instance, in a 2011 survey of migrant workers, the most common reason given for not participating in collective action was that “It does not work” (meiyouyong).46 Yet, at strikes it is common to see workers who are not particularly concerned about the group demands. Some workers do not know what they are fighting for, while others enjoy being part of collective action, regardless of whether the firm raises their wages or not. A worker at a protest in Foshan said, “I actually 39 40 41 42 43 44 45 46

Pun and Huilin, 2010. Chong, 1987. Turner, 1981, p. 11. O’Brien and Li, 2006; Cai, 2006; Chen, 2009; Chen and Xu, 2012. Chen and Xu, 2012. Chen (2009) calls this tendency protest opportunism. Protesters are ready to employ any combination of tactics that they deem useful. Klandermans, 1984; Finkel et al., 1989. The survey was conducted in the Guangzhou train station, during the Spring Festival. Author’s own calculations based on the survey data.

5.3 Understanding Heterogeneous Protest Intention

145

do not benefit much even if the manager agrees to pay the compensation. I still like to be a part of it, because I feel it is actually fun.”47 She added that there were few costs involved: “Why not participate? The production line is shut down, so you cannot work anyway. What should I be afraid of? They cannot fire me because I am not a leader (daitouzhe).” Thus, we could think about the success of collective action in two different ways. One is the success of mobilization, although this does not guarantee that protesters’ demands will be met. As Opp (2009) suggests, the major difference between social movement and protest is that a protest cannot directly produce the collective goods. Whether protesters’ demands will be met depends on the target of the protest, which is the firm in the case of labor protests. When workers are successfully mobilized to go on strike, it does increase the chance that the management will meet their demands, but these two elements are distinct. Those who enjoy the “express benefits” could be satisfied with successful mobilization (production of collective action) itself, even if it does not induce their target to meet their demands (production of collective goods). For workers for whom the costs of initiating collective action are higher, however, there might be more concern about the likelihood of actually getting their demands met. They might only be willing to protest when they believe that their target will respond favorably. In the survey, workers’ estimates of the chances of success were measured using two follow-up questions regarding protest intention. “If a labor protest did take place at your current firm,” the question asked, “how probable is it the following situation would occur?” In response to, “Workers will receive a pay raise,” which measures the perceived likelihood of producing the collective good, about 61 percent of workers replied that it was either highly or somewhat likely. In response to the statement, “All of the workers will participate in the strike,” which measures workers’ perceived likelihood of successful mobilization, about 48.2 percent of workers indicated that it was likely. For both statements, there was no substantial difference between rank-and-file workers and line supervisors when controlling for demographic variables; however, longer firm tenure was associated with a substantially lower expectation of success. An additional year of firm tenure, after controlling for demographic characteristics, reduced the percentage of workers who indicated that the firm would give a pay raise by 2.5 percentage points, and that

47

Interview 01-2013.

146

Workplace Mobilization and Collective Action

of workers who indicated that all of the workers would participate by 3.9 percentage points. 5.3.3.4 Costs of Participation The downside risks to collective action might also be an important determinant of protest participation. There are two major risks for workers that affect whether they initiate or participate in collective action. First, workers might face retribution for taking action, including being dismissed. Second, protest participants could be detained or arrested by police. Studies of collective action in authoritarian regimes treat police repression as the most significant deterrent. Both risks could be higher for protest leaders, because the firm and police use selective repression. Firms tend to fire only protest leaders and active participants.48 The rate and intensity of repression has decreased over the last two decades, but potential protest initiators might still hesitate, out of fear of selective retribution and repression. To analyze the effect of these downside risks on workers’ willingness to strike, I tested two follow-up statements: “Some strike participants will get fired” and “Some strike participants will get arrested by the local police.” A majority of workers anticipated firm-level retribution: About 88.6 percent of workers said that they expected some protesters to be fired. A substantial number of workers, about 62.8 percent, anticipated police repression. The data show that workers think that firm-level retribution is more likely than repression by the local police. We can see that, while atomized incorporation introduced a political environment in which the general level of police repression is low, a substantial percentage of workers still expect selective repression. For these two statements, there was no substantial difference in responses between line supervisors and general production workers. 5.3.3.5 Personal Efficacy Those who stress free-riding as a critical barrier to collective action emphasize personal efficacy – the perception that one’s individual contribution substantially increases the chance of success. An individual who thinks that their contribution is negligible is more likely to free-ride. Empirical studies of collective action find that personal efficacy is an 48

Firms do not always fire the most active participants. For instance, firms seem to be less willing to fire line supervisors when they are integral to production. Instead, those firms choose to fire other participants for sending a signal to workers, and co-opt line supervisors. Nonetheless, workers generally shared the understanding that if you lead a protest or actively support it, it increases your chance of being singled out and dismissed.

5.3 Understanding Heterogeneous Protest Intention

147

important determinant of participation.49 In the context of labor protests and unorganized workers, I argued that free-riding is not a serious issue until after a substantial number of people have joined the protest. If workers believe that firms will accede to protesters’ demands and distribute benefits across the workforce, then some will not join a protest once it reaches a threshold number of participants. This free-riding would be an issue if the firm will only negotiate when the whole workforce strikes. Most firms will react before the entire workforce joins the protest, but individuals’ perceptions of their individual leverage within the movement will still affect their sense of whether protesting has value to them personally. To gauge perceptions of personal efficacy, I used this follow-up statement: “If I do not participate, it will greatly decrease the workers’ strength.” About 51.1 percent of respondents agreed with the statement, and there was no significant difference between line supervisors and rank-and-file workers. Firm tenure, however, had a negative effect: an additional year of firm tenure decreased one’s sense of personal efficacy by 2.6 percentage points. 5.3.4 Comparison of the Two Groups To what extent do line supervisors and rank-and-file workers differ in the hypothesized determinants of protest intention? In Table 5.5, I present the average marginal effects of the five groups of determinants on individual protest intention based on logit regressions.50 The dependent variable is a binary measurement, which takes the value 1 if the respondent is willing to protest and 0 if they are unsure or unwilling. The analyses generally confirm the earlier discussion in the following ways. First, even when controlling for other basic demographic variables, line supervisors are much less willing to strike than general production workers (see the first column). Second, wage grievances do not have strong predictive power for both groups of workers, supporting the theory that wage grievances might be too prevalent to be useful for identifying who is likely to protest. Workplace grievances in general – grievances about the current wage level (desire for pay raise) and lack of future promotion (no promotion) – do not seem to have a substantial effect on line supervisors’ propensity to participate in a labor protest. For rank-and-file workers, however, future career opportunities in the 49 50

Finkel et al., 1989; Muller et al., 1991. The correlation table for the independent variables is presented in the Appendix.

Workplace Mobilization and Collective Action

148 TABLE

5.5 Determinants of protest intention

Age Male Education Rural hukou Wage (log) Supervisor Firm tenure Migration years

All workers

All workers

Line supervisor

Rank-and-file

−0.001 (0.003) 0.152∗∗∗ (0.037) −0.034∗∗ (0.015) −0.000 (0.047) 0.001 (0.099) −0.177∗∗∗ (0.062) −0.017 (0.011) 0.004 (0.005)

−0.000 (0.004) 0.099∗∗∗ (0.033) −0.022∗ (0.012) 0.034 (0.043) 0.062 (0.075) −0.172∗∗∗ (0.045) −0.004 (0.012) 0.001 (0.005) 0.097∗∗∗ (0.037) 0.022 (0.041) 0.112∗∗∗ (0.031) 0.116∗∗∗ (0.040) 0.204∗∗∗ (0.032) 0.019 (0.033) −0.059 (0.039) 0.116∗∗∗ (0.029)

0.001 (0.004) 0.062 (0.055) −0.017 (0.015) 0.005 (0.037) 0.026 (0.058)

0.002 (0.005) 0.145∗∗∗ (0.047) −0.022 (0.018) 0.086 (0.090) 0.122 (0.157)

−0.020∗∗ (0.010) 0.006 (0.005) 0.057 (0.045) 0.017 (0.043) 0.043 (0.033) 0.131∗∗ (0.054) 0.111∗∗∗ (0.043) 0.021 (0.037) −0.103∗∗ (0.041) 0.059 (0.036)

0.016 (0.015) −0.014 (0.009) 0.143∗∗ (0.056) 0.044 (0.059) 0.213∗∗∗ (0.050) 0.122∗ (0.065) 0.297∗∗∗ (0.054) 0.002 (0.051) −0.005 (0.060) 0.176∗∗∗ (0.048)

595 155

592 155

296 155

296 155

Job availability Desire for pay raise No promotion Group efficacy Mobilization Police Fired Self-efficacy Observations Number of firm

Note: Average marginal effects are based on logit regressions with firm-level random effects. ∗∗∗ p < 0.001, ∗∗ p < 0.05, ∗ p < 0.10. Robust standard errors in the parentheses.

firm matter. When rank-and-file workers have zero expectations of future promotion, their inclination to protest increased by 21.3 percentage points. The finding echoes the earlier discussion about the implications of wage grievances during labor contention. Since wage grievances are so

5.3 Understanding Heterogeneous Protest Intention

149

prevalent, they are not sufficient to predict who is more willing to protest for a pay increase. Instead, for rank-and-file workers, lack of upward mobility might be a more important source of workplace grievances. Third, data also show that the tight labor market does not mean increased protest intention for line supervisors. Perceived job availability increases rank-and-file workers’ protest intention by 14.3 percentage points but does not affect line supervisors. The firm-specific embeddedness of line supervisors means that a job change might be more costly for them, and the option of a better job might not translate into exit intention. Rank-and-file workers, on the other hand, find it easier to move since most firms do not train them with firm-specific skills. This means that growing labor shortages and turnover could increase most migrant workers’ overall propensity to resist but would not proportionally trigger more collective action, because the gate-keepers of collective action – who can mobilize and de-mobilize workers – are less strongly affected by the state of the labor market. This might explain why the incidence of collective action – despite the growing bargaining power of migrant workers in the labor market – is still low. Finally, line supervisors appear to be much more goal-oriented than rank-and-file workers. The data show that for both groups, the perceived chance of success has a substantial effect. The perception that collective action would lead to a pay increase (group efficacy) increased the protest intention of line supervisors and rank-and-file workers by 13.1 and 12.2 percentage points, respectively. The perception of successful mobilization (mobilization) also contributed to higher protest intention, and its effect was substantially larger for rank-and-file workers (29.7) than line supervisors (11.1). This finding suggests that rank-and-file workers might enjoy expressive benefits more than line supervisors: Even if they believe that they might not be able to achieve their goal, the expectation of producing collective action itself could induce their participation. This finding echoes the earlier discussion that those who pay the initial costs of mobilization might be more concerned about whether they can satisfy their demands than followers. Tabulation of group efficacy and protest intention supports this point. Among the line supervisors who indicated that the firm would not give a pay raise, only 2.1 percent said that they would protest. The number goes up to 13.6 percent for rank-and-file workers. The costs of participation – police repression (police) and employer retribution (fired) – seem to be only weakly associated with the protest intention of rank-and-file workers; for line supervisors, however, the

150

Workplace Mobilization and Collective Action

expectation that some participants would be fired decreased their protest willingness by 10.3 percentage points. The nonsignificant effect of police repression is expected, given the regime’s tolerance toward workplacebased contention. Workers also rely on diverse strategies to minimize the chance of repression. For instance, protest participants often refuse to identify strike leaders out of fear of targeted repression by the local government. Yet in acts of workplace-based contention such as labor strikes, that do not involve violent or highly disruptive tactics, even if local police appear on the protest scene they rarely use violence, as I show in Chapter 6. The data suggest that potential strike initiators are much more concerned about firm retribution due to their firm-specific embeddedness. Finally, perception of self-efficacy increased protest intention of rank-and-file workers by 17.6 percentage points, while having a positive but statistically nonsignificant effect on line supervisors. While desire for economic gains did not lead to higher willingness to protest among line supervisors, it does not mean that line supervisors are never driven by economic incentives. It might be that line supervisors are not driven by a small wage gain but do respond to larger economic incentives. What the survey data suggest is that for structural reasons, line supervisors have substantially less interest in collective action and they need to perceive a bigger opportunity than rank-and-file workers do, to be willing to protest. Qualitative evidence shows that line supervisors and senior workers sometimes perceive a chance for a substantial financial gain and are motivated to support collective action. A worker activist writes:51 Why would low-level managers initiate strikes? [. . . ] Line supervisors usually enter the firm relatively early; in regards to benefits that are tied to firm tenure, they will objectively benefit more. When workers’ wage decreases (for example, if the overtime standard rates are reduced), its negative effects on line supervisors can be more apparent.

As this activist explains, firm benefits such as year-end bonuses, social insurance, and lay-off compensation are tied to the current wage level and firm tenure. If the firm policies in regards to these issues change in a way that negatively affects line supervisors and those with a longer firm tenure, it generates substantial financial incentives for them to stage a collective protest. For instance, workers’ year-end bonuses might be determined based on workers’ average monthly wage. If the firm announces that they will reduce year-end bonuses from three times the monthly wage 51

Hao, 2016, p. 61.

5.3 Understanding Heterogeneous Protest Intention

151

to one month’s wage, then those who earn more experience a larger loss, and the loss is not small. Similarly, when firms end their labor contract with workers, firms need to provide compensation equivalent to workers’ average monthly wage multiplied by workers’ firm tenure. Those who stayed in the firm for many years will receive a larger amount of compensation than other workers. As a result, line supervisors and experienced workers could be motivated to initiate collective action when a firm changes its company name, merges with another firm or relocates to a different location, demanding that the firm first end the labor contract with the workers and offer employment compensation – or “buy off” workers’ firm tenure (maiduan gongling). By ending the labor contract before the contract expires, workers with many years of firm tenure can receive a large sum of compensation. In these cases, line supervisors and senior workers are less concerned about getting fired, because they want the termination of the employment relationship. 5.3.5 Effect of Blame Attribution on Protest Intention What are the effects of blame attribution on protest intention? If the previous analyses are right that rank-and-file workers reflect a more diffuse set of grievances in their protest intention – while line supervisors are much more concerned about firm-specific treatment – then we would expect that workers’ attribution of blame to the central government would affect line supervisors less strongly than rank-and-file workers. Figure 5.2 displays a simple tabulation of protest intention, blame attribution, and job position. The tabulation highlights a few interesting trends. First, for line supervisors, the largest influence over their willingness to protest seems to be their attribution of blame to firms; the mean level of protest intention is substantially higher for those who blame the firm compared to those who do not. Their attribution of blame to the central government, local government, and workers themselves seems to lower protest intention, although marginally. For rank-and-file workers, all types of external attribution appear to increase protest intention. At the same time, unlike line supervisors, their attribution of blame to firms does not appear to be a stronger factor than attribution of blame to the central or local government. In fact, among those rank-and-file workers who do not blame the firm, more than 20 percent still indicated their willingness to protest at the firm. In other words, while the protest intention of line supervisors is mostly driven by their perception that their firm is

Workplace Mobilization and Collective Action mean (protest intention) .1 .2 .3 .4 .5 blame−cent=0

0

0

line supervisors

blame−local = 0 rank−and−file

blame−local =1 line supervisors

mean (protest intention) .1 .2 .3 .4 .5

mean (protest intention) .1 .2 .3 .4 .5

rank−and−file

blame−cent =1

0

0

0

blame−firm=0 rank−and−file

FIGURE

blame−firm=1 line supervisors

0

0

mean (protest intention) .1 .2 .3 .4 .5

152

0

blame−self = 0 rank−and−file

blame−self=1 line supervisors

5.2 Protest intention by blame attribution and job position

to blame, that of rank-and-file workers could be driven by attribution of blame to government. Next, I ran logistic regressions to see if these patterns hold after controlling for demographic variables and job-specific grievances.52 As displayed in Table 5.6, when we look at both line supervisors and general production workers, the only blame attribution variable that predicts protest intention is attribution of blame to firms: blame-firm increases protest intention by 11.8 percentage points. Yet, as expected by the tabulation of the raw data, there is a substantial interaction effect between blame attribution and job position. Model 2 shows that when rankand-file workers attribute blame to the central government, it increases their willingness to protest by 8.5 percentage points; for line supervisors, the effect switches the sign and decreases the protest intention by about 5.6 percentage points.53 The effect of blame-firm is positive and significant for both groups, but its size is substantially greater for line supervisors than rank-and-file workers. This also shows that the differences between the two groups in their protest intention magnify among 52 53

The findings remain similar when controlling for the other determinants of protest intention, as introduced in the previous section. Due to the relatively high level of correlation between blame-cent and blame-local, which is about 0.52, taking blame-cent out of the models makes the effect of blame-cent larger and more significant.

5.3 Understanding Heterogeneous Protest Intention TABLE

5.6 Effect of blame attribution on protest intention

Age Male Education Rural hukou Wage (log) Supervisor Firm tenure Migration years Job availability Desire for pay raise No promotion Blame-cent Blame-local Blame-firm Blame-self Blame-cent ∗ supervisor

Model 1

Model 2

0.001 (0.004) 0.157∗∗∗ (0.037) −0.029∗∗ (0.014) −0.001 (0.047) −0.007 (0.101) −0.144∗∗ (0.060) −0.042 (0.010) 0.000 (0.005) 0.106∗∗ (0.043) 0.056 (0.047) 0.064∗ (0.036) 0.035 (0.044) −0.035 (0.043) 0.118∗∗∗ (0.039) −0.046 (0.039)

0.002 (0.004) 0.143∗∗∗ (0.037) −0.030∗∗ (0.014) 0.003 (0.048) 0.004 (0.105) −0.170 (0.112) −0.012 (0.010) 0.001 (0.005) 0.111∗∗ (0.043) 0.066 (0.048) 0.074∗∗ (0.036) 0.085 (0.053) −0.029 (0.051) 0.068 (0.050) −0.043 (0.048) −0.141∗ (0.084) −0.032 (0.083) 0.161* (0.095) −0.005 (0.088)

593

593

Blame-local ∗ supervisor Blame-firm ∗ supervisor Blame-self

∗

supervisor

Observations

Note: Average marginal effects are based on logit regressions with firm-level random effects. ∗∗∗ p < 0.001, ∗∗ p < 0.05, ∗ p < 0.10. Robust standard errors in the parentheses.

153

154

Workplace Mobilization and Collective Action

those who blame the central government, and almost disappear among those who blame the firm. The interaction effects for the other two variables – blame-local and blame-self – are negligible. The analyses seem to reinforce the earlier findings, that line supervisors are much more goal-driven and strategic about their protest behavior than rank-and-file workers. Since atomized protests cannot resolve politicized grievances, line supervisors are not motivated to protest when they blame the central government. Rank-and-file workers, on the other hand, are willing to protest even when they do not blame the firm necessarily and rather blame the other actors such as the central government. 5.3.5.1 Workers’ Strategic Agency and Its Political Implications These findings point to structural factors that connect workers’ protest intention, mobilizational capacity, sources of grievances, and strategic agency. Line supervisors’ social embeddedness in the firm endows them with mobilizational resources but also reduces their exit intention because they have a higher investment in the firm. This makes them less prone to participate in collective action and more strategic about their decision to protest. Rank-and-file workers, on the other hand, have not experienced upward mobility in the firm, and pursue economic opportunities via horizontal mobility. This makes them more likely to generalize their grievances but also limits their access to the social capital required to mobilize a workplace. The “core” group of workers – those with mobilizational resources – have reasons to be more goal-oriented. The threshold requirement for mobilizational resources means that collective action by migrant workers is not always a product of legal rights consciousness, or a spontaneous reaction to their marginalized conditions. Workers can also be strategic about gauging the likelihood of success. While the perceived likelihood of success was an important factor for both groups of workers, as individual groups; for line supervisors, the perceived group efficacy was the most important factor, and for rank-and-file workers, the perceived likelihood of successful mobilization was the most important factor. This means that free-riding is not an important impediment to collective action among migrant workers. Free-riders would be discouraged from participating in a protest when they believe that the collective protest will succeed. However, according to both survey data and qualitative evidence, a majority of workers express higher willingness to join a protest if they believe that it is going to succeed.

5.3 Understanding Heterogeneous Protest Intention TABLE

155

5.7 Why did you not join the collective action?

Statement

Agreement

The case was not related to me. Many people already participated. My nonparticipation would not matter. Collective action does not work and would not solve the problem. Even if I do not join, I can still enjoy the results. I was afraid that the firm would harass me.

63.1% 38.5%

26.2% 36.9%

Observations

65

60.0%

Note: The data are from the 2011 Guangzhou Train Station Survey collected by Sun Yat-sen University.

Evidence from the 2011 train station survey data confirms these observations. Workers were asked whether their firm had experienced collective action during the previous year and why they had or had not participated. About 10 percent of workers – 113 out of 1118 respondents – answered that they had experienced an incidence of collective action; out of the 109 workers who gave a valid answer, 69 workers indicated that they had not participated themselves. The survey asked why they had not joined by asking respondents to rate how much they agreed with five statements. The free-rider’s logic for nonparticipation – even if I do not participate, I can still enjoy the results – had the most “strongly disagree” responses. As displayed in Table 5.7, the two most agreed to reasons for nonparticipation were that the protest was not related to one’s own interests and that the worker did not believe in the success of collective action. While workers’ real rationale for their nonparticipation might differ from their answers to the survey question, the data at least suggest that workers do not want to be perceived as free-riders. Instead, many nonparticipants questioned the relevance of the protest to them personally, and the effectiveness of protest. The strategic agency of workers, and that of line supervisors, in particular, suggests that the distinction between economic and political issues under atomized incorporation could lower the chance of attracting politically aggrieved workers to the open space of contention; while workers’ grievances about the central government are relatively common, only a small fraction of them have mobilizational resources and many are not interested in atomized protests in the first place. Even when these workers blame the central government, they become less willing to protest in the

156

Workplace Mobilization and Collective Action

workplace, perhaps because they perceive that atomized protests cannot resolve broader grievances, beyond the boundaries of their individual workplaces. These findings, however, do not suggest that atomized incorporation is ineffective in undermining the potential revolutionary threat from workers. On the contrary, the analyses show that atomized incorporation has been successful at materially rewarding those with mobilizational resources. A substantial number of line supervisors do blame the central government, even if the percentage is lower than rank-and-file workers. Among the line supervisors who blame the central government, about 9.8 percent are still willing to protest at the workplace (see Figure 5.2). These individuals – who are able to mobilize their workplaces, aggrieved about the central government, and still willing to engage in atomized protests – could pose a significant threat to the regime. By letting them express their demands and receive rewards, atomized incorporation could be helpful in demobilizing the labor movement and undermining the more immediate threat. Yet, this does not mean that atomized incorporation has been able to collect information on those who have social grievances beyond the workplace – who could potentially be a bigger threat to the regime once mobilized – because it is structurally designed to reflect economic grievances about specific job conditions. 5.3.6 Past Protest Behavior and Protest Intention How do workers’ previous experiences of collective action affect their preferences for collective action? The extant literature suggests that participating in a protest could influence attitudes and perceptions that encourage future participation in protests. Empirical studies based on survey data find that protest participation increases one’s perception of both personal and group efficacy, and sometimes induces militancy.54 The mechanism through which protest behavior engenders such effects has been discussed in the context of social movements. Snow and Oliver (1995), for instance, argue that social movement activists could change protest participants’ beliefs and attitudes by “identifying a problematic condition and defining it as unjust, intolerable, and deserving of corrective action” (p. 587). The experience of a protest could yield cognitive changes that bolster future inclination to protest. However, it is not clear whether the findings apply to protests such as workplace contention in China, 54

Finkel, 1985; Finkel, 1987; Sherkat and Blocker, 1997; Finkel and Muller, 1998; Kelloway et al., 2007.

5.3 Understanding Heterogeneous Protest Intention

157

Age

Male

Education

Rural

Income

Supervisor

Tenure

Migration years −0.2

−0.1

0

0.1

0.2

FIGURE 5.3 Dependent variable: Have you participated in collective action? Note: The variables on the y-axis indicate the independent variable. The dependent variable is the respondent’s past participation in collective action. Average marginal effects are based on a logit regression analysis with firm-level random effects. Robust standard errors are clustered at the firm level.

where few incidents are led by dedicated activists and well-organized protest campaigns. Recall that respondents were asked whether they had participated in any collective action. Although line supervisors were less willing to protest than rank-and-file workers, they were much more likely to report past protest behavior (see Figure 5.3). This could be due to their longer firm tenure, but even after controlling for firm tenure, the effect of job position is large and statistically significant. Line supervisors were 7.0 percentage points more likely to have participated in the past. This shows that despite line supervisors’ weaker intention to join a protest, due to the collective action dynamics within workplaces, they end up experiencing more labor protests. Those with higher income and education, however, were less like to have engaged in collective action. The result reflects the conclusions in the literature that those with more knowledge and social resources to solve their problems might not join collective action.55 55

Becker, 2012.

158

Workplace Mobilization and Collective Action Protest intention Job availability

Desire for a pay raise No promotion Group efficacy Mobilization Fired Police Self-efficacy −1

0

1

2

3

FIGURE 5.4 Predicted effect of past participation on protest intention and determinants Note: The variables on the y-axis indicate the dependent variables. The independent variable is the respondent’s past participation in collective action. Average marginal effects are based on logit regression analyses with firm-level random effects. Basic demographic controls are included in the model. Robust standard errors are clustered at the firm level.

Those who had participated in a protest were much more likely to indicate future protest intention. Figure 5.4 displays the predicted effect of past participation on protest intention and its determinants based on logit regressions. When controlling for demographic variables, past participation increased one’s inclination to protest by 25.2 percentage points, which is a substantial effect. One of the most important channels through which the past behavior affects future intention seems to be by increasing one’s perception of group efficacy, which confirms the findings in the literature. When controlling for demographic variables, past behavior also increased one’s expectation of successful mobilization and self-efficacy. The findings might originate from the protest participants’ experience of success during collective action. Qualitative evidence supports the relatively high success rate of labor protests: A former official of Guangzhou Trade Union, for instance, revealed during a labor relations workshop in 2013 that about 80 percent of labor protests end in workers’ favor.

5.4 Conclusion

159

The close link between past protest behavior and future protest intention shows that protest participants are not discouraged from protesting again. On the contrary, past participation is the strongest predictor of protest intention. While the analyses do not reveal the causal direction of the relationship between the two, they suggest that atomized incorporation is not a system designed to minimize labor unrest. This could explain why certain factories experience repeated episodes of collective action. Yet, as demonstrated in Chapter 4, those with protest experience are less likely to blame the central and local government and more likely to blame the firm. Thus, while atomized incorporation could induce unstable labor relations by increasing workers’ future protest intention, it may not necessarily mean a substantial political threat for the regime. In this sense, the findings confirm the extant literature’s conclusion that social protests in post-reform China might be a source of political instability but not necessarily a threat to the regime’s political legitimacy.56 5.4 CONCLUSION

In this chapter, I explored the barriers to collective action for atomized protests and probed why workers rarely protest, despite their structural empowerment. I highlighted how worker heterogeneity in grievances, resources, and attitudes toward protests, is channeled through firm-level organization to induce collective action. While the literature has proposed that changes to the socioeconomic environment are favorable for labor protests, these changes haven’t led to proportionately more collective action, because they have not had the same effect on all workers. In particular, I argue that the growing labor turnover, coupled with the regime’s restrictions on broader mobilization, made it difficult for workers with stronger grievances about limited economic opportunities – and thus social grievances – to initiate a protest. Those with mobilizational resources, including low-level managers and workers with long tenure at the firm, are less affected by the tight labor market due to their firm-specific embeddedness. A combination of factors – labor shortage, firms’ lack of interest in providing opportunities for promotion, and the regime’s restrictions on broader mobilization beyond the firm boundary – have contributed to the situation where those with the high levels of generalized discontent have limited mobilizational resources to initiate

56

Perry, 2008; Tong and Lei, 2013.

160

Workplace Mobilization and Collective Action

collective action. Those with mobilizational capacity, on the other hand, are generally uninterested in atomized protests. The analysis I present in this chapter shows that it is difficult to initiate a protest, but that once a protest is initiated it is likely to grow. Those who have mobilizational capacity are substantially less inclined to protest, but ordinary production workers with limited firm tenure – who rarely have the capacity to mobilize – still can easily become enthusiastic followers. These workers are less goal-oriented than line supervisors, and some enjoy the expressive benefits of joining a collective protest. They are also animated more by social grievances beyond job-specific conditions. For line supervisors, the perception that protesters will be able to achieve their goal is one of the most important determinants of protest intention; they are more likely to approach collective action strategically than rank-and-file workers. The relationship between past protest experience and protest intention shows that atomized incorporation does not minimize labor contention: Those who had participated in a protest before were much more inclined to protest in the future and also perceived higher levels of group and personal efficacy. The positive relationship in part originates from the core logic of atomized incorporation, which hinges on both the positive and negative inducements. The positive rewards to atomized protesters increase their willingness to protest again. Yet, this does not mean that the regime’s strategy and its interaction with the socioeconomic environment have translated into politically threatening circumstances. It is still difficult to produce collective action, and the workers with the most resources to initiate atomized protests are not interested in protesting when they blame the central government. Those who have not experienced and do not expect to experience upward mobility in the workplace, on the other hand, are more interested in atomized protests and more likely to blame the central government, but don’t have enough social capital to initiate one.

6 Interest-Based Collective Action and Firm-Level Patterns of Labor Protests

In the mid-2010s, a new conversation emerged about “interest-based” protests and what they mean for China’s political future. Interestbased protests often occur at law-abiding firms and advance extra-legal demands (fawai suqiu) that go beyond the mandated minimum wages and standards. NGOs and activists report that there are a substantial number of labor protests at law-abiding firms. Hao (2016) writes that labor strikes for wage increases “[have] not necessarily taken place in factories with the worst working conditions” (p. 183). There have been high-profile protests at foreign-invested and large-scale enterprises that abide by labor laws. According to the China Labour Bulletin (CLB), a Hong Kong – based labor NGO, many labor strikes took place at “wellestablished and profitable” factories.1 In other words, it does not appear that labor protests were taking place only at factories with the worst labor conditions and illegal labor practices. Scholars echoed these observations and argued that “interest-based” protests – as opposed to “rights-based” protests – were on the rise.2 Some see this change as reflecting workers’ growing capacity to “challenge Beijing’s authority.”3 As the waves of strikes demanding higher wages swept China, Economic Daily from Hong Kong warned that amid rising social tensions in mainland China, migrant workers’ protests for higher wages “could link with other social grievances and develop into a demonstration against the society or even the central government.”4 1 2 3 4

China Labour Bulletin, 2011. Butollo and Brink, 2012; Chen and Tang, 2013; Elfstrom and Kuruvilla, 2014. Shirouzu, 2010. Rapoza, 2011.

161

162

Interest-Based Collective Action and Firm-Level Patterns

Butollo and Brink (2012) characterize the shift in protest patterns in 2010 as a sign of “class formation.” Chen and Tang (2013) predict that the interest-based protests “would have enormous implications for China’s labor movement in the years to come, as it might motivate more workers to seek collective rights and generate the state’s institutional response.”5 In other words, it is postulated that interest-based protests might have greater potential for developing into politicized activism than law-based protests. The debate about interest-based protests highlights the inherent risk of atomized incorporation. As discussed in Chapter 3, the “incorporating” aspect of atomized incorporation – the adoption of pro-labor laws and more tolerant attitudes toward atomized protests – could pose a risk to the regime when it encourages workers to demand more than jobspecific gains. If it is true that interest-based protests signal workers’ political consciousness and potential to challenge the regime’s atomization, this constitutes an important dilemma for the regime: The protests provide important information about the location of a potential threat but could develop into politicized activism. The regime has developed a set of institutions and mechanisms – such as courts, trade union, and legal aid services – to resolve workers’ legal demands but rely mostly only on ad hoc bargaining for extra-legal demands. Interest-based protests, therefore, could pose a more significant threat than law-based protests if they are initiated by those interested in seeking a broader change. We can look at where different types of collective action arise to understand the political risk of atomized incorporation. What drives workers to engage in interest-based collective action? The emergence of interestbased collective action is puzzling, given the barriers to collective action among unorganized workers. When workers suffer from acute barriers to collective action, how can they motivate collective action around interestbased goals? At the same time, does interest-based collective action signal the politicization of labor conflicts? When workers move from “rightful resistance” to making demands that are not legally guaranteed, might those workers also seek a broader change beyond the workplace? In other words, what does the regime learn from an interest-based protest? I focus on analyzing firm-level patterns of labor protests and understanding the characteristics of interest-based collective action. I use both my survey data and an online database of labor protests in Chinese manufacturing sectors collected by the CLB. I use the datasets to chart 5

Chen and Tang, 2013, p. 566.

6.1 The Rise of Interest-Based Collective Action

163

firm-level patterns of collective action and investigate how collective action for law-based and interest-based demands differs in its tactics and state responses. I find that labor protests for interest-based demands have been much more prevalent at law-abiding firms. Collective action for interest-based demands also uses different protest tactics, because it does not rely on state authorities to solicit concessions. These findings suggest that interest-based collective action could fit the political logic of atomized incorporation better than law-based collective action, as it is easier to frame as a firm-specific issue. 6.1 THE RISE OF INTEREST-BASED COLLECTIVE ACTION

In the post-reform era, the Chinese regime has encouraged an increased role for law and judicial processes in labor politics. The legal procedures for labor dispute resolution were reintroduced in 1993, and the first labor law was promulgated in 1994. New labor laws introduced in the last decade include the Labor Contract Law (2008), the Law on Mediation and Arbitration of Labor Disputes (2008), the Employment Promotion Law (2008), and the Social Security Law (2011). These laws have defined legal rights for workers and provided an institutional mechanism for resolving labor conflicts. The Labor Contract Law has been particularly beneficial for migrant labor, because it brings workers into formal labor contracts and gives them access to official channels for raising grievances.6 Chinese workers have increasingly relied on legal and administrative channels to deal with labor disputes. The number of cases accepted by the labor dispute arbitration committee increased from 48,121 in 1996 to 894,053 in 2018. Workers have adopted legal channels as their “weapon” as their awareness of their legal rights increases.7 An experience of legal mobilization makes workers more prone to use legal channels again.8 Legal mobilizing has been a key strategy for labor NGOs and trade unions, because it is considered to be the most legitimate, and thus politically safe, way to assist workers.9 6

7 8 9

The 1994 Labor Law did not substantially benefit migrant labor. The law had ambiguities that made it difficult to enforce and the majority of migrant workers did not have a labor contract. Without a contract, migrant workers could not demonstrate their employment relations and thus were discouraged from pursuing official channels of conflict resolution. For a detailed review of the 1994 Labor Law and its effects, see Ngok (2008). Gallagher, 2005. Gallagher, 2006. Chen, 2004; Xu, 2013.

164

Interest-Based Collective Action and Firm-Level Patterns

If collective action is perceived as a form of legal mobilization, then workers’ grievances, in court or on the street, are seen as staying within the legal boundary. If law determines legitimacy, workers might only engage in collective action to realize legal rights. Scholars have suggested that Chinese citizens perceive an opportunity for contention when they see “a gap between rights promised and rights delivered.”10 If workers’ collective action is primarily driven by the gap between the legally guaranteed rights and their actual implementation, we would expect to see very little collective action at law-abiding firms. Newspaper reports on workers’ protests in China reflect this rhetoric: In a New York Times article, for instance, Geoffrey Crothall argues that labor unrest is the price that the government pays “for decades of not enforcing the regulations,” and that “the only way to prevent more unrest is, simply, for the government to enforce the laws.”11 Yet even before the Honda strike in 2010 that spawned many academic studies on the changing nature of labor protests in China, local government authorities had registered a rise in interest-based collective action. A 2008 report by a district-level labor bureau in Shenzhen, for instance, notes that one of the major characteristics of the labor protests at the time is their pursuit of “illegal” (feifa) or “extra-legal” (fawai) demands.12 According to the report, labor protests in export-oriented regions were increasing in frequency and size, and changing in terms of their goals, tactics, and effectiveness. The report noted, “often protesters make demands that are not based on law and insert them between legal ones.” In addition to the extra-legal demands, the report emphasizes the new characteristics of labor protests: (1) Protesters have become more prone to engage in collective action outside the officially sanctioned channels, including strikes, road blockages, and petitioning higher-level governments; (2) while protests have become more collective, protesters often refuse to select their representatives to negotiate with the firm or the government. Instead, they try to use their collective power to satisfy their demands; (3) protesters use various tactics strategically to directly harm the firms’ profits. Interest-based collective action can be more difficult to initiate for several reasons. First, it is harder for workers to achieve consensus on 10 11 12

O’Brien and Li, 2006. Crothall, 2016. The report does not distinguish between the terms “illegal” and “extra-legal” but uses them interchangeably (Longgang District Government Labor Bureau, 2008, pp. 428–438).

6.1 The Rise of Interest-Based Collective Action

165

their demands. It is easy for workers to articulate legal demands because they can be anchored to legal rights. Workers fighting for payment of wage arrears, for instance, do not need to go through a lengthy process of communication and persuasion to identify major goals and mobilize protesters. Using similar logic, Leung (2015) finds that defensive strikes to restore lost entitlements can be spontaneous and unorganized, as they generally do not require planning. Interest-based collective action, on the other hand, often involves more careful planning, negotiating within the organizing group to agree on the demands, and persuading others to join the collective action. While a significant proportion of workers support joining a labor protest, as I described in Chapter 5, this does not mean that protest initiators can organize around every possible demand. Second, interest-based collective action might require different tactics. Workers fighting for legal rights use tactics directed toward securing sympathy and concessions from the state authorities. Protesters often skip the administrative level and petition a higher level of government – called “skip-level petitioning” – to enhance their chance of success.13 Studies show that protesters perceive the higher-level government to be more benevolent than local government and use this to pressure the higher authorities to look at what the local government is doing.14 Other tactics such as contacting the media or blocking the road are used mainly to get the attention of state authorities. Despite the state’s attempt to describe mass incidents as “internal conflicts,” most of the protests target local governments.15 In nearly all scenarios of law-based protest, the action is designed to engage the state government as both the party responsible for seeing that rights are enforced, and – along with the individual firm – the party that wants to see the conflict resolved. During interest-based collective action, however, workers cannot target on state authorities as the final authority on whether concessions are granted, since it is ultimately the firm’s decision whether to meet protesters’ demands. When the firm violates labor laws, the local authorities can use legal grounds to pressure the firm to meet protesters’ claims and resolve disputes, but the local government has limited capacity to support protesters’ extralegal demands. The Labor Bureau of the Longgang district in Shenzhen, for instance, recommends that officials deal with protests with the principle of “fairness based on law.”16 According to the guideline, the relevant 13 14 15 16

Chen, 2009; Cai, 2010; Chen and Xu, 2012. Cai, 2010. O’Brien and Li, 2006; Yu, 2010. Longgang District Government Labor Bureau, 2008.

166

Interest-Based Collective Action and Firm-Level Patterns

authorities should protect workers’ legal rights but should not support their extra-legal demands. Even when workers adopt disruptive tactics or persist with their extra-legal claims with the widespread belief that punishment is not possible when the law is popularly ignored (fabuzezhong), the local government should not succumb to the pressure. Instead, the local government is directed to spot troublemaking workers and conduct legal education to clear their “misunderstanding” about extra-legal rights. The protocol suggests that interest-based collective action creates a dilemma for local governments. On the one hand, local officials aim to minimize the incidence of collective action in their jurisdiction, because it makes them look bad to higher levels of government. Since the cadre responsibility system in China prioritizes maintenance of the social order – in addition to other important goals such as economic development – local government authorities have sought to discourage disgruntled citizens from taking their grievances to the street, and particularly from appealing to higher levels of government.17 Local government officials are expected to prioritize the prevention of labor protests, rather than waiting until after they occur to resolve them. While some local governments have succeeded in preventing workers from protesting for legal rights such as payment of wage arrears, they lack effective tools for discouraging workers from protesting for interest-based goals. Perhaps the only way for local governments to prevent interest-based collective action is by disseminating information about the law and creating the perception that only legal demands are supported. For instance, a labor activist writes that during a labor protest for a wage increase at an electronics factory, the local union cadres visited the protest scene and commented that the workers’ request was “illegal” since the factory already paid the legal minimum wage.18 Because local governments have limited capacity to mediate labor disputes for extra-legal demands, workers’ traditional troublemaking tactics to appeal to state authorities, such as skip-level petitioning and blocking roads, are less effective and less likely to be deployed during interest-based collective action. 6.2 FIRM-LEVEL PATTERNS OF LABOR PROTESTS

Are interest-based protests an unintended consequence of atomized incorporation that could threaten the regime’s resilience? The political implications of interest-based collective action depend on whether the 17 18

Edin, 2003; Wang, 2015. Hao, 2016, p. 207.

6.2 Firm-Level Patterns of Labor Protests

167

demands of the action can be diverted into a channel that the state controls. State authorities have at their disposal administrative and bureaucratic channels that they can try to use to dissolve law-based collective action. They can respond to collective action by trying to direct protesters back to legal channels and break the protest down to individual cases. In this way, according to Chen and Gallagher (2018), the Chinese regime has been able to “fix” labor insurgency. Yet, interestbased collective action cannot easily be fixed because it cannot be diverted into these channels. Proactive intervention by court judges to resolve collective action, as described in Su and He (2010), is also not feasible with protesters who make extra-legal demands. State bureaucrats resort to informal mediation and bargaining to resolve law-based collective action19 but often have limited tools and resources when it comes to interest-based collective action. This means that interest-based collective action might escalate more easily, which could be a source of a threat to the regime if the protesters are interested in seeking a broader change. Understanding firm-level patterns of collective action, therefore, could be critical in unveiling whether it is the most threatening groups of workers that engage in interest-based collective action. Existing studies have not provided an in-depth analysis of firm-level patterns, as most focus on either individual or macrolevel factors. Some studies have probed the differences between the protests of laid-off workers at SOEs and those of migrant workers in the private sector,20 but there has been limited attention to firm-level differences in collective action within private industries. Given that the unit of the collective action is the firm, understanding the dynamics of collective action at the firm level is imperative. This information is also critical for understanding which workers are being rewarded by engaging in atomized protests. 6.2.1 Labor Relations at Foreign-Invested Firms Among firm-level variables, foreign ownership received scholarly attention in the 1990s, when post-reform China opened its doors to foreign direct investment. Some studies have analyzed labor relations at firms that received foreign investment and made comparisons with other types of firms. Some observed that labor relations at foreign firms were much more unstable than at other types of firms.21 According to Gallagher (2005), for 19 20 21

Lee and Zhang, 2013. Lee, 2007a. Guo et al., 1996; Gallagher, 2005; Cooke, 2008.

168

Interest-Based Collective Action and Firm-Level Patterns

instance, foreign enterprises had experienced more frequent and collective labor disputes in the late 1990s, which “include[d] more workers, and [. . . ] led to strikes, slowdowns, and general social instability” (p. 99). Guo et al. (1996) also show that foreign enterprises experienced the most unstable labor relations during the early and mid-1990s in the Fujian Province.22 In particular, foreign enterprises in coastal regions – which accepted inflows of FDI from neighboring states earlier in the post-reform era – seem to have experienced a higher incidence of protests. However, these studies only examine the varying frequency of labor disputes at firms with different types of ownership and do not explore the nature of labor disputes. One important characteristic of foreign-owned firms is that they are more likely to abide by labor laws than domestic private firms. In early studies of labor relations in post-reform China, a higher incidence of labor contention at foreign-owned enterprises was attributed to their tendency to exploit workers or cultural differences. Studies based on data from the 1990s have pointed to the “race to the bottom” theory and suggested that labor conditions at foreign enterprises were harsher due to the globalization-driven pursuit of sweatshop conditions. These poor labor conditions were postulated to have fomented labor grievances and resistance.23 Other studies argued that the mandatory military training in neighboring Asian countries such as South Korea and Taiwan contributed to a militaristic culture in firms receiving Asian investment and that this provoked discontent among Chinese workers.24 It is undeniable that foreign-owned factories created subpar labor conditions compared to state-owned firms, but compared to domestic private firms, foreigninvested firms in China and elsewhere offered better labor conditions and were more likely to abide by labor laws.25 6.2.2 Survey Evidence A survey from 2011 in Guangzhou demonstrates that foreign-owned enterprises – compared to domestic private firms – experienced a higher incidence of collective action during 2010, despite their stronger tendency to abide by labor laws. As shown in Table 6.1, the results from a logistics regression show that a worker at a foreign-owned firm was more 22 23 24 25

The study shows that among 124 cases of collective labor disputes that occurred in 1994, about 60 percent (or 74 cases) took place at foreign enterprises. Chan and Senser, 1997; Chan, 2001, 2003. Chan and Xiaoyang, 2003. Liu and Guo, 2004; Wan et al., 2006.

6.2 Firm-Level Patterns of Labor Protests

169

TABLE 6.1 Firm-level determinants of labor protests

Variable

Protest

Foreign-owned Collective Firm size (logged) Total N

0.055∗∗ –0.047 0.017∗∗∗ 1017

Note: ∗∗∗ p < 0.01, ∗∗ p < 0.05. Source: Spring Festival Train Survey of Migrant Workers (2011), Sun Yat-sen University, Guangdong, China.

likely to have experienced a labor protest than one at a domestic private firm by 5.5 percentage points.26 Collectively owned firms were not substantially different from domestic private firms. State-owned firms, which comprised 5.9 percent of the dataset, were dropped since none of the workers at state-owned firms reported an incidence of collective action. In addition to firm ownership, firm size also had a positive effect on protest incidence. As displayed in Table 6.1, workers in the largest firm in the sample – with 400,000 employees – were nearly 14 percentage points more likely to report an incidence of collective action than those at the smallest firm, with 100 employees.27 Both foreign ownership and large size positively predicted the firm’s compliance with labor laws.28 Analyses of my survey data from 2013 also confirm the positive relationship between protest incidence and law-abiding behavior. I used logistics regressions to measure the effect of the firm’s tendency to abide by labor laws and occurrence of a labor protest. Firms’ law-abiding behavior (law score) was measured by averaging the respondents’ answers to nine questions in the survey. The nine questions are whether the worker signed a labor contract, has a base salary higher than the legal minimum wage, receives the legally defined overtime payment, is paid on time, and 26

27

28

The dependent variable is respondents’ binary answer to the question, “Did your firm experience workers’ collective action (labor strike, sit-in, marching, etc.) during the previous year?” The baseline category for ownership is domestic private firms. Since few respondents were employed at state-owned firms, the category is omitted from the analysis. Firm size denotes logged employment size and ranges from 0.69 to 12.9. The table reports average marginal effects based on logistics regressions. The survey asked workers if the firm provides a labor contract and injury-related insurance. For both questions, foreign ownership and firm size had a positive and significant effect, after controlling for demographic variables.

170

Interest-Based Collective Action and Firm-Level Patterns

is provided with five types of insurance (medical, pension, injury, unemployment, and maternity) and the housing providence fund (i.e., wuxian yijin).29 Out of 156 firms, 23 firms (14.7 percent) had a perfect law score, meaning that the workers employed at these firms answered yes to all of the nine questions. The law score in the sample was on average 0.74 with a median 0.78.30 Firms with a high law score reported a higher incidence of collective labor protest. The survey asked whether there had been any labor strike or work stoppage (tinggong bagong) since January 2012. The binary variable strike was coded as 1 if any worker employed at the firm replied yes, and 0 otherwise. In total, at twelve firms (7.7 percent), workers reported a strike incident. Of the twelve firms that experienced a labor strike, 3 firms were domestic and nine were foreign enterprises. Those that had experienced a labor protest had on average 1528 employees, while those that had not experienced one had about 460 employees. To test the relationship between firm-level characteristics and strike incidence, I ran logit regressions for the dependent variable strike. The results in Model 1 show that while both foreign ownership and firm size had a positive impact on strike incidence, the effect of foreign ownership was not statistically significant (Table 6.2). This is partly due to the fact that foreign-owned firms are on average larger than domestic private enterprises.31 All of the firms that had experienced a strike during the previous year had an aboveaverage law score. As displayed in the first column of Table 6.2, law score has a large and statistically significant effect on strike incidence: When law score increases by one standard deviation, the likelihood of a strike incident increases by roughly 7.0 percentage points. As shown in Model 3, even after controlling for the firm’s ownership status (foreign) and size, the effect remained substantial and statistically significant. There was a follow-up question in the survey, to test whether the labor strikes occurred for an interest-based goal. Respondents were asked 29

30

31

The law score ranges from 0 to 1: If all of the respondents at a firm replied yes to the nine questions, then the firm would receive 1.0 for its law score; if there were no positive replies to the questions, then the firm’s law score would be 0. The relatively high score reflects the fact that most of the firms paid the local minimum wage (91.7 percent), signed a labor contract (83.5 percent), and provided the insurance for work injury (82.0 percent). The result might also be because the manufacturing sector exceeds others in the labor regulation enforcement. Yet, firms lagged behind in other types of legal rights, most notably in their provision of the housing providence fund and maternity insurance: Only 23.1 percent of the sampled firms provided the housing providence fund while 30.1 percent covered maternity insurance. Foreign-owned firms on average had 749 employees, while domestic private firms had 324 employees.

6.2 Firm-Level Patterns of Labor Protests TABLE

171

6.2 Determinants of strike incidence at the firm

level Model 1

∗∗

Law score Foreign Size Total N

Model 2 0.365 (0.161)

0.044 (0.050) 0.031∗ (0.018) 156

156

Model 3 0.302∗ (0.165) 0.012 (0.050) 0.026 (0.018) 156

Note: The table presents average marginal effects and robust standard errors from logistic regressions. ∗∗ p < 0.05, ∗ p < 0.1.

about the main demand in the labor strike. Of the twelve labor protests, the most common grievance was that the wage was too low and the workers wanted higher wages. Workers at five firms (41.7 percent) gave this answer. The next most common answer was wage arrears, given by the workers at three firms (25.0 percent). The other four labor protests were for other reasons – one for industrial injury-related compensation, another for increased labor intensity, and two for hostile relations between Japan and China.32 The data show that interest-based collective action was prevalent in 2012. The survey also asked whether the workers succeeded in securing any of the demands that they were protesting for: in eight cases (75 percent), workers reported that as a result of collective action their firm met at least some of the protesters’ demands, revealing relatively high success rates. Although the survey data only provide preliminary evidence due to its small sample size, the findings correspond to a comment from a former official of the Guangzhou Trade Union, that more than 80 percent of labor protests end in favor of workers.33 Protests for higher wages occurred predominantly at law-abiding firms. Four out of the five protests for higher wages were at foreign-owned firms; when domestic private firms experienced a labor protest, it was usually because of wage arrears. These results suggest that the high-profile strikes at foreign-owned enterprises in the early 2010s were not only observable due to biased 32

33

In 2012, China’s relations with Japan soured due to their disputes over the Diaoyu/Senkaku island. Many labor protests occurred as a result of the dispute. For a more detailed analysis of how the international dispute prompted a wave of labor strikes at Japanese-owned factories, see Chapter 7. Participant observation 05-2013.

172

Interest-Based Collective Action and Firm-Level Patterns

media reporting. The two survey datasets from this period show that interest-based protests mostly occurred at law-abiding firms, which are usually foreign owned and large scale. Due to the prevalence of interestbased collective action during this period, law-abiding firms experienced overall a higher incidence of collective action than those firms with illegal labor practices. The findings counter the dominant view that these labor protests were a reflection of the regime’s failure to enforce labor laws. On the contrary, workers’ collective action was more prevalent where firms complied with labor laws. 6.2.3 The CLB Strike Map Dataset The two survey data sets, however, might not provide a generalizable answer, since each of them only covers one year and one region. To situate the survey data in a larger context, I analyze an online database of labor protests in mainland China collected by the CLB. The CLB Strike Map does not include every labor protest, but given the lack of official data on collective action, it is still one of the most extensive records available. The majority of labor protests in the database are identified from social media posts made by protesters and occasionally by bystanders. The dataset could suffer from selection bias, since there might be a systematic reason why some cases of collective action were included but not others. For instance, collective action in remote regions may not be registered in the database due to low visibility. However, the dataset still provides a unique resource for comparative analyses, as long as the direction of the bias does not correlate with the firm-level variables. The CLB report reveals that interest-based demands made up only 9 percent of all reported cases of labor protests in 2000 but increased to 30 percent in 2010.34 Since the raw data from this period are not available, I use the Strike Map Data that records labor protests from 2011 to 2019. For each observation of a labor protest, the CLB Strike Map also records information about the firm ownership, location, protest size, major demands of protesters, their tactics, and government responses. In the dataset, there are 3,325 labor protests in the manufacturing sector, and in about 8.4 percent of the protests, there was a pay increase demand. Wage arrears were the most common driver of collective action, comprising about 65.1 percent of the protests. The time-series changes show that interest-based protests became more common up to 2012 but 34

China Labour Bulletin, 2012.

6.2 Firm-Level Patterns of Labor Protests

173

FIGURE 6.1 Percentage of demands by year Source: The CLB Strike Map dataset.

decreased after 2013. As displayed in Figure 6.1, labor protests for a pay increase comprised 38.8 percent of the protests in 2012 but only made up 1.1 percent in 2018. Demands for wage arrears, on the other hand, became much more common after 2012. In 2012, only 20.8 percent of the protests were for wage arrears, but this increased to 79.5 percent in 2018. The data show the prevalence of interest-based protests in the early 2010s, comprising almost 40 percent of the labor protests in a given year. This echoes my survey findings that more than 40 percent of the labor strikes demanded a pay increase in 2012. The CLB dataset puts the survey data into perspective: The strike activity measured in the survey took place during 2012 when labor protests over higher wages were most frequent. The finding confirms conclusions from previous studies that the number of labor protests for higher wages increased up to 2012.35 Earlier studies, however, only looked at the data up to 2012 and could not document the reversing trend. There might be many reasons why interestbased protests have become less common, including economic factors and changes to the political environment. I return to this topic in Chapter 8. 35

Elfstrom and Kuruvilla, 2014.

174

Interest-Based Collective Action and Firm-Level Patterns TABLE 6.3 Ownership demands (% of protests)

and

legal/extra-legal

Ownership

Wage increase

Wage arrears

N

Foreign HK/Taiwan Joint Private State Unknown

21.8 15.3 17.5 3.6 8.4 8.6

21.8 43.6 51.8 82.5 38.2 65.1

357 346 114 1771 249 499

Source: The CLB Strike Map.

While the dataset does not contain any information on firms’ law-abiding behavior, we can still examine the relationship between protesters’ demands and firm-level characteristics. The dataset includes information on firm ownership and the size of protests measured by the number of participants. As demonstrated in Table 6.3, protests for a pay increase were most common at wholly foreign-owned factories, comprising 21.8 percent of all incidents, and least common at domestic private factories where only 3.6 percent of the labor protests demanded a pay increase. Firms owned by investors from Taiwan and Hong Kong (sanzi firms) and jointly owned firms were in the middle: 15.3 percent of labor protests at sanzi firms and 17.5 percent of protests at jointly owned firms demanded a pay increase. Pay increases were also a less common reason for labor protests at state-owned enterprises. Protests for wage arrears, however, were most prevalent at privately owned enterprises, comprising 82.5 percent of all protests. Payment of wage arrears was the least common demand of labor protests at wholly foreign-owned enterprises, making up 21.8 percent of all cases. As shown in Table 6.4, the number of protest participants was also correlated with the frequency of law-based and interest-based protests: Larger protests were more often for a pay increase, while most protests with fewer than 100 participants demanded payment of wages. While the number of protest participants is not the same as the size of the firm, given that these protests are workplace-based collective action, the data still provide support for the hypothesis that it is relatively rare for small-scale firms to experience collective action for an interest-based goal. Together with the effect of firm ownership, the data show that the goal of labor protests might differ depending on firm ownership and size. Collective

6.2 Firm-Level Patterns of Labor Protests

175

TABLE 6.4 Firm size and legal/extra-legal mobilization (% of protests)

Size

Wage increase

arrears arrears

N

1–100 persons 100–1000 persons 1000 or more persons

4.1 13.5 23.8

76.3 50.9 27.0

2102 979 252

Source: The CLB Strike Map.

action for an interest-based cause is unlikely at the domestic-private and small-scale firms that generally provide the worst labor conditions. For a full analysis, I classified protest demands into three categories: extra-legal demands, legal demands, and the demands that could be legal or extra-legal. Out of the eight kinds of protest demands that the CLB Strike Map classifies, two were coded as extra-legal demands (wage increase and managerial supervision), three were coded as legal demands (wage arrears, overtime pay, and social insurance), and the rest were coded in between (relocation, compensation, layoff, and others). Workers’ protests for compensation and layoffs could be related to a legal clause but sometimes do not have a legal basis. For instance, one of the most common demands of protests has been “buying off the tenure” (maiduan gongling) during company restructuring. Legally, the firm does not always have an obligation to compensate the workers according to their tenure as long as the firm maintains the labor contract, but many protesters have asked that the firm terminate their labor contract and pay compensation. The variable extra-legal was created to denote extra-legal demands as 2, legal demands as 0, and the rest as 1. If protesters made both legal and extra-legal demands, the protest was coded as legal. While 10.1 percent of protests made an extra-legal demand, about 6.5 percent of protests – those that made only extra-legal demands – were classified as extra-legal. 72.7 percent were classified as legal. Statistical analyses also confirm the relationship between protest size, firm ownership, and the nature of demands. When included in the same model, both foreign ownership and protest size predict a higher probability of interest-based collective action and a lower probability of law-based action (see Table 6.5). The first two columns of the table demonstrate the effect of firm ownership and protest size on the most common type of demand for interest-based and law-based collective action: wage increase and wage arrears. In all models, the baseline ownership category for firm

176

Interest-Based Collective Action and Firm-Level Patterns TABLE

6.5 Multivariate analysis of protest demands

Foreign Taiwan/HK Joint State Unknown Size Total N

Wage increase

Wage arrears

Extralegal

0.069∗∗∗ (0.013) 0.038∗∗∗ (0.014) 0.065∗∗∗ (0.020) 0.018 (0.018) –0.020 (0.056) 0.028∗∗∗ (0.007) 3,323

–0.355∗∗∗ (0.021) –0.176∗∗∗ (0.021) –0.144∗∗∗ (0.034) –0.276∗∗∗ (0.022) 0.078 (0.094) –0.088∗∗∗ (0.011) 3,323

0.971∗∗∗ (0.075) 0.523∗∗∗ (0.078) 0.633∗∗∗ (0.121) 0.737∗∗∗ (0.085) 0.257∗∗∗ (0.074) 0.220∗∗∗ (0.038) 3,323

Note: The first two columns present average marginal effects based on logistic regressions. The final column is based on an ordered probit model. The baseline ownership is domestic private enterprises. All models include year fixed effects. ∗∗∗ p < 0.01.

ownership is domestic private firms, and protest size is a category variable that ranges from 1 to 3.36 Based on logistics regressions, the results in the first column show that wholly foreign-owned firms were 6.9 percentage points more likely than domestic private firms to experience a wage increase protest. Joint ownership and sanzi ownership by Taiwan and Hong Kong also had a positive effect. Wage increase was also a more common demand in the larger size protests. As demonstrated in the second column, however, the opposite was true for labor protests that demanded payment of wage arrears. Compared to domestic private firms, all types of firms were substantially less likely to experience a protest for wage arrears. Wholly foreign-owned enterprises, in particular, were 35.5 percentage points less likely to experience this than domestic private firms. Finally, the last column shows the effect of firm ownership and protest size on the variable extra-legal based on an ordered probit model. The results also confirm that interest-based collective action is least common at domestic private firms and involves more protest participants. 36

1 is when the protest involves less than 100 participants, and 4 is when the protest involves more than 10,000 participants.

6.3 Characteristics of Interest-Based Collective Action TABLE

177

6.6 Protest tactics

Tactics Strike Sit-in Road block Marching Suicide N

29.1% 29.5% 13.5% 49.4% 2.8% 3,325

Source: The CLB Strike Map. The categories are not mutually exclusive.

6.3 CHARACTERISTICS OF INTEREST-BASED COLLECTIVE ACTION

The same dataset allows us to examine differences in protest tactics and outcomes for collective action with legal or extra-legal demands. For each observation of labor protests, the dataset records the protest tactics used – labor strike, road blockage, sit-in, marching, and suicide attempt – and whether police were present. I use these two variables to investigate differences in interest-based and law-based collective action. 6.3.1 Protest Tactics While a labor strike is one form of disruptive contention, workers can use disruptive tactics to expand the reach of the conflict beyond the factory gate, for example by blocking the traffic or marching to government offices. If local authorities are not willing to or capable of helping workers meet their extra-legal demands, we would hypothesize that workers would not employ these strategies to attract their attention. The CLB dataset shows that, as displayed in Table 6.6, the preferred protest strategy for collective action in manufacturing industries was marching to the government offices. This reflects industrial workers’ continued reliance on local governments as a conduit for contention. Labor strikes and sit-ins were also relatively popular protest tactics, each comprising nearly 30 percent of all protest incidents. Highly disruptive strategies such as blocking the road or threatening to jump off a building were least common. Statistical analyses based on logit regressions of extra-legal on the choice of protest tactics (Figure 6.2) show that protests for an extra-legal cause are significantly more likely to involve a labor strike but less likely

178

Interest-Based Collective Action and Firm-Level Patterns

strike

march

sit−in

blockage

suicide −1

−.5

0

.5

1

1.5

FIGURE 6.2 Effect of extra-legal demands on protest tactics Note: Protest tactics on the y-axis are the binary dependent variables. The independent variable is extra-legal. All of the models have protest size as a control variable and year fixed effects.

to employ more disruptive tactics that take the conflict outside individual workplaces. When the protest demands were exclusively interest-based (extra-legal = 2), the protest would be 30 percentage points more likely to take the form of a labor strike than when at least one of the goals was a legally protected right. A labor strike was the only protest tactic that was positively correlated with extra-legal demands. This shows that interestbased collective action might fit the logic of atomized incorporation better than law-based contention, since those protests are more likely to be contained within the firm boundary. During interest-based collective action, protesters rarely target local government authorities with highly disruptive actions that breach the boundaries of a single factory. The Longgang district government’s assessment in 2008 – that workers increasingly did not want to involve local governments as intermediaries but rather to use their collective power to directly harm the firm’s profits–could reflect the rise of interest-based collective action. During interest-based collective action, workers need to find alternative allies to enhance their bargaining power, as seeking concessions by leveraging the power of the state will not meet their goal.

6.3 Characteristics of Interest-Based Collective Action

179

TABLE 6.7 Determinants of police presence

Extra-legal

Police presence

2 (extra-legal) 1 (uncertain) 0 (legal)

24.8% 36.8% 30.9%

Source: The CLB Strike Map.

6.3.2 Police Presence The CLB Strike Map notes whether police attended the protest scene. Since the dataset is based on the accounts of workers and observers, it could suffer from type 1 statistical error: Even if police were present at the protest, it would only be recorded if a worker or an observer makes a note of it. Thus, the frequency of police presence could be underestimated. Regardless, the data can be useful for comparing local governments’ responses to different types of protests, unless there is a reason to believe that the direction of the statistical error is correlated with the type of protest. For example, police presence might only be recorded when the protest is highly salient and receives high levels of attention from multiple sources over an extended period of time. In that case, police are more likely to appear at protests at well-established firms. However, there are not sufficient data to support interpretation. Protests at state-owned firms, for instance, were rarely high-profile, but were more likely to register a police presence in the dataset. Since the majority of cases in the dataset originate from workers’ own posts, the selection bias should at least be smaller than those protest cases screened and selected by the traditional media outlets. Out of the 3,339 instances of labor protests, police presence was recorded at about 31.8 percent of the incidents. Table 6.7 shows the percentage of the protests where police presence was registered, according to whether the protest was law- or interestbased. The data show that labor protests for an extra-legal cause were the least likely to invite police presence. In about a quarter of labor protests for extra-legal demands, a police presence was recorded. After controlling for the size of the protest, the difference becomes even more pronounced, since large-scale protests are more likely to have extra-legal demands and draw the attention of the local police. As shown in Figure 6.3, a protest for an extra-legal demand is 14.0 percentage points less likely to attract a police presence than a protest for a legal demand. A larger protest, as expected, is more likely to record police presence.

180

Interest-Based Collective Action and Firm-Level Patterns

extra−legal=1 extra−legal=2 size foreign taiwanhk joint state unknown −.2

−.1

0 Effects on Pr(Police)

.1

.2

FIGURE 6.3 Effects of protest demands and size on police presence Note: The baseline conditions are a protest of fewer than 100 participants, legal demands (extra-legal = 0), and domestic private firms. The reported effects are average marginal effects based on logit regressions with 95 percent confidence intervals.

The results confirm the intuition that policing will be selective under atomized incorporation. When protests do not expand beyond the factory gate, police are less likely to attend. When workers leave the factory compound to protest, police are more likely to intervene. Qualitative evidence also supports this finding. In my interviews with firm managers, they confirmed that local state authorities are often indifferent to protests within the factory compound, if workers do not use violence or highly disruptive tactics. A manager at a small-scale LCD factory in Guangzhou, for instance, said the local labor bureau and police would not usually show any interest in a labor strike at his firm and, in many cases, the state authorities would not even know it was happening. Even when the labor bureau learns about the protest, he said, the local authorities would probably call the firm to ask about the situation rather than actively intervene.37 This could be because his firm was relatively small with only 137 employees and thus not likely to disrupt the social order. 37

Interview 11-2012.

6.4 Firm-Level Patterns of Blame Attribution

181

Another firm manager at a clothing factory in Guangzhou largely concurred and said that when her factory experienced labor strikes she did not let the local authorities know, as they would not show up “unless workers break something or beat someone.” This was based on her previous experience, when workers powered down the electricity in the factory building and asked for an increase in the piece rate. When she called the police, “they showed up but nothing got resolved, because they tried not to intervene.”38 Observations by workers and labor activists and local policy documents confirm these attitudes of local government authorities toward labor strikes within a single workplace. During a labor strike in a Taiwanese-owned electronics factory, the company called the local authorities. When the police and security officials examined the situation they said, “There is no trouble; it is an orderly strike.”39 A policy document on labor mass incidents by a district government in Wenzhou states that the district police should get involved to maintain the public order when a large or extremely large (zhongda or teda) incident takes place. Those protests with fewer than 1,000 participants or those that block traffic for less than four hours are classified as “relatively large” (jiaoda) incidents that only require intervention by street-level personnel.40 These observations could explain why police presence is more likely to be recorded at large protests and at those using disruptive tactics. Although interest-based protests are generally larger in size, they could be regarded as less disruptive than law-based protests, since they tend to take place within the bounds of the workplace. 6.4 FIRM-LEVEL PATTERNS OF BLAME ATTRIBUTION

The tactics employed might not reveal the full picture about whether interest-based collective action could escalate and evolve into more politicized activism. In Chapter 4, I examined individual-level differences in blame attribution and found that line supervisors – who are in a more privileged position than rank-and-file workers – are more likely to adopt internal attribution of blame. These gate-keepers of collective action within individual workplaces are much less inclined to blame the central government for migrant workers’ plight than rank-and-file workers. 38 39 40

Interview 03-2012. Hao, 2016, p. 139. Ouhai District, 2012.

182

Interest-Based Collective Action and Firm-Level Patterns

In this section, I examine the political implications of the firm-level patterns by analyzing workers’ blame attribution at law-abiding firms. One could hypothesize that workers at law-abiding firms are more supportive of the central government because they experience the relatively better labor protection provided by the state law. These workers might also attribute less blame to the local government if they believe their firm’s compliance with labor laws is due to enforcement by the local government. The same argument could be applied to firms, since firms are the ones that spend resources to comply with the laws. Working at relatively more compliant firms, these workers might be more inclined to blame themselves for their difficulties. If these hypotheses hold, then even if interest-based collective action takes place at law-abiding firms, it would not undermine the logic of atomized incorporation, since workers at those firms are not necessarily interested in directing their discontent at the central government. Statistical analyses confirm that workers at law-abiding firms are substantially less likely to blame the central government. As demonstrated in Figure 6.4, the effect of law score on the direction of blame attribution shows that those at law-abiding firms are more prone to blame internally than externally, although the only statistically significant difference was the effect on blame attribution to the central government. An increase in the law score by one standard deviation would mean a decrease in the likelihood of blaming the central government by roughly 6.2 percentage points. However, firm-level legal protection did not significantly affect workers’ attribution of blame to the local government or firms. The finding mirrors the argument in the literature on hierarchical trust, that policy success is often attributed to the central government rather than the local government.41 The substantial and significant effect of law score on reducing attribution of blame to the central government might mean that the state’s better enforcement of labor laws could help the central government to avoid blame. It does not necessarily help firms to avoid blame, however, probably because legal rights protection per se does not induce satisfaction with their wages and working conditions among workers. This shows that atomized protests could have different political implications depending on the nature of the demands. On the one hand, workers who engage in law-based collective action are more likely to target the local government and make their protest relevant to political authorities; they are also more likely to blame the central government compared 41

Lü, 2014.

6.4 Firm-Level Patterns of Blame Attribution

−0.6

−0.4

−0.2 blame-cent blame-firm

0

0.2

183

0.4

blame-local blame-self

FIGURE 6.4 Effect of law score on blame attribution Note: The legend shows the four dependent variables. The independent variable in four models is law score. The x-axis reports the average marginal effects of law score based on logistics regressions with firm-level random effects. Basic demographic characteristics are controlled.

to those engaging in interest-based collective action. Yet local governments have access to institutional channels to resolve these conflicts and break down the protest before it escalates. Interest-based collective action, on the other hand, takes place at firms where workers are less likely to blame the central government; the protest tactics also are less likely to be disruptive or to target the local government. Interest-based protests are more likely to be treated as “economic” protests that should be resolved by individual firms. Local governments have limited ways to intervene to resolve the conflict, as the state’s tools are not designed for this type of contention. This means that interest-based collective action is more likely to be apolitical but could mean larger uncertainty. I continue to investigate these political implications in Chapter 7 by probing what kind of mobilizational resources enable interest-based collective action, and

184

Interest-Based Collective Action and Firm-Level Patterns

whether these resources could contribute to the emergence of politicized labor activism. 6.5 CONCLUSION

In this chapter, I examined firm-level patterns of labor protests, with a focus on understanding where interest-based collective action takes place and how it differs from law-based collective action. I find that interestbased collective action – even though it appears to be more proactive in nature – does not appear to pose a more substantial threat to the regime than law-based protests. First, law-abiding firms such as foreigninvested and large-scale firms are more likely to experience interest-based collective action than firms with illegal labor practices but – due to their relatively privileged position – workers at law-abiding firms are substantially less likely to blame the central government for their difficulties. Second, interest-based collective action rarely employs disruptive tactics that enlarge the physical boundary of the conflict, because protesters seeking extra-legal gains have little to gain by drawing the local government into the conflict. Instead, interest-based protests use labor strikes to put pressure on firms. The rise of interest-based protests demonstrates that migrant workers can no longer be characterized as rightful resisters who force companies to meet their state-mandated obligations. This might be a reflection of a change in state – society relations in China more broadly. A study of contention among urban citizens, for instance, shows that citizens are increasingly willing to protest for a “rational” outcome even when it goes beyond the legally defined minimum.42 Despite the fact that the regime does not have an institutionalized mechanism to resolve disputes over interest-based demands, the findings in this chapter show that interestbased collective action alone does not seem to threaten the regime’s political resilience, at least in the short run. Protests over legal demands are more likely to target state authorities and escalate into disruptive actions, but the regime has institutionalized means to break down lawbased collective action via courts, trade unions, and legal aid services.43 This is what Chen and Gallagher (2018) call “procedural fix.” The way the regime deals with atomized protests demonstrates the logic of atomized incorporation: During interest-based collective action, the state does not necessarily intervene with police as workers do not 42 43

Zhang, 2014. Su and He, 2010; Chen and Gallagher, 2018; Xu, 2013.

6.5 Conclusion

185

interact with state actors as much as they do when defending their legal rights. This suggests that due to the market reform, the state is now better able to distance itself from “economic” disputes. During the Mao era, Zhou (1993) argues the socialist state’s penetration into social organization contributed to the “large numbers” phenomenon. While citizens have limited capacity to organize their interests, a large number of individuals often engage in the same patterns of resistance. The target of their resistance is usually the state, due to the overlap between the public and private spheres. While it is easier to escalate protests for legal rights protection into more disruptive tactics – because protesters believe that the local authorities have an obligation to meet their demands – interest-based collective action could remain a private matter between employers and employees. The analyses in this chapter alone, however, do not fully reveal the kind of information provided by interest-based collective action. In Chapter 7, I discuss why law-abiding firms are more prone to experience interest-based collective action and what this implies for atomized incorporation.

7 Discursive Opportunities and Collective Action at Law-Abiding Firms

While workers can use a collective labor strike to create leverage, merely initiating collective action does not guarantee its success. Individuals are keen to maximize the probability of success of any protest they initiate, and their decision to protest also depends on their ex ante perception of their group efficacy. In this chapter, I continue the firm-level analyses of labor protests and look into how workers use collective action to advance interest-based claims at law-abiding firms. When workers know that they cannot engage state actors to support their interest-based claims, what resources do they have to enhance their bargaining power, and how can they assess the strength of these resources? I focus on firm-level distribution of discursive resources – the resources for creating discursive opportunities – with the understanding that workers with limited bargaining power rely on third-party actors for support. In Lipsky’s language, the “reference publics” who exercise a significant influence over protest targets enhance the chance of success.1 Chinese protesters have also made good use of “publicity tactics” to gain the attention and sympathy of potential allies.2 While more direct and disruptive tactics could generate concessions, they are the last resort of protesters because of the risk involved.3 When publicity tactics can be used effectively, protesters do not need to adopt more costly strategies. Distribution of discursive resources at the firm level is an important determinant of workers’ ex ante perception of group efficacy. A map 1 2

3

Lipsky, 1968, p. 1146. Chen (2009) classifies petitioners’ troublemaking tactics into four categories: elite advocacy, publicity, persuasion, and disruption. The study finds that publicity and disruptive tactics are effective in generating positive responses from the local government. Cai, 2008.

186

7.1 Media Reports of Labor Protests

187

of protest incidence onto discursive resources shows that workers at law-abiding firms – which are more often the targets of interest-based collective action – also perceive that they have more discursive resources. During interest-based collective action, workers use the threat of reputation damage to push firms to respond and media is a major channel for creating such pressure. Collective action during the China-Japan disputes over the Diaoyu/Senkaku islands shows how workers at many Japanese-owned firms used the international dispute strategically to create discursive opportunities. This shows that, for the state, the negative political implications of atomized incorporation are more likely to be realized when workers’ access to discursive resources improves. 7.1 MEDIA REPORTS OF LABOR PROTESTS

Scholarship on protest has long demonstrated that media interest has a substantial effect on the success of a protest.4 Media has been postulated as one of the most important channels for protesters to create discursive opportunities, “the aspects of the public discourse that determine a message’s chance of diffusion in the public sphere.”5 It is well documented that Chinese protesters often contact media companies before and during collective action to gain publicity.6 Contrary to the assumption that media is simply a propaganda tool controlled by the regime, a growing body of literature suggests that authoritarian regimes do allow some amount of reporting on relatively sensitive topics,7 although media reports on mass incidents are still quite limited. King et al. (2013), for example, show that while seemingly sensitive online postings in China – such as criticism of government officials – remain uncensored, the events that have “collective action potential” are highly censored. Chen and Xu (2012) find that petitioners rarely gain access to the media because journalists do not cover mass incidents. Despite the restrictions on journalistic freedom, however, labor protests in the 2010s received substantial media attention. Scholars and labor activists note that the Chinese regime became more tolerant about reporting collective action by migrant workers from the mid-2000s.8 4 5 6 7 8

Lipsky, 1968; Gamson and Wolfsfeld, 1993; Koopmans, 2004a. Koopmans and Olzak, 2004, p. 202. Zhao and Wusan, 2009; Cai, 2010. Schatz, 2009; Egorov et al., 2009; Stockmann, 2012; King et al., 2013; Lorentzen, 2014; Repnikova, 2017b. China Labour Bulletin, 2012; Elfstrom and Kuruvilla, 2014.

188

Discursive Opportunities and Collective Action

According to a report by CLB, while only 8.3 percent of labor protests in its database were reported between 2000 and 2007, roughly 57.4 percent of the cases between 2007 and 2010 were covered by official media.9 There was some speculation that when then-Prime Minister Wen Jiabao publicly denounced the widespread problem of migrant workers’ wage arrears in 2003, it changed in the way these issues were covered. Domestic media started to carry stories of migrant workers’ contention around wage arrears. Increased media coverage of labor protests could reflect a stronger effort by the regime to shape public opinion amid rising social tensions. Beginning in the 1990s, as China dealt with the negative consequences of marketization and economic integration in the post-reform era, the CCP started to take advantage of media as a tool for “public opinion supervision” (yulun jiandu).10 It learned that uncovering social problems at the local level not only revealed information about the behavior of local government officials but could also be used to shape public opinion in favor of the regime by helping the central government to claim the moral high ground.11 A 2005 policy document of the Shenzhen municipal government states that while media reports on mass incidents should not be allowed in principle, the local government can arrange for reporters to visit protest scenes to create a “good environment for public opinion.”12 The space for critical media reports, however, has been subject to intensive monitoring and management by the party state.13 In the case of labor protests, journalists have not been able to report certain types of labor disputes that are deemed highly sensitive. For instance, a journalist from a news company in Guangdong Province said the central government regards any labor dispute at state-owned enterprises to be politically sensitive.14 As a result, media companies rarely produce investigative reports on labor issues at state-owned enterprises. The central government also restricts media reporting when a protest persists and has the potential for a large social impact. When 40,000 workers at the Yue Yuen shoe factory in Dongguan staged a strike in 2014 media companies, including foreign outlets such as the BBC, reported from the protest scene. Due to the large 9 10 11 12 13 14

China Labour Bulletin, 2012. Zhao and Wusan (2007) defines yulun jiandu as “the use of critical media reports to supervise government officials.” Zhao and Wusan, 2007; Lorentzen, 2014. Shenzhen Emergency Command Center, 2005. Repnikova, 2017a; Roberts, 2018. Interview 06-2013.

7.1 Media Reports of Labor Protests

189

size of the protest, it might have been judged impossible to contain the spread of information. Even the English channel of the CCTV reported on the protest in its news program China 24, two days after the BBC aired its coverage. But as the protest dragged on, the propaganda department issued guidance to censor reports about the protest, and Chinese-language posts about the protest were removed. While the central government might tolerate some reports on labor protests for media oversight, local governments consider journalists reporting on mass incidents in their jurisdiction to be undermining them. Local governments often exercise their influence over the local propaganda department to block news coverage of mass incidents in their jurisdiction, including labor protests. My interviews and past research have uncovered ways that journalists try to circumvent the censorship. For instance, they can employ “reporting in a different jurisdiction” (yidi baodao) to avoid local censorship. Journalists reporting on a protest in a different jurisdiction have more freedom, since the local propaganda department cannot directly control media companies from a different region.15 While yidi baodao was officially banned by the central government in 2005, it seems to have continued in practice.16 When a labor protest occurs at the township or city level, journalists from a provincial media outlet also could enjoy leverage in reporting the incident.17 Studies find that some cities such as Guangzhou provide a better environment for investigative reporting due to the fragmented political structure.18 Finding ways to circumvent local censorship, however, incurs costs for journalists. This means that a labor protest will have to reach a threshold of newsworthiness before journalists will consider it to be worth covering. As the editor-in-chief of the magazine Caijing explains, journalists in China “calculate an issue’s importance and then weight it against the risks

15

16 17

18

Media companies operating in China need a government sponsor, whose administrative level determines the media companies’ administrative rank. The government sponsor has direct control over the media companies’ personnel appointments and dismissals (Stockmann, 2012). Interview 03-2014. For instance, the Nanfang Daily Group, controlled by the Guangdong Communist Party Committee, is better able to cover labor protests in Shenzhen than the Shenzhen Newspaper Group, because Guangdong Province is at a higher administrative level. Studies find that journalists in one locality sometimes ask colleagues in a different location or who answer to a higher level government to report on politically sensitive topics. See Shirk (2010). Lei, 2017, Chapter 4.

190

Discursive Opportunities and Collective Action

of reporting.”19 An issue’s importance can be evaluated in many different ways, but the most critical standard will be whether it is of interest to their readers. Past scholarship on media attention in liberal democracies suggests that media companies and journalists show a selection bias in determining whether or how to cover protest events. Due to journalistic standards and news companies’ pursuit of profits, news reports tend to be skewed toward the protests that interest a large number of readers.20 For instance, studies have consistently found that a large-scale protest is more likely to receive media coverage.21 Others have found that journalists are more likely to cover protests that last a long time or involve violence.22 The existing literature on media commercialization in China highlights that Chinese journalists compete for ratings.23 Media commercialization and partial privatization since the 1990s have incentivized news companies to secure their own financial resources. Competition for advertisers and readers has intensified over time and pushed journalists to find stories with high news value. Some scholars suggest that market incentives create conflicts of interest for investigative journalists. On the one hand, the public wants to “get the truth” about major social incidents and expects critical reporting. On the other hand, the state emphasizes “positive spin” on the news stories and prevents journalists from fully engaging in investigative journalism.24 My interviews with a former editor of a major magazine in China reveal that some journalists are willing to push the limit of state censorship to pursue stories with high news value, including workers’ collective action.25 However, not all labor protests are of interest to them. The interactive effect of media censorship and commercialization means that some protests have a better chance of receiving media attention than others, due to their higher news value. 7.1.1 Media Exposure at Law-Abiding Firms The discussion suggests that those workers engaging in “newsworthy” protests are better able to attract attention from media outlets and potentially from relevant third-party allies. So are there firm-level structural 19 20 21 22 23 24 25

Hu, 2010, p. 86. Andrews and Caren, 2010. Snyder and Kelly, 1977; McCarthy et al., 1996; Oliver and Myers, 1999. Snyder and Kelly, 1977; McCarthy et al., 1996; Barranco and Wisler, 1999; Oliver and Myers, 1999. Shirk, 2007; Shirk, 2010; Liebman, 2010; Stockmann and Gallagher, 2011; Stockmann, 2012. Stockmann, 2012, p. 98. Interview 05-2014.

7.1 Media Reports of Labor Protests

191

factors that determine a protest’s newsworthiness? Following the discussion in Chapter 6, we can also ask whether labor protests at law-abiding firms would be considered more newsworthy. To answer these questions, we must discuss why firms abide by labor laws in the first place when there is limited top-down enforcement. As noted in Chapter 3, local authorities responsible for enforcing labor laws lack the financial and personnel resources to monitor effectively. In 2011, for example, the number of officials working in the labor inspection bureau was only 23,000, meaning one inspector for every 20,000 workers.26 Local inspection bureaus conduct regular and random audits but due to their limited capacity, many firms are not exposed to top-down pressure.27 In interviews, factory managers confirmed that firms abide by labor laws mostly to protect their reputation. The costs of being labelled as a “sweatshop” factory can be substantial even if the likelihood of getting caught is small. The companies that produce internationally renowned brands, for instance, are highly concerned about their reputation and do not want to tarnish it by being caught violating the local laws. An HR manager at a packaging material factory in Zhongshan stated during an interview that the company had always abided by the local labor laws, not because they were afraid of punishment from the local government, but because they were concerned about their reputation.28 Ninety percent of their buyers were multinational companies that are internationally recognized, and the buyers’ concerns about reputation automatically trickle down to the subcontracting factory. The factory manager emphasized that one of their core business values was social responsibility. The factory was externally audited four to six times a year and established their own programs for social responsibility, such as building schools for workers’ home villages. Concerns for reputation and the image they were presenting as a socially responsible factory created an incentive to comply with the laws. “Even if our customers might not immediately know of the violations,” the manager said, “we will have to pay a high cost once they do.” This means that workers at subcontracting factories that produce goods for internationally renowned firms could perceive a higher chance, and higher cost, of media exposure. Foreign-owned firms might also perceive that their risk of being exposed is high. They are profitable and sufficiently well established to 26 27 28

Yang, 2011. Ngok and Zhuang, 2009. Interview 09-2013.

192

Discursive Opportunities and Collective Action

invest overseas and thus often visible to the public. Any bad publicity could also stir up nationalist sentiments among Chinese consumers. After the Foxconn suicide scandal in 2010, for example, the head office at Samsung Electronics wrote to all of the factories they contracted production to, instructing them to decrease overtime hours to the legal limit of 36 hours per month. According to an HR manager at Samsung Electronics in Guangdong, the Foxconn incident was a wake-up call for all foreign-invested firms in China. “After the Foxconn incident,” he said, “the headquarters really understood how risky it was to have any illegal labor practices even if every other firm does the same. Especially as a foreign-owned firm, we do not want to create the perception that a well-known Korean company exploits Chinese workers.”29 Since the majority of factories in China are not operating under the 36-hour limit, the company’s decision reflects its strong commitment to reduce risk.30 The discussion highlights that media reports of labor protests at wellknown firms – or those that produce for well-known firms – could create bad publicity. Even when the firm abides by labor laws, they might still be labelled as “sweatshop” factories if workers are protesting for higher wages. The suicide incidents at the Foxconn factory in Shenzhen were widely reported in the domestic and international media in 2010 and led to the denouncement of the factory’s major business partners, including Apple, HP, and Dell. Media reports on labor protests at wellestablished firms also draw the attention of nonprofit labor organizations and activists, who can harm the firm’s reputation by increasing customers’ awareness of labor conditions in factories. For protesters who make extra-legal demands, publicity tactics are particularly important because they can target the actors who have direct leverage over the firm’s policy decisions. There are workers who have a clear understanding of their firm’s position in the global economy and its impact on their protest leverage and link their protest to the lead firm of the global supply chain or the headquarters. At an American-owned factory that produced agricultural machinery, workers used social media to urge each other to contact their American headquarters via Google and Facebook. Since the firm was a Fortune Global 500 Company, according to one worker, the headquarters “would not let the mess continue for the 29 30

Interview 08-2012. The overtime hour limit is perhaps one major exception that local governments implicitly allow as an illegal labor practice. According to a firm manager of a small clothing company in Shenzhen, after the Labor Contract Law was first implemented, the local officials circulated a small leaflet on the new law. When it came to the overtime hours, however, the local officials said they could offer more leeway (Interview 01-2012).

7.2 Perception of Media Exposure at Different Firms

193

sake of their brand image.”31 Even when firms do not produce consumer goods such as cell phones or computers, their business partners might be concerned about bad publicity adversely affecting their own reputation for social responsibility. Firms listed on the stock market could also be sensitive about any information about unstable labor relations affecting their share price or reputation among shareholders. Likewise, firms that make corporate social responsibility part of their brand make themselves more vulnerable to pressure from workers undertaking collective action: Workers at these firms might perceive that the threat of bad publicity could generate a response from the firm management and enhance the chance of success. The very factors that make firms abide by labor laws, therefore, could enhance workers’ ex ante perception that their protest would receive media attention and generate success. 7.2 PERCEPTION OF MEDIA EXPOSURE AT DIFFERENT FIRMS

7.2.1 Group Efficacy at Law-Abiding Firms Are discursive resources important for bolstering workers’ ex ante perception of group efficacy? Do workers at different types of firms have different perceptions of their access to discursive resources during collective action? To investigate the relationships between workers’ sense of group efficacy, expectation of media coverage, and firm-level variables, I first measure whether workers at law-abiding firms show stronger group efficacy by estimating the effect of law score on group efficacy. The estimation results based on logit regressions are displayed in Figure 7.1. As expected, after controlling for demographic variables, the effect of law score on group efficacy was positive and statistically significant. Workers at fully law-abiding firms – where the law score was 100 – were 28.3 percentage points more likely to indicate that their protest would succeed than those at the firms where the law score was 0. The effect was larger and more statistically significant for line supervisors compared to rankand-file workers. The effect of firm-level factors could be more salient to line supervisors, who are more experienced and thus more informed about the environment in which their firm is situated. 7.2.2 Group Efficacy and Media Exposure I measured workers’ expectations about media exposure by asking respondents whether media would report their strike incident if a labor 31

The information is from protesting workers’ online posts on Weibo. For the original writings, see http://wickedonna2.tumblr.com/post/83071564817/4-17 (last checked on January 15, 2020).

Discursive Opportunities and Collective Action

194

−0.2

0

0.2

All workers

0.4

0.6

Line supervisors

Rank-and-file workers FIGURE

7.1 Effect of law score on perceived group efficacy

strike took place at the current firm. A binary variable media exposure denotes whether the respondent expects media exposure to be likely or not. About 50.7 percent of respondents replied that media exposure was likely. Only a small fraction of workers (4.4 percent), however, were certain about media exposure; a larger number of respondents replied positively but with uncertainty (46.3 percent). Statistical analyses confirm that workers’ expectations about media exposure increase their perception of group efficacy, particularly among line supervisors. As displayed in Figure 7.2, when we consider the entire sample, the expectation that a labor protest would be reported by the media increased perceived group efficacy by 12.2 percentage points. The effect increased to 16.7 percentage points for line supervisors, which was substantially stronger than among rank-and-file workers (6.1 percentage points). The results again demonstrate that line supervisors are more likely to connect media reports to the success of collective action. While the data cannot establish the direction of the causal relationship, it is unlikely that the analysis suffers from endogeneity. It would be difficult

7.2 Perception of Media Exposure at Different Firms

−0.1

0 All workers

0.1

0.2

195

0.3

Line supervisors

Rank-and-file workers FIGURE

7.2 Marginal effect of expected media exposure on perceived group

efficacy

for journalists to select only the successful protests, because they generally report the protest incident at the onset and do not follow up on outcomes.32 7.2.3 Marginal Effect of Firm-Level Variables on Expected Media Coverage Finally, I estimate the effect of firm-level variables on expected media exposure. Table 7.1 shows the average marginal effect of firm-level variables on media exposure based on logit regressions. As expected, law score substantially increased workers’ expectations of media exposure (see Model 4). The effect, however, was largely driven by line supervisors (see Figure 7.3). When I confined the sample to line supervisors, 32

Scholars such as Chan (2001) suggest that media reports play a limited role in bringing about long-term changes to labor conditions, since journalists in China do not pay attention to the final outcome of labor disputes.

196

Discursive Opportunities and Collective Action TABLE 7.1 Effects of firm ownership and size on workers’ perceived chance of media attention

Variable

Model 1

Model 2

Model 3

Foreign

0.135∗∗ (0.059)

0.055 (0.072) 0.078∗∗ (0.034)

0.027 (0.072) 0.063∗ (0.035) 0.128∗∗ (0.065)

Firm size Foreign brand Law score N = 603

N = 602

N = 591

Model 4

0.255∗∗ (0.164) N = 603

Note: ∗∗∗ p < 0.01, ∗∗ p < 0.05, ∗ p < 0.10. Basic demographic controls are included in the models. Average marginal effects are based on logit regressions with firm-level random effects.

law score increased the perception of media exposure by 38.6 percentage points. For rank-and-file workers, the effect is positive but weaker and not statistically significant. The results show that line supervisors are more likely to form their judgement about available discursive resources based on their firm’s socioeconomic environment. The two firm-level variables that are correlated with a higher law score – foreign ownership and firm size – also had a substantial effect on media exposure, although the effect of firm size was larger. Foreign ownership increased workers’ expectations of media exposure substantially, by 13.5 percentage points (Model 1). Since foreign firms are on average larger than domestic firms, the effect of foreign ownership decreased after controlling for firm size and lost its statistical significance (Model 2). The results generally support the observational intuition that foreign-invested and large-scale firms receive more attention from the media. The positive effect of firm size was strikingly large: When controlling for foreign ownership, if the logged firm size increased from minimum (2.89) to maximum (9.28), the respondents’ expectation of media exposure rose by nearly 50 percentage points (Model 2). Firms’ newsworthiness could also be measured by the visibility of their products. The survey asked whether their firm produced products for brands with an international reputation. Roughly 64.6 percent of workers replied positively, demonstrating the prevalence of firms engaging in global production in Southern China. As shown in Figure 7.4, the ratio

7.2 Perception of Media Exposure at Different Firms

−0.2

0 Line supervisors

0.2

0.4

197

0.6

Rank-and-file workers

FIGURE 7.3 Marginal effect of law score on expected media exposure by job position Note: average marginal effects are based on logit regression analyses controlling for age, gender, income, wage gap, years of migration, firm tenure, job position, and firm profit with firm-level random effects.

was significantly larger for foreign-owned enterprises, but almost half of workers (45.1 percent) at domestic firms also reported that their firm produced internationally reputable brands. Based on the answer, I created a binary variable foreign brand that denotes the firm’s production of internationally recognized brands. As expected, foreign brand had a large and significant effect on workers’ expectations of media exposure (Model 3 in Table 7.1). Even after controlling for firm ownership and size, foreign brand increased the number of workers who expected media exposure by 12.8 percentage points. Regardless of firm ownership, being part of a global value chain seems to generate discursive resources for workers when there are recognizable firms in the chain. Studies find that subcontracting relationships are

Discursive Opportunities and Collective Action

0.7 0.6 0.5 0.4 0.2

0.3

Percentage of respondents

0.8

0.9

198

Domestic firms FIGURE

Foreign firms

7.4 Does your firm produce internationally renowned brands?

susceptible to a “race to the bottom” in labor conditions, because subcontracting firms compete with each other to offer cheaper labor costs.33 Yet, the results indicate that if a firm is part of a global value chain, workers can engage with third-party allies as part of their collective action as long as the allies – such as the lead firm and consumers – are concerned about labor conditions at subcontracting facilities. 7.3 CASES OF LABOR PROTESTS

I present two case studies of labor protests to demonstrate how workers engage in interest-based collective action, and their interactions with third-party actors. The two protests illustrate the similarities and differences between interest-based and law-based collective action. The first case study is an interest-based protest at an American-owned packaging material factory in Guangzhou. The protesters mainly interacted with media companies and personnel from the Asian headquarters to enhance their bargaining power, without targeting state authorities. The second 33

Mosley and Uno, 2007.

7.3 Cases of Labor Protests

199

case study, of a Hong Kong-owned printer factory, was not a clear case of interest-based collective action. The protesters used the legal ambiguities to pressure the local government for intervention and mainly relied on disruptive tactics. In both cases, protesters sought media attention, but in the latter protest, there was limited media attention and workers mainly focused on pressuring the local government. 7.3.1 A Labor Strike at a Paper Factory in Guangzhou In February 2013, there was a labor strike at an American-owned factory in Guangzhou that produced packaging materials. Workers returned to the factory after the Spring Festival to find that their year-end bonus had decreased substantially, without a clear explanation from management: management only said that the firm did not make enough profit. Workers were supposed to receive the year-end bonuses on the 27th of January, but the factory postponed the payment. According to a worker, it was the factory’s deliberate strategy to avoid disputes before the holidays and safeguard the production schedule. Another subsidiary factory of the same company had gone on strike before the Spring Festival over the same issue, and the incident made the firm management realize that they should delay the announcement about reduced bonuses until the end of the Spring Festival holidays. While the main demand was to increase the year-end bonuses, workers expressed widespread grievances about declining welfare benefits since the firm’s takeover in 2010. Before the takeover, workers received double salary (shuangxin) as the year-end bonus, which was equivalent to double the average monthly wage. After the takeover, the firm still provided up to 2000 yuan as the year-end bonus, but in 2013, the bonus dropped to 600 or 700 yuan. The factory had written a letter of agreement at the point of takeover and promised that welfare benefits would be maintained, but workers experienced deterioration in conditions with respect to protective gear, labor intensity, and year-end bonuses. While the factory paid the local minimum wage, a large number of workers at the factory were relatively skilled, with multiple years of experience. The baseline wage had not changed since 2007 when the factory decided to adopt a performance-based salary system. The strike was initiated by line managers at the factory. Line leaders told workers not to enter the factory the next day and to report in front of the factory gate. One of the participants said that their line leaders lived outside the factory compound and in proximity to one another. They were

200

Discursive Opportunities and Collective Action

on very good terms, which made it easier for the line leaders to initiate the strike of more than 200 workers at the factory. Workers learned via social media that the other subsidiary factory had staged a strike and succeeded in receiving a 60 percent higher year-end bonus. Many workers said that because of the other strike, they had high levels of confidence that their strike would succeed. The factory’s headquarters in the United States belonged to Fortune 100 companies and the factory manager took great pride in the firm’s reputation as a socially responsible firm. When I interviewed the firm manager in 2012, he stated that the headquarters was considered to be one of the most ethical businesses globally and had been selected as one of the most admired companies. The company owned twenty-two subsidiary factories around the world in the United States, Latin America, Europe, China, and Africa. In China alone, the firm employed 9,000 workers and enjoyed a good reputation as an environmental-friendly company. According to the firm manager, their major customers were well-established international businesses that would pay more to ensure socially responsible production. Labor costs increased after the Labor Contract Law, according to the manager, but their products did not lose their competitive edge as their competitors were also well-established factories that cared about reputation. The factory, however, was trying to introduce automation to lower costs. His remarks also mirrored workers’ grievances about declining welfare benefits. For instance, he said that the factory tried to enhance efficiency by adopting a 3-4-5 approach to human resources: by hiring three workers, you complete the work of five but pay the salary of four. The factory was paying more to workers compared to other factories, but labor intensity grew without a proportional increase in salary. Out of concern for their reputation, the factory managers made sure that their labor practices did not violate the labor laws. The firm’s international reputation appears to have contributed to workers’ strategic decision to initiate the strike. In addition to the other factory’s experience of success, some workers mentioned that their factory had a lot to lose, due to its reputation as a model (guifan) firm. One worker said that he did not consider the risk of getting fired or repressed because his factory was a guifan firm that was expected to behave well. Strike initiators contacted media companies before the strike began, and on the first day of the strike, the Guangzhou TV station was present to air images of the strike. It was not clear whether the media company had confirmed their presence before the strike, but the timing of the report

7.3 Cases of Labor Protests

201

suggests that the media company considered the labor strike highly newsworthy; it did not appear that workers spontaneously contacted the media after the collective action began, but rather that the workers had chosen to invite the media before the strike started. Although workers had relatively high levels of confidence due to the information from the other strike, they still wanted media support to enhance their bargaining position. Workers knew that the strike would draw attention from external actors, and used it as a tool. Strike participants did not seem to care whether their demands were legally supported. One of the most common words that workers used to describe their discontent was “unreasonable” (buheli). There was substantial variation in their understanding of why the firm decreased the year-end bonuses. Some suggested that while the firm was not profitable the previous year, it still had an obligation to maintain the level of welfare benefits. Others stated that their factory might have lost profits, but the headquarters still made two million dollars in profits and would be capable of paying workers more. One worker told me that he wanted to contact the US headquarters but could not do so due to the language barrier. The discussion showed that at least some workers were well aware of their factory’s position in the larger corporate structure and how they could leverage it to their advantage. The labor bureau visited the firm on the first day of the strike but did not make any comment. Later that year, when I returned to the factory to interview the head of the factory’s trade union, she told me that due to the media coverage, the factory was pressured by the local labor bureau and the local trade union to resolve the issue quickly. The districtlevel government, she said, was anxious to resolve the conflict because “the government officials disliked media exposure.” However, she also stated that the local government could not do anything to resolve the issue because there was nothing illegal about reducing the amount of the year-end bonuses: “I showed them our employee handbook that clearly said welfare benefits could change depending on the firm’s profits. We did nothing wrong and had no reason to listen to the government.” For the firm management, more significant pressure came from the firm’s Asian headquarters in Shanghai. After the strike was reported, the headquarters dispatched a dispute resolution team from Shanghai, a sign that the headquarters was taking the issue seriously. On the second day of the strike, the team met with workers and persuaded them to go back to work. According to some workers, the Shanghai team was hinting that if the workers were to go back to work immediately, they would not

202

Discursive Opportunities and Collective Action

hold the workers responsible, but if workers continued with the strike there would be repercussions. The threat was effective and the strike ended on the second day. According to the firm manager, the headquarters’ intervention might have been intended to prevent chain effects. Since the corporation had many factories in China, they had to demonstrate a tough stance to discourage future disputes. Some workers suggested that the firm management provided financial incentives to line leaders and made them go back to work. Although the strike did not succeed, in this protest it was the corporation’s headquarters, not the local government, that had influence over the factory management. The workers did not pursue any typical strategies to appeal to the local government. They did not visit the labor bureau or the petitioning bureau but relied mostly on publicity tactics. Their discussion about strategies for enhancing the chance of success mostly focused on utilizing discursive resources and reaching those third-party allies who could pressure their firm. Thus, the strike did not physically expand beyond the factory gate. One of the major reasons why the headquarters’ visit was so effective in discouraging workers from sustaining their strike was the realization that, even after the media exposure and intervention by the headquarters, workers could not achieve success. Workers were left with few other options to enhance their bargaining power. The protest also showed that media was the main channel for engaging with third-party actors who had substantial leverage over the factory, such as the headquarters and buyers. The workers publicized their cause on social media but the posts did not influence the way the firm dealt with the protest. This might be because the factory did not produce consumer goods and its name was not instantly recognizable to the public. When I mentioned the strike, the head of the firm’s trade union immediately asked if I learned about the strike on the Guangzhou TV. Even if the strike ended fairly quickly, she said that the firm management went through immense anxiety and stress because they feared negative reactions from their buyers. When I asked how buyers usually learn about a protest, she said the main channel was, “of course, the media.” 7.3.2 A Labor Strike at a Printer Factory in Shenzhen The second case study is a labor strike that occurred at a printer factory employing about 4,000 workers in Shenzhen. The Hong Kong–owned factory was established in 1998 and had operated as an export-processing

7.3 Cases of Labor Protests

203

facility (a sanlaiyibu firm) of its parent company. Due to the government policy that all sanlaiyibu firms needed to re-register as a wholly owned firm, the factory initiated a restructure and changed its name and firm representative. According to workers, the rumors about the restructure started in November 2012, sparking discussions among the workers about the change. The striking workers’ main goal was tenure compensation (maiduan gongling). It was unclear whether the demand had a legal basis. When a firm terminates its labor contract with employees, it is legally obliged to pay compensation based on workers’ firm tenure and wage level. It is also stipulated that when firms change name or representative, the change should not affect the existing labor relations. The law is unclear about whether this clause can be applied to sanlaiyibu firms under restructuring. A restructure could be viewed as continuation of existing labor relations, as long as the newly registered firm takes over the old labor contracts. On the other hand, since the firm is newly registered as a different firm, this could be used as a cover for labor contract termination. Workers at many sanlaiyibu firms took this issue to the labor arbitration committees, requesting contract termination and the corresponding compensation. The dispute had become so prevalent in export-oriented regions that the Shenzhen municipal government issued a document, “Summary of the Seminar on Difficult Issues of Shenzhen Municipal Labor Contract Management” in 2012. The municipal government stated that workers’ demands for compensation should not be supported when sanlaiyibu firms restructure, as long as the firms roll over the labor contracts and ensure that there are no changes to existing rights and obligations. Due to legal ambiguities, however, many workers still protested in the expectation that they could receive tenure compensation. On the morning of the strike, workers submitted a demand letter with three main demands. First, the firm should pay tenure compensation. Second, the firm should provide social insurance to all workers. Third, the overtime pay should be re-calibrated, because skilled and unskilled workers were receiving the same overtime pay. The workers who wrote this letter were senior workers who had stayed in the firm for more than ten years, including line managers. The firm management did not respond to the letter. Before noon that day, workers started their strike. Workers did not have any formal representatives, but it was clear that the initiators and core participants had been the ones to write the letter. They had a large stake in tenure compensation because of their own long tenure. There were many workers who had worked in the firm for more than

204

Discursive Opportunities and Collective Action

five years and workers’ solidarity seemed to be quite high. Protesters on the strike scene estimated that around 80–90 percent of the workforce joined the strike. When asked why some did not participate, workers generally stated that those with a shorter firm tenure might not have a strong incentive to join. However, there were also many participants who did not know what they were demanding or who said that they did not care about the goal of the strike but still joined because they had nothing else to do. The strike also received support from the firm’s trade union. Protesters hung a banner on the factory gate that said, “Trade union supports our strike, Protect legal rights, Pay for workers’ social insurance, Pay reasonable overtime fees, Compensate workers’ tenure. Never give up, persist till the end!” The banner showed that, in publicizing their protest, workers prioritized their legal demands, although most discussions at the scene of the strike was about tenure compensation. Participants varied in their beliefs about whether their demand was legally supported or not. Some thought the firm was legally obliged to pay compensation even if it maintained the existing labor contracts, and some were concerned that a new labor contract would not correctly reflect their firm tenure. A firm that carried over the original contract might not need to pay compensation, but workers were skeptical that the firm would honor this commitment. To gain trust, the firm manager issued a letter of agreement that guaranteed the maintenance of existing labor relations, but workers were not sure if the letter would be legally binding. Participants seemed ready to pursue every strategy to make the firm terminate the contracts and pay compensation. The strike was definitely not a spontaneous reaction to the firm’s restructuring. Since the rumors about restructuring started to circulate in November 2012, workers had already had substantial time to discuss their strategies. Interviews revealed that two factors had largely contributed to their decision to protest. First, senior workers at the factory perceived a clear financial incentive in tenure compensation and were confident in their mobilizational capacity. One of the senior workers stated, “The stakes are too high to back down. As a peasant worker, this kind of opportunity to receive a large sum of money is rare. We all spent our young and healthy years at this firm, and we deserve the compensation.” Among senior workers, it was evident that their grievances were not merely about compensation per se: Years of migration had taken a toll on their health, youth, and life. Another senior worker said that “we knew every worker would feel the same way about this issue. We spent many years together and are aware of each other’s thoughts. It was not that

7.3 Cases of Labor Protests

205

some people forced others to join. Everyone voluntarily did so because they knew the firm’s dealing of the matter was wrong.” Protesters contacted media companies and the labor bureau before and during collective action. On the first day, protesters heard from a journalist at a local newspaper company that he would take a look at the case, but there was no media coverage until the fifth day of the strike. Workers had visited the labor bureau before they initiated the strike, but the bureau had not responded to them. The protesters’ main strategy was to pressure the local government. On the second day of the strike, an official from the local labor bureau visited the factory but only stated that the legal basis of tenure compensation was ambiguous. The statement seemed to have contributed to workers’ anger but also motivated them to confront the local government further. Workers marched to the local petitioning bureau and a large number of police were summoned to maintain order. There was no visible clash between protesters and police, and workers generally seemed to believe that police would not forcefully repress them, although the presence of police clearly created more anxiety at the protest scene. After hearing no response from the petitioning bureau, some workers went to the labor bureau but could not get a satisfactory answer there either. On the fourth day, workers marched to the local petitioning bureau again. A few workers kept shouting, “Do not go down to the traffic road,” to make workers stay on the pedestrian road. In the meantime, the firm management introduced schemes to discourage the collective action. On the first day, the firm said it would provide an additional bonus of 100 yuan for those who went back to work. The bonus increased to 300 yuan in the morning of the fourth day and to 800 yuan for some departments that afternoon. Some protesters went back to work, but those who remained on strike said that only those workers on good terms with the firm managers went back to production. The strike ended on the fifth day, after the labor bureau officials and trade union cadres visited the factory and mediated the dispute. The officials supported workers’ demands for social insurance but said that there was no legal basis for tenure compensation. The strike ended in a partial victory for workers, but the major demand was not met. This case differs in several ways from the Guangzhou strike. First, since there was legal ambiguity about the basis for the claim, protesters relied on a mixture of publicity and disruptive tactics. These protesters were more interested in getting the local state authorities involved than the protesters at the Guangzhou firm but still contacted media companies,

206

Discursive Opportunities and Collective Action

since they were not sure that the local authorities would support their main demand. Particularly after the official from the local labor bureau declared that the case was legally ambiguous, protesters’ discussions about how to achieve success centered around getting state concessions; the officials’ ambiguous attitudes signaled that workers could get the local government’s support if they maintained the pressure. Because protesters hoped that the local government might intervene, they enlarged the physical scope of conflict beyond the factory gate and marched to the local labor bureau. Second, the firm had violated the law by not providing social insurance for workers. Workers strategically highlighted this point in their banner, when their major demand was that their tenure should be paid out. The firm’s violation of labor laws gave workers leverage to involve state authorities, because the local labor bureau and trade union had to intervene when a protest included legal demands. This strategy confirms the Longgang district government’s assessment in 2008 that one of the new tactics of workers’ collective action seemed to be to insert extra-legal demands among legal ones. In this case, workers’ ex ante perception of group efficacy originated in major part from their assessment of local governments’ reactions rather than those of non-state third-party allies. Third, due to the high stakes involved, workers with a longer firm tenure were willing to take on a higher risk struggle. This complements the findings in Chapter 6 that even those workers who are socially embedded in the firm can be incentivized to launch collective action when the demand disproportionately affects those with a longer firm tenure. Some workers expressed concern that their employment might end as a result of the collective action. “If this fails,” said one worker who had been employed for six years at the firm, “most will quit the job one by one.” Yet, as older workers stated explicitly, they were aware of the large stakes involved in claiming compensation for tenure. Finally, the case also shows why media attention is sparse for lawbased protests. Since law-based protests occur at less newsworthy firms – because newsworthy firms are ex ante incentivized to abide by laws – workers at the most downtrodden firms do not have access to discursive resources that would enable them to use publicity as a tactic. At these firms, workers need to use disruptive tactics to attract publicity, but this carries higher risk than protests that draw media attention because of the profile of the firm. This could explain why high-profile protests that were reported in the domestic media have mostly been at foreign-owned and large-scale firms.

7.3 Cases of Labor Protests

207

The discussion also implies that firm-level patterns of law-based collective action should depend on local governments’ responsiveness. When local governments are responsive to law-based collective action, even workers at firms without much discursive resources would be able to launch collective action in the expectation that they could get state authorities involved on their side. If the state actors are not responsive to protesters’ asking for legal demands, firm-level patterns of law-based collective action should also hinge on the availability of discursive resources, but those resources would mostly be used to draw the attention of state actors. The cases also show that interest-based protests may be less destabilizing to local governments, at least in the short run. 7.3.3 Diaoyu-Senkaku Dispute and Labor Strikes at Japanese Firms Labor strikes that occurred following the dispute between Japan and China over the ownership of the Diaoyu/Senkaku Island offer another example of migrant workers engaging in interest-based collective action. After the announcement that Japan “purchased” the island in September 2011, Chinese workers staged collective action at large-scale Japanese enterprises including Canon, Panasonic, Mazda, Nissan, and Ohm Electronics. Omron Corporation, for example, experienced a three-day strike right after the announcement and suffered a loss of about five million dollars. After its three factories experienced workers’ protests, Panasonic temporarily closed down the production facilities and gave employees paid leave. Studies on anti-Japan protests in 2012 found that migrant workers made up a large part of the group of nationalist protesters.34 While the strike wave had its origin in political anti-foreign demonstrations, workers at many firms demanded a wage increase during the protests. The Shenzhen Association of Foreign Enterprises reports that about 20 percent of the 56 Japanese firms that the association surveyed had experienced workers protesting for a wage increase. A report issued by the association states that “some workers turned their nationalism into demands about a wage increase and created new labor conflicts.”35 Migrant workers might not be the only social group to have advanced unrelated demands during the anti-Japan protests. Wallace and Weiss (2015) argues that “within limited alternative channels for expressing discontent, nationalist protests may also give aggrieved citizens, including dispossessed farmers, migrant workers and laid-off workers, the opportunity to air their grievances” (p. 410). 34 35

Wallace and Weiss, 2015. Shenzhen Foreign Invested Firm Association, 2012.

208

Discursive Opportunities and Collective Action

Labor protests at foreign-invested firms are often prompted by the events that incite Chinese workers’ nationalism, such as the rumor that Chinese workers are being discriminated against, or international conflicts that involve hostile behavior of other states toward China. Workers display anti-foreign sentiment during collective action. Protesters refer to those who do not cooperate as national traitors (hanjian) or running dogs (zougou). On the sites of some strikes, workers hang Chinese national flags. There are some suggestions that Chinese workers at foreign enterprises perceive a labor dispute as a conflict between Chinese people and foreigners.36 In Chan’s analysis of a labor dispute at a Korean-owned factory in 1995, a worker described his resistance against his Korean manager’s order to kneel down as a patriotic act; he said, “I had only one thing in mind; even if I died I wouldn’t kneel! I’m Chinese.”37 The author concludes that the Chinese government’s education in patriotism made the worker believe that to kneel to the Korean manager “not only insulted him personally, but also insulted the national character of the Chinese people.” According to Pun and Chan’s study of a labor strike at a Taiwanese factory, workers publicly expressed discontent at the firm’s management style by linking it to the management’s Taiwanese identity: one worker, for instance, chanted “You Taiwanese guys do not treat us [mainland Chinese] as human,” while others referred to their manager as “the Taiwanese lao,” a name that embodies disrespect and disdain.38 Such observations align with some managers’ views that their behavior at foreign enterprises can be seen as representing their country. A firm manager of a Koreanowned factory in Guangzhou, for example, expressed concerns over labor disputes at his factory, emphasizing that “we (Korean managers) need to be careful when dealing with labor conflicts in China, because what we do can be easily translated into what Koreans do.”39 These arguments generally suggest that anti-foreign labor protests are an expression of workers’ patriotism. Yet, it is difficult to conclude that these labor protests originate from workers’ patriotism alone. If the main goal is to express patriotism, why are some of the workers demanding a wage increase during the protests? The workers’ advancement of wage-specific demands during the Diaoyu/Senkaku Island dispute hints at the strategic motives 36 37 38 39

Chan, 2001, 2009 Ibid., p. 58. Chan, 2009, p. 294. Interview 01-2013.

7.3 Cases of Labor Protests

209

driving the protests. Evidence suggests that anti-foreign protests are often the product of protest leaders’ strategic calculations about the efficacy of using anti-foreign framing. Protest leaders understand the effect of antiforeign slogans on mobilization. A study by Chan and Pun (2009) on a labor strike at a German factory, for example, shows anti-foreign slogans at work. Strike leaders circulated a pamphlet that said, “if you want to be [. . . ] a shameful Hanjian, then you can sell your body before we get our wage demand” (p. 300). The original meaning of Hanjian is a national traitor who sells their country to foreigners, and the term was used to denounce any workers who had returned to work before the strike succeeded. The strike leaders at the Yue Yuen factory in May 2014 also targeted noncooperators with anti-foreign slogans. Strike leaders posted leaflets that said, “Down with Taiwanese exploiters, down with mainland running dogs. If things do not get resolved, those who go back to work will die without offspring.” Workers also posted the pictures of Chinese managers who were on the side of the Taiwanese management, calling them “Hanjian.” During a labor strike at Philips – an electronics company owned by Netherlands – in 2014, workers circulated “a letter to all brothers and sisters” to solicit mobilization, saying that “We need to protect our own rights! We are Chinese people. Chinese people should not suppress fellow Chinese people.” The anti-foreign slogans are also used to elicit the support of external actors. The main goal advanced during the nationalist protests at Japanese-owned factories was an interest-based demand that would not usually be supported by state actors. In ordinary circumstances, workers would not be able to rely on state actors for support. However, during anti-foreign protests, protesters appeared to succeed in expanding discursive opportunities by tapping into widespread nationalism among the public and its effect on their firm and the local government. Some workers noted that the anti-Japan strike wave in 2012 gave them an opportunity to make economic gains. During a labor strike at Canon in Zhuhai, a strike participant stated, “workers [were] trying to seize this opportunity to get better working conditions.”40 In my interviews, workers argued that foreign slogans could be effective mainly because widespread nationalism made the public sympathetic toward anti-foreign protests. Public opinion could put pressure on local governments to side with workers because the officials would not want to be perceived as helping foreigners; firms would also be pressured to meet workers’ demands because 40

“Japan-related books disappear in Beijing; Chinese demand pay hikes from Japanese employers” 2012.

210

Discursive Opportunities and Collective Action

they do not want to be perceived as foreigners exploiting Chinese workers. A worker at a Japanese-owned automobile factory stated that when protesters at their firm used anti-foreign slogans, the local government officials could not stop them because the officials were “concerned about the public opinion.”41 “Because of the acute anti-Japan sentiments these days,” the worker added, “both the Japanese managers and the local officials are extremely careful in dealing with labor conflicts at my firm.” There was a learning effect from the 2012 strike wave, as workers realized that anti-foreign slogans could be an effective tool. After the Diaoyu/Senkaku Island dispute, workers at some Japanese enterprises staged a strike on the 18th of September the following year to coincide with the date commemorating the 1931 Japan–China war. Some firms anticipated a labor protest on that day and offered the entire factory paid leave. Even though the war had occurred more than eighty years before, workers only started to engage with the 9.18 strike after the 2012 Diaoyu/Senkaku dispute. This shows that the 2012 strike wave taught workers that anti-Japan slogans could be used as leverage against the firm, by using nationalist sentiment to attract the attention and support of the public and local authorities for the workers’ cause. Workers’ expectations of support during anti-foreign protests sometimes lead to anger, when local governments failed to meet the expectations. The 2010 Honda strike shows that the local government could be the target of tremendous pressure if they were criticized for not helping Chinese workers who strike against foreign managers. On the fifth day of the strike, a group of people with yellow caps appeared on the strike scene claiming to be from the local trade union. These people tried to persuade workers to go back to work, but as workers continued to protest, the people in caps turned to violence and started to beat some strike participants. Infuriated by such a repressive response, the workers blamed the local trade union chanting, “Down with national traitors.” After the events were reported in the news, the local trade union was harshly criticized by the public. The trade union at the district level issued a public statement of apology the next day. The example shows that public nationalism creates pressure for local governments to support workers during anti-foreign collective action at foreign-invested enterprises. The labor strikes at Japanese-owned firms following the international dispute, and the use of anti-foreign slogans during collective action at foreign-owned enterprises more broadly, demonstrate migrant workers’ 41

Interview 09-2013.

7.4 Broader Implications

211

sophistication in protest tactics and their strategic understanding of their environment. Some workers display a clear understanding of the strategic benefits of using anti-foreign slogans, particularly at a time when the Chinese public is highly sensitive about China’s involvement with other countries. The labor strikes at Japanese enterprises show that anti-foreign framing can be used during interest-based collective action to garner the support of state authorities while also appealing to the public and domestic consumers’ nationalism. Nationalism could become a tool for opening up discursive opportunities and drawing the attention of nonstate actors for those engaging in interest-based collective action. In other words, workers not only utilize the discursive opportunities present in their environment but sometimes create them. 7.4 BROADER IMPLICATIONS

In this chapter, I analyzed sources of perceived group efficacy and found that workers have higher expectations of attracting the attention of external actors when their firms are law-abiding and produce goods for internationally recognized brands. Based on their strategic understanding of their firm’s position in the global supply chains and in relations to the local environment, workers use innovative publicity tactics. Some attempt to draw the attention of nonstate external actors such as the lead firm in the supply chain, or consumers in China and abroad. Others use anti-foreign priming to appeal to the public’s nationalistic tendencies. These tactics seem less disruptive compared to the traditional approach of protesting in front of the local government offices or blocking the road. The tactics that create market pressure for firms to meet workers’ demands demonstrate both the advantages and risks of atomized incorporation. On the one hand, the tactics are not necessarily regime-threatening as they remain within the boundary of “economic” struggles. Instead of seeking to involve the government, workers try to find innovative ways to expand their discursive opportunities and pressure those nonstate thirdparty allies who can directly or indirectly impact their firm’s decisions. While workers engaging in traditional forms of protest in pursuit of legal claims also seek media attention, those firms are not so newsworthy and the workers often resort to disruptive tactics to get the attention of the local government, which places a burden on the local state. Thus, the interest-based protests that take place at law-abiding firms are easier for the state to distance itself from, compared to law-based protests.

212

Discursive Opportunities and Collective Action

The role of discursive resources in enabling interest-based protests also suggests that atomized incorporation can distribute material rewards to those with more capacity to influence social expectations. China scholars have disagreed about the implications of increasing social unrest for citizens’ political consciousness.42 Lorentzen and Scoggins (2015) suggest that the apparent increase in rights discourse among the public could reflect one of three scenarios. First, it could be mirroring citizens’ growing interests in challenging the state to protect their rights. Second, it could simply be a consequence of the state’s changing policies and its increased tolerance of popular claim-making. Finally, it might be due to the changes in society’s shared expectations about what their rights are. Citizens might be more inclined to protest for their rights when they believe that their behavior would receive public support, in the form of support from nonstate third-party actors. The third scenario, the authors contend, might have the most implications for China’s political stability. The findings in this chapter show that public support – and workers’ perceived sense of shared values – shapes collective action. The presence of discursive resources at certain types of firms could mean that workers at those firms might have a stronger capacity to influence the shared social expectations about citizens’ rights. Atomized incorporation then could be an effective approach to manage different groups of workers. Selective material co-optation of the workers at these firms with discursive opportunities could help the regime to detect those workers who are aggrieved and influential. On the other hand, however, the uneven distribution of discursive resources also means that a substantial number of workers with strong grievances about the central government are unable to access the discursive resources to generate bargaining power. The findings parallel what Wolfsfeld (1997) refers to as the “principle of accumulative inequality,” which posits that “those who most need the news media are the ones who find it the most difficult to obtain them” (p. 24). Workers employed at firms with the most lagging conditions might need the support of third-party allies the most, but they are the ones who lack the discursive resources to tap into the tolerated space of contention. This reinforces the finding in previous chapters that atomized incorporation could be effective at demobilizing the labor movement by co-opting those with more resources for successful collective action but might not work as a long-term approach to appease disgruntled workers. 42

Pei, 2010; Perry, 2008; Li, 2010a.

7.4 Broader Implications

213

The extent to which the regime enforces the labor laws also affects the importance of discursive resources for law-based protests. If the regime were to strictly enforce labor laws, workers at firms with illegal labor practices would not have to engage in publicity tactics – they could simply inform the state authorities about the illegal practices. In this case, workers at the least renowned firms would expect their legal collective action to be as likely to succeed as action at the most renowned firms. We would not necessarily see more collective action taking place at law-abiding firms in this situation. This suggests that the regime is able to influence the incidence and nature of collective action by calibrating the legal and discursive resources available for collective action. By influencing the civil society actors that help workers generate discursive opportunities, the regime is able to encourage or discourage interest-based protests. This also has important implications for the patterns of collective action during the Xi era, which I discuss in Chapter 8.

8 State–Labor Relations in the Xi Era and Beyond

China is often portrayed as a winner in the globalization stakes, benefiting from the relocation of manufacturing jobs away from other industrial economies. Internationally, it has generated import competition in advanced economies and contributed to both the backlash against globalization and the rise of populism. Domestically, China’s rise is postulated to have provided performance legitimacy for the CCP and strengthened its one-party rule. China’s path of economic reform has been characterized as “reform without losers.”1 It is argued that Chinese citizens perceive the CCP as the facilitator of economic development.2 China’s rise, however, also induced domestic political instability. It ushered in a new era of state–labor relations, as the rapid economic growth absorbed surplus labor from the countryside and reduced the efficiency of labor-intensive industries. There were losers from economic reform and one of the most disadvantaged groups was migrant laborers, whose wellbeing was neglected in the pursuit of export-led growth. While searching for a more sustainable model of economic development, the Chinese regime has been dealing with growing labor discontent among migrant workers, who have become more educated, more technologically savvy, and more ambitious in what they expect from life. In this book, I theorize a particular form of state–labor relations that emerged in this new environment as atomized incorporation. I argue that the Chinese regime faces a difficult task of controlling unorganized 1

2

The term was specifically used to characterize the dual-track approach to reform in the initial stage (Lau et al., 2000). It has been applied more broadly to describe the evolutionary style of Chinese economic reform compared to the big bang approach in Eastern Europe. Wright, 2010.

214

State–Labor Relations in the Xi Era and Beyond

215

workers who are increasingly vocal about their dissatisfaction. Because it has not enabled meaningful organizational representation, the regime has limited tools of co-optation and relies on a dual approach. On the one hand, the regime offers very narrow channels for claim-making by individual or small groups of workers, which has restricted access to economic and advocacy resources to those who express their claims within the state-sanctioned boundary. On the other hand, the regime has repressed any organized interests and broader structures of mobilization that try to emerge outside of this boundary. Using my analysis at the firm and individual levels, I have argued that this limited form of incorporation has mainly targeted those with resources for successful collective action. Due to the many barriers to labor organizing, not all workers are capable of generating collective action. This outcome has its important merits for the regime given that a substantial number of workers blame the central government for their plight. By rewarding those with the resources to initiate collective action, and thus co-opting them, atomized incorporation is able to reduce the immediate threat of a widespread labor movement. Those who are capable of generating collective action have come to view atomized protests as an effective channel to resolve their firm-specific demands; both at the individual and firm levels, collective action resources predict a higher incidence of experiencing atomized protests and a stronger belief that atomized protests achieve job-specific material gains. The findings in this chapter thus demonstrate that atomized incorporation has been a relatively successful approach to undermine the revolutionary threat from below by demobilizing labor movement. Yet there are also signs to suggest that atomized incorporation has not been effective at enhancing migrant workers’ political loyalty in general. Despite the regime’s efforts to frame labor issues as depoliticized internal disputes that affect a few individual workplaces, I showed that both younger and more experienced migrants have come to perceive the plight of migrant labor as a systemic condition. Atomized labor protests remain depoliticized but have not been a channel for workers to express their social grievances about the broader society. Those workers with high levels of social grievances, which could be politicized, often do not have the capacity to mobilize other workers in the current environment. My research shows that a growing number of workers have political grievances that remain out of sight of the regime. Projections of China’s political future, therefore, might need to consider what is not visible in the space of explicit contention.

216

State–Labor Relations in the Xi Era and Beyond

My findings suggest that the incorporation of disgruntled workers as economic agents, but not as political citizens, has its limits. The current approach to managing labor discontent only takes into account workers’ immediate needs for material gains, but not their other, equally important needs as social and political beings – such as the need for personal growth, recognition, community, and expression. By allowing individual or small groups of workers to contend for short-term economic gains, the regime has successfully created a depoliticized space for contention, but the information constraint that conceals the extent and patterns of broader grievances still remains. Limited incorporation thus amplifies the need for atomization. Atomized contention is politically safe when it only attracts workers who do not have substantial interests in seeking change that reaches beyond a single workplace. The widespread grievances about the central government, however, show that the regime needs to carefully monitor the emergence of atomized protests and make sure to isolate each protest from other protests and from potential external mobilizers. Disgruntled workers are susceptible to external mobilization led by insurgent activists and organizations with more far-reaching goals. When those seeking political change acquire and control the organizational resources to facilitate atomized protests, politicized labor activism could emerge. As the regime continues to allow workers to express their short-term economic grievances, it needs to ensure that they remain sufficiently atomized and that those who contend do not move on to anti-regime activities with a broader agenda. 8.1 ATOMIZED INCORPORATION IN THE XI ERA

This dilemma is already well reflected in the waves of crackdowns on labor activists in the Xi era. Under Xi Jinping’s leadership, the Chinese government has adopted a more repressive approach to civil society in general, but labor activists and organizations have been a particular target of increased repression. Two large waves of crackdowns since 2014 have led to the disappearance, detention, and arrest of prominent figures in labor activism. The limited space that had been allowed to these individuals and organizations for their participation in workers’ collective struggles is now gone. How has this restriction affected atomized protests? 8.1.1 Continuation of Atomized Protests Despite the increased repression of labor activists, there does not appear to have been an increase in police repression of atomized protests or

8.1 Atomized Incorporation in the Xi Era

0

0.2

0.4

0.6

0.8

217

2011

2012

2013

2014

2015

Foreign-invested Private FIGURE

2016

2017

2018

2019

Taiwan and HK

8.1 Cases of labor protests by ownership

a reduction in the number of protests. As discussed in Chapter 6, the number of protests recorded on the CLB Strike Map has increased continuously since 2011 (see Section 6.1). The CLB Strike Map is not a complete dataset of labor protests and so is not conclusive evidence of an overall trend, but the data support the observation that atomized protests are still frequent.3 When analyzing the frequency of police presence per year, the same dataset also does not confirm that police repression of atomized protests is substantially higher in recent years. Between 2011 and 2019, there is no notable trend of increasing police presence. On the contrary, police presence recorded in the CLB Strike Map has decreased since 2014. When we break down the nature of labor protests, however, the changing trend is evident. Not only have protests largely been focused on law-based claims since 2015, they have mostly been occurring at domestic private and small-scale firms (see Figures 8.1 and 8.2).4 The major tactic used in collective action was a labor strike until 2014, but this was replaced by marching in the street – a more disruptive tactic associated 3 4

Hernandez, 2018. The drop in the percentage of protests at foreign-invested firms reflects not only a change in the percentage but also a change in the absolute number. While about ninetyfour cases occurred at foreign-owned enterprises in 2014, the number dropped to only fourteen in 2019.

State–Labor Relations in the Xi Era and Beyond

0

0.2

0.4

0.6

0.8

1

218

2011

2012

2013 FIGURE

2014

2015

2016

2017

2018

2019

8.2 Cases of small-scale protests

with law-based collective action (Figure 8.3). The trend suggests that atomized protests still occur frequently but the basis of collective action has largely changed from interest-based to law-based. The majority of labor protests in recent years were organized around the payment of wage arrears. Several factors could explain the change. First, due to the increasing costs of inputs such as labor and raw materials, as well as growing competition in export markets, labor-intensive firms in the manufacturing sector have faced increasing financial pressure in China. When firms are less profitable, workers are less inclined to protest for higher wages but they do engage in defensive strikes. The increasing number of protests for wage arrears could thus reflect the declining profitability of labor-intensive sectors in China. For instance, total employment in Chinese manufacturing peaked in 2014 but has declined since then, dropping by 20.1 percent in 2018.5 Second, the Xi regime’s repression of civil society more generally could have affected workers’ access to discursive resources, which I found to 5

The PRC Statistical Yearbook provides information on the size of employment in different sectors. The data however only captures firms whose annual revenues exceed 5 million yuan.

8.1 Atomized Incorporation in the Xi Era

0

0.2

0.4

0.6

0.8

219

2011

2012

2013

2014

2015

2016

2017

Strike Marching in the street FIGURE

2018

2019

Road blockage

8.3 Cases of labor protests by tactic

be important for enabling interest-based collective action. Lei (2017), for instance, argues that the leverage of rights lawyers, media journalists, and public opinion leaders – which constitute the contentious public sphere – has been greatly diminishing in the Xi era. Journalists enjoy less freedom in investigative reporting due to the leadership’s tighter control over media companies.6 This could make it more difficult for workers to reach the third-party allies that could enhance their bargaining power. When workers perceive neither an opportunity for a big gain – due to declining firm profits – nor access to the discursive resources that strengthen their bargaining power, labor protests might only occur at the firms where workers’ survival is threatened. 8.1.2 Increased Atomization The increasing trend of state repression has been most evident for labor organizations and activists. Multiple waves of crackdowns – most notably, the 2015 crackdown on labor NGOs in Guangdong and the 2019 crackdown in the aftermath of the Jasic incident – show that the regime is increasingly concerned about the unintended consequences of 6

Repnikova, 2017b.

220

State–Labor Relations in the Xi Era and Beyond

the way in which atomized incorporation has been implemented. It might be convenient to attribute the increased repression to a fundamental transformation of state–labor relations in the Xi era, but it is important to highlight that the change was not necessarily abrupt. It is a continuation of many elements that had already surfaced in the late Hu era.7 Atomized incorporation, like other policies in post-reform China, was not a singular or uniform phenomenon. Localities experimented with different policies and approaches to channel citizens’ grievances into politically safe spaces. While the process had been carefully monitored and managed by the state, it also yielded unintended consequences. Waves of crackdowns on labor activists and organizations reveal a shift toward increased atomization as a response to the unintended consequences but do not necessarily signal a departure from atomized incorporation. One of the unintended consequences has been the emergence of labor NGOs that challenge atomization – albeit covertly. It has largely been assumed that labor NGOs and activists in the post-reform era play a limited role in labor activism, since they stay away from politically sensitive areas, most notably workers’ collective action. It has been rare for labor activists or NGO staff to be seen at protests, for fear of unwanted government attention. However, evidence suggests that some NGOs had adapted to the restrictive political environment and played a more covert political role. As more individuals and small groups of workers utilized atomized channels of claim-making, NGOs and activists began to provide not only legal advice but also training in how to maximize the chance of success.8 The training appeared to be depoliticized but worked to counteract the effects of atomization. Fu (2018), for instance, argues that some NGOs provided a pedagogical process that fostered “collective consciousness through constructing frames.”9 They only seem to assist atomized claimmaking, but helped to re-frame “individual grievances as part of broader common problems shared by millions of others,” which is critical for the politicization of labor discontent, as I discussed in Chapter 4. In the PRD region, a substantial number of labor NGOs began to pursue workers’ collective interests. In 2015, Duan Yi, a well-known human rights lawyer and labor advocate, stated that some labor NGOs in China were no longer “service-oriented” but had evolved to be “movementoriented.” Rather than focus on providing legal aid to workers, these 7 8 9

For instance, recent studies on the Xi regime’s dealings with civil society also stress both continuation and change (Lei, 2017; Fu, 2018). Chen et al., 2015; Leung, 2015; Fu, 2017. Fu, 2018, p. 18.

8.1 Atomized Incorporation in the Xi Era

221

movement-oriented NGOs advanced collective bargaining as a new way to promote workers’ interests.10 The change was possible in part due to the local government’s more lenient attitudes toward new initiatives, such as collective bargaining and the democratic election of trade union leaders within enterprises. In the early 2010s, local governments in Guangdong began to experiment with these initiatives as a way to stabilize labor relations. High-profile labor strikes in the region revealed that legal channels alone could not prevent collective action. The rationale behind the experiments was to establish an institutionalized mechanism to represent and negotiate labor demands within workplaces, while discouraging workers from taking their grievances to the street. It seemed that collective bargaining in Guangdong could be a new model of more inclusive state-labor relations. The Guangdong Provincial Government introduced the Guangdong Regulations on Collective Contracts in Enterprises in 2014 as a framework for enterprise-level collective bargaining. The effectiveness of their regulations was questioned, because enterprise unions either did not exist or were controlled by firm management in the majority of enterprises where they did exist. Yet it was still a positive signal to many NGOs that increasingly perceived their future role to lie in collective bargaining. Particularly in Guangzhou, where a wave of labor strikes swept the automobile industry in 2010, the city government demonstrated greater tolerance for labor NGOs’ involvement in collective bargaining. Labor activists in the region actively trained and coached workers during collective struggles and disseminated information about best practices. They issued their own grassroots handbook of collective bargaining, called the Code of Collective Bargaining, to offer a practical guide to workers.11 The NGOs’ involvement led to many successful cases of collective bargaining in the PRD region.12 8.1.2.1 Crackdown on Labor NGOs in 2015 Yet in 2015, the state changed its stance and targeted the movementoriented NGOs during its crackdown on labor activists. While collective bargaining was perceived as a middle road by the labor NGOs – which do not directly threaten the regime’s political stability but could contribute to workers’ collective interests – in late 2015, the Guangdong Provincial government, city governments, and the central government clearly signaled their disagreement. While it is still not clear why the state launched 10 11 12

Chen and Yang, 2017; Franceschini and Lin, 2019. China Labour Bulletin, 2013. Chan and Hui, 2014; Kuruvilla and Zhang, 2016; Franceschini and Lin, 2019.

222

State–Labor Relations in the Xi Era and Beyond

the attack at that moment despite many mixed signals, we can guess based on available information that the state’s aversion toward collective bargaining, particularly the involvement of movement-oriented NGOs, originated in the growing risk of atomized contention being politicized. For one, it seemed that collective bargaining did not reduce the incidence of collective action, as hypothesized by local governments at the onset of their experiment. Most importantly, it was difficult for workers to initiate collective bargaining without engaging in some form of collective action, because management rarely agreed to sit at the negotiation table. Chan and Hui (2014) call this “collective bargaining by riot,” suggesting that one often induces the other. While collective bargaining could help workers and firm management to de-escalate after collective action had commenced, workers found that firms were unwilling to negotiate unless compelled to do so by actual protests. Labor NGOs did not openly encourage workers to engage in labor strikes, but their approach to collective action became much more permissive than state authorities’ responses. Their code acknowledges workers’ rights to strike – albeit as a last resort – which sets it apart from the state’s guidelines on collective bargaining. In prominent cases of collective bargaining, workers adopted a combination of collective action and negotiation to advance their demands. During the process, the role of some labor NGOs was critical in strengthening workers’ organizational capacity and sustaining their collective struggle. Perhaps encouraged by the lenient attitudes of local governments and trade unions in the early 2010s, the movement-oriented labor NGOs started to play a more visible role in workers’ collective struggles. In 2014, when the Yue Yuen strike occurred, activists from Chunfeng – a wellknown NGO in Shenzhen – provided advice on how to organize workers and bargain with firm management. Since labor NGOs rarely appeared at protests, Chunfeng’s more visible role during the Yue Yuen protest was widely considered to be a bold move. Not surprisingly, the local police warned the activists from Chunfeng not to intervene in the protest and later detained them when they persisted with their involvement.13 In the Lide struggle – considered to be the incident that promoted the 2015 crackdown – movement-oriented labor NGOs like the Panyu Migrant Workers’ Service Center were seen to play a much more pronounced role in workers’ collective struggles, including labor strikes, than they did in the past. The Panyu Center in Guangzhou was the most prominent labor NGO to promote collective bargaining in the PRD area. 13

Sevastopulo, 2014.

8.1 Atomized Incorporation in the Xi Era

223

Established in 1997, the Panyu Center first began as an organization to provide legal advice to workers in labor disputes but changed its main agenda to focus on collective bargaining. In the early 2010s, the Center assisted in many prominent cases of successful collective bargaining in various industries, such as jewelry and sanitation. During the Lide struggle, the Panyu Center played a critical role in mobilizing and organizing workers throughout their eight-month journey. The Lide struggle shows that workers’ collective action assisted by labor NGOs – even if occurring in a single workplace – differs from typical atomized collective action in several important ways. First, in the Lide struggle, the Panyu Center functioned as a mobilizing vehicle from outside the organization of production, which meant that workers had more than just their own resources to rely on for mobilization within the firm. The Panyu Center had a good reputation among migrant workers due to its history of involvement in many successful cases of collective bargaining, which gave strength to workers’ expectations and motivation to be part of the struggle. Two main members of staff who helped the workers at the Lide factory were workers-turned-activists who joined the Panyu Center in 2014. Due to their worker background and former experience in labor contention, they were able to gain workers’ trust. Their successful roles as mobilizers show how movement-oriented NGOs such as the Panyu Center can provide mobilizational resources to those workers who are interested in seeking a broader change but otherwise lack the mobilizational capacity within the firm. The presence of external mobilizational resources also undermines the tactics usually used by firms to discourage collective action – such as side payments to protest leaders and the threat of dismissal. When an atomized protest is led by people who are more embedded in the firm, management’s undermining tactics are more effective, as discussed in Chapter 5. When an external organization can provide the mobilizational capacity that sustains the collective struggle, however, the movement no longer has to rely on leaders who are embedded in the firm. The first thing that NGOs train workers to do is to democratically elect their own delegates. Even when some delegates succumb to firm management’s pressure, the NGO helps workers to elect new ones. During the Lide struggle, when some of the worker representatives lost momentum due to management pressure, the staffs quickly stepped in and organized the election of new representatives. The use of labor strikes in the Lide struggle was much more planned and strategic than in a typical atomized protest. The Panyu Center played a critical role by providing timely advice on effective strategies. They

224

State–Labor Relations in the Xi Era and Beyond

knew the importance of launching labor strikes when the management was refusing to come to the negotiation table. In total, workers at Lide engaged in three labor strikes that were “strategically-timed” to pressure management.14 During labor strikes, the staff made sure that workers did not become too agitated or emotionally charged, so that the strike could be used as a bargaining tool. While many atomized protests are more organized and planned than wildcat strikes that occur out of anger, it is often difficult to sustain collective action because of the tactics that the firm and local government use to dissolve the struggle. The Panyu Center assisted workers to stay organized, adopt strategic actions, and sustain their struggle for a long period of time, which differentiated the struggle from other episodes of atomized contention. The Lide case hints at why the state might have chosen the Panyu Center as its main target in the crackdown. The regime allows episodes of atomized contention under the assumption that they are isolated incidents with little potential to develop into a broader movement. While labor NGOs are potentially more threatening to the regime, not all labor NGOs have the capacity that the Panyu Center did to mobilize workers and influence their collective behavior. While their actions might only generate collective action at a single workplace at a time, their reputation among workers, as well as the demonstrated ability to organize workers for collective action, could signal the Panyu Center’s potential as a mobilizing platform. Furthermore, the Panyu Center did not function in isolation but formed a tight network of labor NGOs in the region, including those that were founded by their former staff, such as the Sunflower Women Workers’ Center. Some workers who had come into contact with these labor NGOs during their own struggle became activists themselves, forming an important link in the growing network of labor activists dedicated to workers’ collective action. The Panyu Center’s network also reached Hong Kong. Zeng, the director of the Panyu Center, began to receive funding from the China Labour Bulletin (CLB), based in Hong Kong, and receipt of foreign funding was cited as the main reason behind the state’s crackdown on the organization. It seems that Zeng was harassed by local security officials for this collaboration from the onset. Han Dongfang – the founder of the CLB – recalls that Zeng had already opened up the relationship with the CLB in 2011. Although he faced constant intimidation, the local government did not formally arrest Zeng. 14

China Labour Bulletin, 2015.

8.1 Atomized Incorporation in the Xi Era

225

During this period, state security officers would sometimes threaten Mr. Zeng during their so-called tea-times, demanding that he immediately cease cooperation with China Labour Bulletin. At other times they would ease up, telling Mr. Zeng that the collective bargaining cases undertaken by the Panyu Center had attracted too much attention and upset relevant officials—he should be more low profile, they suggested.

Yet it is clear that the Panyu Center – and other movement-oriented NGOs for that matter – received mixed signals from the state. During collective bargaining, NGO staff had to deal constantly with threats, although the threats were not always explicit. For instance, after the Panyu Center’s involvement in Lide, their staff were questioned and summoned for investigation multiple times. Zeng was even attacked by unknown thugs in December.15 Zhang Zhiru from Chunfeng had to move more than ten times, as the local security officials exercised indirect pressure through landlords. However, even in 2014, some parts of the state – both at the national and local levels – seemed to have a positive attitude toward these movement-oriented NGOs. The Panyu Center’s involvement in the Lide struggle was praised by the ACFTU, and the Global Times – an English news outlet under People’s Daily – ran a report complimenting Zhang’s work. In the end, while it is not entirely clear what promoted the state’s coordinated crackdown on these NGOs in 2015, the Panyu Center’s sustained and in-depth involvement in workers’ collective action – including the labor strikes – seems to have played an important role. 8.1.2.2 The 2018 Crackdown in the Aftermath of the Jasic Incident The Jasic labor incident might be the most prominent case of labor activism and the crackdown that followed it was the largest in scale. During the summer of 2018, workers at Jasic Technology in Shenzhen protested against the factory’s poor working conditions and lack of funding for housing. The Jasic incident was substantially different from other atomized protests in two ways. First, the protesters wanted to establish an independent trade union, which is not a common claim in migrant workers’ protests. Second, not only workers but college students – mostly from Beijing University – joined the collective struggle by forming the Jasic Workers Solidarity Group. The explicit involvement of the college students from Beijing attracted wide attention, including from intellectuals and state officials. The aftermath of the Jasic incident was enormous. 15

Harney, 2015.

226

State–Labor Relations in the Xi Era and Beyond

About thirty workers from Jasic were arrested. Police raided university dormitories in Beijing and rounded up the student activists. The raid coordinated by the regime, spanning from Shenzhen, Huizhou, to Beijing, seemed to be a disproportionate response. It was not common for a large number of workers to be arrested and charged as a result of collective action, and even less so for college students who supported the protests to be caught up in the suppression. In the Lide struggle, for instance, the main target of repression was labor activists, not the workers at the Lide factory. Beneath the surface, however, the Jasic incident might have been a fundamentally different type of collective action compared to an average atomized protest. There is substantial evidence that the Jasic workers’ quest to establish an independent union had its origins in an underground network run by students-turned-worker activists. The network – whose identity was revealed in the aftermath of the Jasic incident – had been nurturing grassroots labor activism in the PRD region for nearly a decade. Being mostly university students, members of the network went into factories with the goal of educating workers with class theories and awakening their class consciousness. Their activities resembled those of the Korean student activists in the 1970s and 1980s, who believed in the potential for factory workers to be part of the foundation of a prodemocracy movement. In South Korea, especially under the military rule of Chun, a group of pro-democracy student activists became strong advocates of engaging with workers on the ground – known as “field theory” (hyun jang non) – and permeating the workplace. A substantial number of prominent labor strikes in the 1980s were led by students-turned-workers who worked in factories for a number of years and accumulated enough social capital to mobilize other workers. Not surprisingly, many studentsturned-workers in China – including those who were part of the Jasic struggle – cited the Korean labor activism as a source of inspiration. The student – worker alliance in the Jasic struggle should have been perceived as a serious threat to the regime for several important reasons, regardless of whether it could actually develop a politically threatening situation. One, these underground activists were known to target factories with lagging labor conditions – the kinds of factories where workers are more aggrieved but lack resources for successful collective action, according to my findings. They would infiltrate workplaces with high levels of social grievances. Two, the activists were directly challenging the effect of atomization by providing a broader context for workers’ everyday struggles. The students took Lenin’s What is to be done as the guide

8.2 Beyond Protests

227

for approaching workers and providing political education. Finally, some of the students involved in the Jasic incident were also active in other social movements. At the time the Me Too movement was popular in China. One of the most prominent figures of the movement, Yue Xin – a student of Beijing University who initiated a campus-wide campaign against a professor accused of rape – was also a core member of the Jasic Workers’ Solidarity Group. She was one of the fifty students who were arrested on campus. The Me Too movement attracted substantial attention nationally that year and was also the target of a state crackdown. A major reason given for such repression was the wider network and broader influence of feminist activists. The fact that some of those active in other social activism were involved could have been perceived as a threat, with greater potential to attract attention and supporters. The students’ involvement in the Jasic struggle was not without its controversies – about their strategy, effectiveness, and motives.16 Their deep commitment to the Maoist ideology also brings up the question of whether their involvement could in any way threaten the CCP’s oneparty rule. As the labor movement in Korea during the 1970s and 80s demonstrate, ideologically charged students had only limited influence over workers. It was only when students and activists understood workers’ everyday experiences from workers’ perspectives and affected their perception of identity not only inside but outside the workplace that the student involvement created a meaningful influence.17 Yet, even if the students’ involvement in the Jasic struggle was somewhat disconnected from workers and could not gain their support, it provided enough reason for the regime to launch a large-scale crackdown. Maoist or not, the students’ intervention in workers’ struggles seems to have signaled to the regime that the Jasic struggle is not an atomized action and presents a risk of escalation. 8.2 BEYOND PROTESTS

These two crackdowns demonstrate that the regime has the coercive capacity to repress external groups that attempt to mobilize workers from outside. As I have argued in this book, atomized incorporation has not been able to resolve workers’ social grievances in a lasting way. Workers’ grievances about society alone, however, cannot automatically predict the onset of a political movement. The emergence of a broader movement 16 17

Loong-Yu, 2019; Gongchao, 2020; Gongchao, 2020a. Koo, 2001; Lee, 2007b.

228

State–Labor Relations in the Xi Era and Beyond

requires its own mobilizational structures, which the regime has succeeded at undermining via atomization. What could the social grievances of workers mean for China’s long-term resilience? If the regime is successful at monitoring the potential development of cross-group alliances and discouraging broader mobilization, the grievances might not pose a risk to the regime. So do these grievances still matter for our understanding of China’s political future? Recent developments suggest that they do. The social grievances, while not contested during collective action, have been manifested as passive forms of resistance. Many migrant workers are dissatisfied with their life and the broader society but do not expect that they could have an impact and change the society. As a result, they appear disenchanted and to have given up on their aspirations. The designation “sanhe youth,” for instance, refers to young migrant workers who gather at the Sanhe labor market in Shenzhen; they are known to engage only in casual labor – “work one day and rest for three days” – and do not aspire to achieve a better life. The contrast between the typical image of young migrant workers as those with high aspirations for success and the apathetic sanhe youth has received widespread attention both domestically and internationally. In their book Why wouldn’t I want to go back? (qibihuaigui), sociologists Tian Feng and Lin Kaixuan argue that the workers’ seemingly detached attitudes do not necessarily mean that they lack socioeconomic aspirations. The sanhe workers do have high social aspirations and are sensitive to social unfairness but believe that they do not have the means to make a change; thus, they choose to engage in a passive form of resistance by defying the social norms about success.18 Similarly, other migrants have also joined the “lying flat” (tangping) movement by leaving the workplace for an idle and minimal life in the countryside.19 These workers no longer strive to get ahead in the existing system; they reject the system by disengagement. Although they seem passive, these actions still have a social impact. Instead of overt rebellion, the workers are wielding what Scott called “weapons of the weak.”20 These subtle forms of noncooperation may or may not channel into political activism but nonetheless pose a considerable challenge to the central leadership that has been calling on the youth to facilitate the “China Dream.” This is why the regime has been alarmed by the “lying flat” movement and urged the young workers to become 18 19 20

Tian and Lin, 2020. Ji et al., 2021. Scott, 1985.

8.2 Beyond Protests

229

productive members of the society. The original post that called for the “lying flat” movement was quickly taken down by state authorities; state media also criticized the movement as “shameful.”21 In the aftermath of rapid economic expansion, the regime faces an important task of upgrading the economy to generate more sustainable development. It depends on the cooperation of the younger generation to support the ever-increasing older generations. The young workers’ refusal to comply with the regime’s plan to reinvigorate the economy could be a source of political instability. Even if workers do not go so far as to leave for the countryside, those who remain in the labor market could engage in many more subtle forms of noncooperation. When atomized incorporation only allows limited types of communication channels and workers are unable to raise their collective voice about life and society, they could be discouraged from directly challenging the regime – but this might encourage instead the subtle expression of noncooperation that the regime might not be able to repress. At the same time, the young workers’ disengagement might not necessarily signal that they are disinterested in raising their voice altogether; in appearance, their behavior seems to suggest that they are “exiting” from mainstream society, mentally and physically, and thus would not be interested in changing the status quo. Yet, the “exit” here is merely symbolic and does not demonstrate that those social grievances have dissolved. If limited opportunities for upward mobility and self-realization are the main sources of frustration, choosing to do nothing may not be a sustainable solution to the problem. This means that the widespread grievances of younger workers could still be mobilized if political opportunities arise. While it is not clear if and when those opportunities could be present in China, given the repressive political environment for any type of political activism, this means that the regime needs to keep improving its monitoring and coercive capacity for the entire society. The ongoing episodes of crackdowns do show that the regime has substantial coercive capacity, yet they also reveal that the regime is not always able to prevent the emergence of activism in the first place. From the regime’s perspective, even if workers at the moment do not openly display political antagonism, the need for monitoring and suppression remains important.

21

Wang, 2021.

Appendix I: Survey Dataset

In July 2013, in collaboration with three academic institutions in China, I organized a survey team and collected survey data. The survey team comprised 23 undergraduate and graduate students, and administered surveys of 618 factory workers from 155 factories in Shenzhen, China. Scholars interested in labor issues in China have collected individuallevel survey data about migrant workers. However, one of the important challenges has been generating a representative sample, as it is difficult to obtain a list of factory workers with high levels of labor mobility. Most factory workers are internal migrants who don’t have an official registration. Previous sampling methods mostly relied on migrant neighborhood sampling, which selected certain streets from the districts with the highest migrant populations. Survey enumerators interviewed the workers they encountered in the sampled streets. This sampling method covers workers with a wider variety of employment backgrounds, including those who are unemployed or in an informal job. However, the sampling method also makes it difficult for interviewers to reach workers in less-populated neighborhoods or gated factories. For this research, a sampling scheme that captures workers’ behaviors and attitudes across firms was important. The neighborhood sampling does not permit systematic firm-level comparison because it is not based on a random sample of firms. At the same time, previous sampling strategies did not allow us to look at the effect of firms’ law-abiding behavior, since they make up a small subset of all firms in the region. Foreignowned enterprises, in particular, were more prone to both comply with labor laws and experience collective action but only comprised a fraction of sampled firms in surveys. In most previous survey datasets, less than 231

232

Appendix I: Survey Dataset

10 percent of workers surveyed were employed by foreign enterprises. To examine why collective action is more frequent at foreign-owned firms despite their more law-abiding behavior, in this study I adopted oversampling. I generated a random sample of manufacturing firms in Shenzhen and sampled roughly the same number of domestic and foreign enterprises. I obtained a list of all firms in the manufacturing sector through a private company that compiles lists of firms in China.1 From the firm list, I generated a list that includes private and foreign (wholly owned) enterprises, excluding state, collective, joint, and individually owned enterprises. State and collective enterprises are outside the scope of my research and so were excluded from the population. I also did not include “semi-foreign” categories: joint ventures and enterprises owned by Taiwan, Hong Kong, and Macau. Since Shenzhen covers about 790 square miles, it was difficult to visit factories in the districts too far from where the survey team was based. I restricted the sample to the areas that the survey team could reach by public transportation in less than two and a half hours. As four districts in the southern part of the city are mostly urban areas with no industrial zones, these districts were also excluded from my sample. The map above shows the geographical scope of the population in this research (see Figure A.1). The target sample size was 600 workers from 150 firms with 75 private domestic enterprises and 75 foreign enterprises. After mapping out the 150 randomly selected firms, the survey team visited each factory and interviewed four workers, two general production workers (pugong) and two line managers (lazhang) from each factory. Since one of the goals of the survey was to see if responses from general production workers differ from line supervisors, I oversampled line supervisors from each factory. To encourage truthful responses, the survey team did not interview workers right at the factory gate, but at least 100 meters away from the gate. One of the biggest challenges during the survey came as a consequence of incorrect or outdated information about the firms in the firm list. Since every year many factories went bankrupt or moved to other regions, the most up-to-date firm list still included some firms that did not exist at the provided address. In these situations, the survey team moved onto 1

One of the major compilers of lists of firms in China is Emage Company Limited based in Hong Kong. The company claims to be one of the three biggest companies to collect firm directories in the region.

Appendix I: Survey Dataset

233

FIGURE A .1 Geographic coverage of survey firms Note: The area within the line denotes the scope of the firm population. The star indicates where the survey team was located.

the next firm in the sample list. Since some firms were very small in size, the survey team could not always recruit four respondents from a firm. In cases where the survey team was not able to recruit more than two workers from the assigned factory, the survey team visited an additional firm in the firm list to meet the target sample size. In total, the survey team visited 155 firms, of which 79 firms are foreign and 76 are domestic enterprises, and interviewed 618 workers, 301 from domestic enterprises and 317 from foreign enterprises. Table A.1 summarizes basic characteristics of firms and respondents by ownership at the firm level. The reported firm size varied from 18 employees to 10,750 employees, with a mean size of about 551 employees. Workers’ ages ranged from 18 to 55, and more than 90 percent of workers were in their 20s and 30s. There were more male than female respondents in the sample, but the higher percentage of male workers was due to the fact that line supervisors are overwhelmingly male. Among the general production workers, the number of female respondents roughly matched the number of male respondents. Workers on average were highschool graduates, although line supervisors had more years of education than general production workers. There are limits to my dataset insofar as it is not a representative dataset of all factory workers in China. It is only based on one city and therefore does not represent workers from all factories around the country. I chose Shenzhen as my survey site because it is representative of China in terms of its openness to foreign capital and export-oriented

Appendix I: Survey Dataset

234 TABLE

A.1 Basic firm-level characteristics by ownership

N. of firms N. of respondents Firm size Age Male (%) Firm tenure (years)

Domestic

Foreign

All

76 301 327 29.0 66 3.03

79 317 751 29.1 67 2.97

155 618 551 29.1 66.5 3.00

Note: Firm size is the reported number of employees in each firm averaged by firm ownership.

industrialization. Shenzhen was the first city to be designated as a special economic zone and was allowed to open its economy to foreign capital immediately after economic reform started in China in 1979. Since then, it has remained a major manufacturing hub and a major destination for foreign direct investment. At the same time, Shenzhen has seen some of the highest numbers of labor disputes. While the city presents an important case to study labor politics in China, the insights from Shenzhen might not be applicable to the political economy of other regions. Another consideration with the dataset concerns selection bias, because I was not able to randomly select respondents from each firm. Because those who agreed to take the survey might be more politically conscious or more contentious than other workers, in order to minimize such selection bias, survey enumerators told workers during the recruitment process that the survey was about migrant workers’ life experiences generally in Shenzhen.

References

Acemoglu, Daron and James Robinson (2009). Economic Origins of Dictatorship and Democracy. Cambridge University Press. Acemoglu, Daron, Thierry Verdier, and James A. Robinson (2004). “Kleptocracy and divide-and-rule: A model of personal rule.” en. In: Journal of the European Economic Association 2.2–3, pp. 162–192. Ahlquist, John S and Margaret Levi (2013). In the Interest of Others. Organizations and Social Activism. Princeton University Press. Ajzen, Icek (1991). “The theory of planned behavior.” en. In: Organizational Behavior and Human Decision Processes. Theories of Cognitive Self-Regulation 50.2, pp. 179–211. Ajzen, Icek and Martin Fishbein (1980). Understanding Attitudes and Predicting Social Behavior. Prentice-Hall. Albertus, Michael, Sofia Fenner, and Dan Slater (2018). Coercive Distribution. 1st ed. Cambridge University Press. Alexander, Peter and Anita Chan (2004). “Does China have an apartheid pass system?” In: Journal of Ethnic and Migration Studies 30.4, pp. 609–629. “An in-depth discussion with a veteran worker at Yue Yuen” (2014). In: Worker View Point. Andrews, Kenneth T and Neal Caren (2010). “Making the news movement organizations, media attention, and the public agenda.” In: American Sociological Review 75.6, pp. 841–866. Anner, Mark (2015). “Labor control regimes and worker resistance in global supply chains.” In: Labor History 56.3, pp. 292–307. Arceneaux, Kevin (2016). “The conditional impact of blame attribution on the relationship between economic adversity and turnout.” In: Political Research Quarterly 56.1, pp. 67–75. Baekkeskov, Erik and Olivier Rubin (2017). “Information dilemmas and blameavoidance strategies: From secrecy to lightning rods in Chinese health crises.” In: Governance 30.3, pp. 425–443.

235

236

References

Barranco, Jose and Dominique Wisler (1999). “Validity and Systematicity of newspaper data in event analysis.” In: European Sociological Review 15.3, pp. 301–322. Becker, Jeffrey (2012). “The knowledge to act: Chinese migrant labor protests in comparative perspective.” In: Comparative Political Studies 45.11, pp. 1379– 1404. (2014). Social Ties, Resources, and Migrant Labor Contention in Contemporary China: From Peasants to Protesters. Lexington Books. Beinin, Joel (2009). “Workers’ protest in Egypt: Neo-liberalism and class struggle in 21st century.” In: Annual Review of Sociology 8.4, pp. 449–454. Béja, Jean-Philippe (2011). “The new working class renews the repertoire of social conflict.” In: China Perspectives 2011.2, pp. 3–7. Blaydes, Lisa (2018). State of Repression. Princeton University Press. Boix, Carles (2003). Democracy and Redistribution. Cambridge University Press. Brändström, Annika and Sanneke Kuipers (2003). “From ‘normal incidents’ to political crises: Understanding the selective politicization of policy failures 1.” In: Government and Opposition 38.3, pp. 279–305. Brown, Adam R (2015). “Are governors responsible for the state economy? Partisanship, blame, and divided federalism.” In: The Journal of Politics 72.3, pp. 605–615. Browning, Rufus P, Dale Rogers Marshall, and David H Tabb (1984). Protest Is Not Enough: The Struggle of Blacks and Hispanics for Equality in Urban Politics. University of California. Bueono de Mesquita, Bruce, James D Morrow, and Randolph M Siverson (2005). The Logic of Political Survival. MIT Press. Butollo, Florian and Tobias ten Brink (2012). “Challenging the atomization of discontent.” In: Critical Asian Studies 44.3, pp. 419–440. Cai, Fang and Yang Du (2011). “Wage increases, wage convergence, and the Lewis turning point in China.” In: China Economic Review 22.4, pp. 601–610. Cai, Fang and Wang Meiyan (2012). “Labour market changes, labour disputes and social cohesion in China.” In: OECD Development Centre Working Papers 307, p. 1. Cai, Yongshun (2002). “The resistance of Chinese laid-off workers in the reform period.” In: The China Quarterly 170, pp. 327–344. (2006). State and Laid-Off Workers in Reform China. The Silence and Collective Action of the Retrenched. Routledge. (2008). “Disruptive collective action in the reform era.” In: Popular Protest in China. Ed. by Kevin J O’Brien. Harvard University Press, pp. 163–178. (2010). Collective Resistance in China: Why Popular Protests Succeed or Fail. Stanford University Press. Chan, Anita (1993). “Revolution or corporatism? Workers and trade unions in post-Mao China.” In: The China Journal 29. (2001). China’s Workers under Assault: The Exploitation of Labor in a Globalizing Economy. ME Sharpe. (2003). “A ‘Race to the Bottom.’ Globalisation and China’s labour standards.” In: China Perspectives 46.

References

237

(2010a). “Racing to the bottom: International trade without a social clause.” In: Third World Quarterly 24.6, pp. 1011–1028. Chan, Anita and Robert A Senser (1997). “China’s troubled workers.” In: Foreign Affairs 76.2, pp. 104–117. Chan, Anita and Kaxton Siu (2010). “Analyzing exploitation.” In: Critical Asian Studies 42.2, pp. 167–190. (2012). “Chinese migrant workers: Factors constraining the emergence of class consciousness.” In: Chinas Peasants and Workers. Ed. by Beatriz Carrillo and David SG Goodman. Far Eastern Economic Review, pp. 79–101. Chan, Anita and Zhu Xiaoyang (2003). “Disciplinary labor regimes in Chinese factories.” In: Critical Asian Studies 35.4, pp. 559–584. Chan, Chris (2009). “Strike and changing workplace relations in a Chinese global factory.” In: Industrial relations journal 40.1, pp. 60–77. (2010b). The Challenge of Labour in China : Strikes and the Changing Labour Regime in Global Factories. 1st ed. Strikes and the Changing Labour Regime in Global Factories. Routledge. (2012a). “Class or citizenship? Debating workplace conflict in China.” In: Journal of Contemporary Asia 42.2, pp. 308–327. Chan, Chris and Ngai Pun (2009). “The making of a new working class? A study of collective actions of migrant workers in South China.” In: The China Quarterly 198.1, pp. 287–303. Chan, Chris King-Chi and Elaine Sio-Ieng Hui (2012). “The dynamics and dilemma of workplace trade union reform in China: The case of the honda workers’ strike.” In: Journal of Industrial Relations 54.5, pp. 653–668. Chan, Chris King-Chi and Elaine Sio-leng Hui (2014). “The development of collective bargaining in China: From collective bargaining by riot to party state-led wage bargaining.” In: China Quarterly 2014.217, pp. 221–242. Chan, Jenny and Ngai Pun (2010). “Suicide as protest for the new generation of Chinese migrant workers: Foxconn, global capital, and the state.” In: The Asian-Pacific Journal 37.2, pp. 1–33. Chan, Kam Wing (1994). Cities with Invisible Walls: Reinterpreting Urbanization in Post-1949 China. Oxford University Press. (1996). “Post-Mao China: A two-class urban society in the making.” In: International Journal of Urban and Regional Research 20.1, pp. 134–150. (2011). “Internal migration in China: trends, geography and policies.” In: Population Distribution, Urbanization, Internal Migration and Development: An International Perspective. United Nations Department of Economic and Social Affairs Population Division, pp. 81–109. (2012b). “Migration and development in China: trends, geography and current issues.” In: Migration and Development 1.2, pp. 187–205. Chan, Kam Wing, Ta Liu, and Yunyan Yang (1999). “Hukou and non-hukou migrations in China: Comparisons and contrasts.” In: International Journal of Population Geography 5.6, p. 425. Che, Xiaohui and Bo Kong (2008). “Bankrupcy? Relocation? Closure? PRD Bears the Pain of Industrial Upgrading (daobi? waiqian? guanting? zhusanjiao chengshou chanyeshengji zhentong).” In: Economic Information Daily.

238

References

Chen, Dan (2016). “Review essay: The safety valve analogy in Chinese politics.” en. In: Journal of East Asian Studies 16.2, pp. 281–294. Chen, Feng (2003). “Between the state and labour: The conflict of Chinese trade unions’ double identity in market reform.” In: The China Quarterly 176, pp. 1006–1028. (2004). “Legal mobilization by trade unions: the case of Shanghai.” In: The China Journal 52, pp. 27–45. (2008). “Worker leaders and framing factory-based resistance.” In: Popular Protest in China. Ed. by Kevin O’Brien. Harvard University Press, pp. 88–108. Chen, Feng and Mengxiao Tang (2013). “Labor conflicts in China.” In: Asian Survey 53.3, pp. 559–583. Chen, Feng and Xin Xu (2012). “Active judiciary: Judicial dismantling of workers’ collective action in China.” In: The China Journal 67, pp. 87–108. Chen, Feng and Xuehui Yang (2017). “Movement-oriented labour NGOs in South China: Exit with voice and displaced unionism.” In: China Information 31.2, pp. 1–21. Chen, Juan et al. (2015). “Life satisfaction in urbanizing China: The effect of city size and pathways to urban residency.” en. In: Cities 49, pp. 88–97. Chen, Patricia and Mary Gallagher (2018). “Mobilization without movement: How the Chinese state ‘Fixed’ labor insurgency.” In: ILR Review 71.5, pp. 1029–1052. Chen, Tain-Jy and Ying-Hua Ku (2014). “Indigenous innovation vs. teng-long huan-niao: policy conflicts in the development of China’s flat panel industry.” In: Industrial and Corporate Change 23.6, pp. 1445–1467. Chen, Xi (2009). “Between defiance and obedience: Protest opportunism in China.” In: Grassroots Political Reform in Contemporary China. Ed. by Elizabeth J Perry and Merle Goldman. Harvard University Press, pp. 253– 281. (2012). Social Protest and Contentious Authoritarianism in China. Cambridge University Press. (2018). “The logic of fragmented activism among Chinese state-owned enterprise workers.” In: The China Journal 81, pp. 58–80. Chen, Yan yuan and Liu Yang (2010). “The impact of labor contract law’s enforcement on Chinese enterprises’ human resources management.” In: Economic Theory and Economic Management 7, p. 66. Cheng, Tiejun and Mark Selden (1994). “The origins and social consequences of China’s hukou system.” In: The China Quarterly 139, pp. 644–668. Cheng, Zhiming (2014). “The new generation of migrant workers in urban China.” In: Urban China in the New Era. Ed. by Zhiming Cheng, Mark Wang, and Junhua Chen. Springer, Berlin, Heidelberg, pp. 125–153. China Labour Bulletin (2011). “Unity is strength.” In: China Labour Bulletin Research Reports. (2012). “A decade of change: The workers’ movement in China 2000–2010.” In: China Labour Bulletin Research Reports. (2013). Developing a Blueprint for Collective Bargaining in China.

References

239

(2015). Shoe Factory Workers Show that Collective Bargaining is Already a Reality in Guangdong. China State Council (2003). “Notice on doing a better job concerning the employment management of and services for migrant workers.” Chong, Dennis (1987). Collective Action and the Civil Rights Movement. University of Chicago Press. Chou, Kwok Ping and Zhuoyi Wen (2007). “The major change of China’s policy of migrant workers: Towards a new paradigm.” In: Chinese Public Policy Review 1, pp. 157–169. Chu, Dazhi and Zhibo Zhang (2010). “Business conditions, issues, and solutions in Shenzhen Bao’an district during the post-crisis period (houjinrongweiji shidai shenzhen baoanqu sanlaiyibu qiye xianzhuang, cunzaiwenti jiduice).” In: Special Zone Economy (tequ jingji) 7, pp. 36–38. Chung, Olivia (2008). “Last call for Guangdong shoemakers.” In: Asia Times. CMAHK (2008). Survey on Members’ Views on the Business Environment in the Pearl River Delta. (2010). Survey on Members’ Views on the Business Environment in the Pearl River Delta. (2011). Survey on 2011 Business Strategies for CMA Members Operation in the PRD. (2013). Survey on 2013 CMA Members’ Views on the Business Environment in the PRD. Cohen, Youssef (1982). “‘The Benevolent Leviathan’: Political consciousness among urban workers under state corporatism.” In: American Political Science Review 76.1, pp. 46–59. Collier, Ruth Berins (1982). “Popular sector incorporation and political supremacy.” In: Brazil and Mexico: Patterns in late development. Ed. by Sylvia Ann Hewlett and Richard S Weinert. Institute for the Study of Human Issues, pp. 57–109. Collier, Ruth Berins and David Collier (1979). “Inducements versus constraints: Disaggregating ‘Corporatism’.” In: American Political Science Review 73.4, pp. 967–986. (2002). Shaping the Political Arena: Critical Junctures, the Labor Movement, and Regime Dynamics in Latin America. University of Notre Dame Press. Collier, Ruth Berins and James Mahoney (1997). “Adding collective actors to collective outcomes: Labor and recent democratization in South America and Southern Europe.” In: Comparative Politics 29.3, pp. 285–303. Conner, Neil (2016). “Three labour three rights activists sentenced in China amid worker unrest.” In: The Telegraph. Cooke, Fang Lee (2008). “The changing dynamics of employment relations in China: An evaluation of the rising level of labour disputes.” In: Journal of Industrial Relations 50.1, pp. 111–138. Cornelius, Wayne A. (1975). Politics and the Migrant Poor in Mexico City. Stanford University Press. “Creating a Modern Industrial System: PRD Prepares to Move Labor-intensive Sectors” (2007). In: Southern Daily.

240

References

Crothall, Geoffrey (2016). “Refusing to honor labor rights backfires on China.” In: New York Times. Cui, Chuanyi (2007). “On the paradigmatic changes of the peasant workers for in China.” In: Chinese Public Policy Review. Ed. by Kinglun Ngok and Weiqing Guo. Shanghai People’s Publisher, pp. 133–156. Dahlum, Sirianne, Carl Henrik Knutsen, and Tore Wig (2019). “Who revolts? Empirically revisiting the social origins of democracy.” In: The Journal of Politics 81.4, pp. 1494–1499. Daming, Zhou and Sun Xiaoyun (2014). “Research on “job hopping” by migrant workers from the countryside.” In: Chinese Sociology & Anthropology 43.2, pp. 51–69. Deyo, Frederic C (1989). Beneath the Miracle: Labor Subordination in the New Asian Industrialism. University of California Press. Dickson, Bruce J (2000). “Cooptation and corporatism in China: The logic of party adaptation.” In: Political Science Quarterly 115.4, pp. 517–540. (2017). “Cooptation and corporatism in China: The logic of party adaptation.” In: Critical Readings on the Chinese Communist Party. Leiden, The Netherlands: Brill, pp. 949–977. Dimitrov, Martin K, ed. (2013). Why Communism Did Not Collapse: Understanding Authoritarian Regime Resilience in Asia and Europe. Cambridge University Press. Du, Yang and Weiguang Pan (2009). “Minimum Wage Regulation in China and Its Applications to Migrant Workers in the Urban Labor Market.” In: China & World Economy 17.2, pp. 79–93. Easterly, William and Stanley Fischer (1995). “The Soviet Economic Decline.” In: The World Bank Economic Review 9.3, pp. 341–371. Easton, David (1979). A Systems Analysis of Political Life. Chicago University Press. Edin, Maria (2003). “Remaking the communist party-state: The cadre responsibility system at the local level in China.” en. In: China: An International Journal 1.1, pp. 1–15. Egorov, Georgy, Sergei Guriev, and Konstantin Sonin (2009). “Why resourcepoor dictators allow freer media: A theory and evidence from panel data.” en. In: American Political Science Review 103.4, pp. 645–668. Elfstrom, Manfred and Sarosh Kuruvilla (2014). “The changing nature of labor unrest in China.” In: ILR Review 67.2, pp. 453–480. Ellis, Richard (1994). Presidential Lightning Rods the Politics of Blame Avoidance. University Press of Kansas. Fan, C Cindy (1996). “Economic opportunities and internal migration: A case study of Guangdong Province, China*.” In: The Professional Geographer 48.1, pp. 28–45. (2007). China on the Move: Migration, the State, and the Household. Routledge. Fan, C Cindy and Chen Chen (2013). “The new-generation migrant workers in China.” In: Rural Migrants in Urban China. Ed. by Fulong Wu, Fangzhu Zhang, and Chris Webster. Routledge, pp. 41–59. Feigon, Lee (1990). China Rising: The Meaning of Tiananmen. Ivan R. Dee.

References

241

Feng, J (2008). “How action is possible.” PhD Thesis. Sun Yat-sen University. Feng, Wang (2010). “Boundaries of inequality.” In: One Country, Two Societies. Ed. by Martin K Whyte. Harvard University Press. Feng, Wang, Xuejin Zuo, and Danching Ruan (2002). “Rural Migrants in Shanghai: Living under the Shadow of Socialism.” In: International Migration Review 36.2, pp. 520–545. Ferree, Myra Marx and Frederick D Miller (1985). “Mobilization and meaning: Toward an integration of social psychological and resource perspectives on social movements*.” In: Sociological Inquiry 55.1, pp. 38–61. FHKI (2010). Honh Kong Manufacturing SMEs: Preparing for the Future. Fields, Gary S (1994). “Changing labor market conditions and economic development in Hong Kong, the Republic of Korea, Singapore, and Taiwan, China.” In: The World Bank Economic Review 8.3, pp. 395–414. Finkel, Steven E (1985). “Reciprocal effects of participation and political efficacy: A panel analysis.” In: American Journal of Political Science 29.4, p. 891. (1987). “The effects of participation on political efficacy and political support: Evidence from a West German Panel.” In: The Journal of Politics 49.2, pp. 441–464. Finkel, Steven E and Edward N Muller (1998). “Rational choice and the dynamics of collective political action: Evaluating alternative models with panel data.” In: American Political Science Review 92.1, pp. 37–49. Finkel, Steven E, Edward N Muller, and Karl-Dieter Opp (1989). “Personal influence, collective rationality, and mass political action.” In: The American Political Science Review 83.3, pp. 885–903. Franceschini, Ivan and Kevin Lin (2019). “Labour NGOs in China. From legal mobilisation to collective struggle (and Back?)” fr. In: China Perspectives. 1, pp. 75–84. Freire, Paulo (1968). Pedagogy of the Oppressed. Bloomsbury Publishing. (1974). Education for Critical Consciousness. Bloomsbury Publishing. Friedman, Eli (2014). Insurgency Trap: Labor Politics in Postsocialist China. Cornell University Press. Friedman, Eli and Ching Kwan Lee (2010). “Remaking the world of Chinese labour: A 30-year retrospective.” In: British Journal of Industrial Relations 48.3, pp. 507–533. Froissart, Chloé (2005). “The rise of social movements among migrant workers. Uncertain strivings for autonomy.” In: China Perspectives 5, pp. 1–15. (2017). “Negotiating authoritarianism and its limits: Worker-led collective bargaining in Guangdong Province.” In: China Information 32.1, pp. 23–45. Fu, Diana (2017). “Disguised collective action in China.” In: Comparative Political Studies 50.4, pp. 499–527. (2018). Mobilizing without the Masses: Control and Contention in China. Cambridge University Press. Fu, Qiang and Qiang Ren (2010). “Educational Inequality under China’s rural–urban divide: The hukou system and return to education.” In: Environment and Planning A 42.3, pp. 592–610.

242

References

Fu, Xiaolan and V N Balasubramanyam (2005). “Exports, foreign direct investment and employment: The case of China.” In: World Economy 28.4, pp. 607–625. Gallagher, Mary (2005). Contagious Capitalism: Globalization and the Politics of Labor in China. Princeton University Press. (2006). “Mobilizing the law in China: ‘informed disenchantment’ and the development of legal consciousness.” In: Law and Society Review 40.4, pp. 783–816. (2017). Authoritarian Legality in China. Cambridge University Press. Gallagher, Mary et al. (2015). “China’s 2008 labor contract law: Implementation and implications for China’s workers.” In: Human Relations 68.2, pp. 197– 235. Gamson, William A and Gadi Wolfsfeld (1993). “Movements and media as interacting systems.” In: The Annals of the American Academy of Political and Social Science 528.1, pp. 114–125. Gao, Qin, Sui Yang, and Shi Li (2017). “Social insurance for migrant workers in China: Impact of the 2008 labour contract law.” In: Economic and Political Studies 5.3, pp. 285–304. Geddes, Barbara (2003). Paradigms and Sand Castles. University of Michigan Press. Geddes, Barbara and John Zaller (1989). “Sources of popular support for authoritarian regimes.” In: American Journal of Political Science 33.2, p. 319. Gerschewski, Johannes (2013). “The three pillars of stability: legitimation, repression, and co-optation in autocratic regimes.” In: Democratization 20.1, pp. 13–38. Global Labor Strategies (2007). “Undue influence: Corporations gain ground in battle over China’s new labor law.” Globalization Monitor (2009). No Choice but to Fight. Globalization Monitor Limited. Goebel, Christopher (2019). “Social unrest in China: A bird’s eye perspective.” In: Handbook on Dissent and Protest in China. Ed. by Teresa Wright. Edward Elgar Publishing, pp. 27–45. Golden, Miriam A (1996). Heroic Defeats: The Politics of Job Loss. Cambridge University Press. Goldman, Merle and Roderick MacFarquhar, eds. (1999). The Paradox of China’s Post-Mao Reforms. en. Harvard University Press. Golley, Jane and Xin Meng (2011). “Has China run out of surplus labour?” In: China Economic Review 22.4, pp. 555–572. Gongchao (2020). “Jasic Movement” Debate – Chris Chan: Student activists who intervene in factories should adjust their steps once more, https://www.gongchao.org/2020/01/18/jasic-movement-debate-chris-chan/. (2020a). “Jasic Movement” Debate – Pun Ngai’s response to criticism: “Don’t let them sacrifice in vain,” https://www.gongchao.org/2020/01/17/jasicmovement-pun-ngais-response-to-criticism/. Gouldner, Alvin W (1954). Wildcat Strike. The Antioch Press.

References

243

Guo, Chengtu, Hua Li, and Helao Ding (1996). “Fujian province foreigninvested enterprises’ labor disputes: Special characteristics and countermeasures (fujianshengsanziqiyelaodongjiufendetedianjiduice).” In: Research and Suggestions 9, pp. 17–21. Gupta, Dipankar (2000). Interrogating Caste. Penguin Books India. Haas, Benjamin (2018). “Student activists detained in China for supporting workers’ rights.” In: The Guardian. Haggard, Stephan and Robert Kaufman (2008). Development, Democracy, and Welfare States. Latin America, East Asia, and Eastern Europe. Princeton University Press. Han, Dongfang (2015). “Witness statement.” Hao, Ren (2016). China on Strike: Narratives of Workers’ Resistance. Ed. by Zhongjin Li and Eli Friedman. Haymarket Books. Harney, Alexandra (2015). “China labor activists say facing unprecedented intimidation.” In: Reuters. Harney, Alexandra and John Ruwitch (2014). “In China, managers are the new labor activists.” In: Reuters. Hassid, Jonathan (2012). “Safety valve or pressure cooker? Blogs in Chinese political life.” In: Journal of Communication 62.2, pp. 212–230. Hernandez, Javier C (2018). “China’s leaders are confronted with an unlikely foe: Ardent young communists.” In: The New York Times. Hirschman, Albert O (1970). Exit, Voice, and Loyalty. Harvard University Press. Hoff, Karla and Priyanka Pandey (2004). Belief Systems and Durable Inequalities: An Experimental Investigation of Indian Caste. Policy Research Working Papers. The World Bank. “Honda Strike’s Deep Secrets: Revealing Pay Slips’ Blood and Tears (bentian bagongmen shendu jiemi: baoguang yuangong gongzibiao ningju xuelei)” (2010). In: National Business Daily. Hong, Shen (2010). “The mystery of China’s labor shortage.” In: The Wall Street Journal. Hood, Christopher (2010). The Blame Game. Princeton University Press. Hou, Y (2019). The private sector in public office: Selective property rights in China. Cambridge University Press. Howell, Jude (2015). “Shall we dance: Welfarist incorporation and the politics of state-labour NGO relations.” eng. In: China Quarterly 2015.223, pp. 702– 723. Hu, S (2010). “The rise of the business media in China.” In: Changing Media, Changing China. Ed. by Susan Shirk. Oxford University Press, pp. 77–90. Huang, Ping (2012). “Dismissal cost, employment and industry transformation and upgrading.” In: Nankai Economic Studies 3, pp. 79–94. Huang, Yasheng (1999). Inflation and Investment Controls in China. The Political Economy of Central-Local Relations during the Reform Era. Cambridge University Press. Huang, Yi, Mr Prakash Loungani, and Gewei Wang (2014). “Minimum wages and firm employment: Evidence from China.” In: IMF Working Paper 14–184.

244

References

Hui, Echo (2014). “Labour advocate helping Dongguan factory strike arrested over Wechat message.” In: South China Morning Post. Hui, Elaine Sio-Ieng and Chris Chan (2014). “The politics of labour legislation in southern China: How foreign chambers of commerce and government agencies influence collective bargaining laws.” In: International Labour Review 153.4, pp. 587–607. Human Rights Watch (2009). Not Yet a Workers’ Paradise. Huntington, Samuel P (1968). Political Order in Changing Societies. Yale University Press. Hurwitz, Jon and Mark Peffley (1987). “The means and ends of foreign policy as determinants of presidential support.” In: American Journal of Political Science 31.2, p. 236. Inglehart, Ronald and Pippa Norris (2017). “Trump and the populist authoritarian parties: The silent revolution in reverse.” en. In: Perspectives on Politics 15.2, pp. 443–454. Iyengar, Shanto (1990). “Framing responsibility for political issues: The case of poverty.” In: Political Behavior 12.1, pp. 19–40. (1994). Is Anyone Responsible? University of Chicago Press. “Japan-related books disappear in Beijing; Chinese demand pay hikes from Japanese employers” (2012). In: The Asahi Shimbun. Javeline, Debra (2003). “The role of blame in collective action: Evidence from Russia.” In: American Political Science Review 97.1, pp. 107–121. Ji, Siqi, Huifeng He, and Brian Peach (2021). “What is ‘lying flat’, and why are Chinese officials standing up to it?” en. In: South China Morning Post. Jung, Chauncey (2018). “Report: Chinese university graduates’ average annual starting salary is barely $8,000.” In: SupChina. Kahn, Joseph and David Barboza (2007). “As unrest rises, China broadens workers’ rights.” In: New York Times. Kamp, Denise van der, Peter Lorentzen, and Daniel Mattingly (2017). “Racing to the bottom or to the top? Decentralization, revenue pressures, and governance reform in China.” en. In: World Development 95, pp. 164–176. Kelloway, E Kevin et al. (2007). “Predicting protest.” In: Basic and Applied Social Psychology 29.1, pp. 13–22. Kim, Jun (1993). “Labor politics and labor movements in Asian authoritarian regimes: A comparative study on Korea and Taiwan.” PhD Thesis. Seoul National University. Kim, Wonik and Jennifer Gandhi (2010). “Coopting workers under dictatorship.” In: The Journal of Politics 72.3, pp. 646–658. King, Gary, Jennifer Pan, and Margaret E Roberts (2013). “How censorship in China allows government criticism but silences collective expression.” In: American Political Science Review 107.2, pp. 326–343. Klandermans, Bert (1984). “Mobilization and participation: Social-psychological expansions of resource mobilization theory.” In: American Sociological Review 49.5, p. 583. (1992). “The social construction of protest and multiorganizational fields.” In: Frontiers in Social Movement Theory. Ed. by Aldon D Morris and Carol McClurg Mueller. Yale University Press, pp. 77–103.

References

245

Knight, John, Quheng Deng, and Shi Li (2011). “The puzzle of migrant labour shortage and rural labour surplus in China.” In: China Economic Review 22.4, pp. 585–600. Knight, John and Li Shi (1996). “Educational attainment and the rural-urban divide in China.” In: Oxford Bulletin of Economics and Statistics 58.1, pp. 83–117. Knight, John, Lina Song, and Jia Huaibin (1999). “Chinese rural migrants in urban enterprises: three perspectives.” In: The Journal of Development Studies 35.3, pp. 73–104. Koo, Hagen (2001). Korean Workers. Cornell University Press. Koopmans, Ruud (2004a). “Movements and media: Selection processes and evolutionary dynamics in the public sphere.” In: Theory and Society 33.3–4, pp. 367–391. (2004b). “Protest in time and space: The evolution of waves of contention.” In: The Blackwell Companion to Social Movements. Ed. by David A Snow and Sarah A. Soule. Blackwell Publishing, pp. 19–46. Koopmans, Ruud and Susan Olzak (2004). “Discursive opportunities and the evolution of right-wing violence in Germany.” In: American Journal of Sociology 110.1, pp. 198–230. Krugman, Paul (1994). “The myth of Asia’s miracle.” In: Foreign Affairs 73.6, p. 62. Kuran, Timur (1991). “Now out of never: The element of surprise in the East European revolution of 1989.” In: World Politics 44.1, pp. 7–48. Kuruvilla, Sarosh (1996). “Linkages between industrialization strategies and industrial relations/human resource policies: Singapore, Malaysia, the Philippines, and India.” In: ILR Review 49.4, pp. 635–657. Kuruvilla, Sarosh and Hao Zhang (2016). “Labor unrest and incipient collective bargaining in China.” In: Management and Organization Review 12.1, pp. 159–187. Labor Bureau Research Team (2004). Report on the Investigation of Shortage of Peasant Farmers (guanyu mingong duanquede diaocha baogao). Tech. rep. Lardy, Nicholas R (1987). “The Chinese economy under stress, 1958–1965.” In: The Cambridge History of China. Ed. by John K Fairbank and Roderick MacFarquhar. Cambridge University Press, pp. 360–397. Lau, Lawrence J, Yingyi Qian, and Gerard Roland (2000). “Reform without losers: An interpretation of China’s dual-track approach to transition.” In: Journal of Political Economy 108.1, pp. 120–143. Lau, Richard R and David O Sears (1981). “Cognitive links between economic grievances and political responses.” In: Political Behavior 3.4, pp. 279–302. Lee, Ching Kwan (2002). “Revisiting the South China miracle.” In: Labor Studies Journal 27.2, pp. 61–64. (2007a). Against the Law: Labor Protests in China’s Rustbelt and Sunbelt. University of California Press. (2016). “Precarization or empowerment? Reflections on recent labor unrest in China.” In: The Journal of Asian Studies 75.2, pp. 317–333.

246

References

Lee, Ching Kwan and Yonghong Zhang (2013). “The power of instability: Unraveling the microfoundations of bargained authoritarianism in China.” In: American Journal of Sociology 118.6, pp. 1475–1508. Lee, Leng (2012). “Decomposing wage differentials between migrant workers and urban workers in urban China’s labor markets.” In: China Economic Review 23.2, pp. 461–470. Lee, Namhee (2007b). The Making of Minjung: Democracy and the Politics of Representation in South Korea. Cornell University Press. Lei, Ya-Wen (2017). The Contentious Public Sphere. Princeton University Press. Leung, Parry (2015). Labor Activists and the New Working Class in China: Strike Leaders’ Struggles. Springer. Li, Cheng (1996). “Surplus rural laborers and internal migration in China: Current status and future prospects.” In: Asian Survey 36.11, pp. 1122– 1145. Li, Gang (2010a). “Handa Nanhai Co. workers engage in a work stoppage for industrial disputes (bentian nanhai lingbujianchang gongren yingongzijiufen tinggong).” In: People’s Daily (renmin ribao). Li, Hongbin et al. (2012). “The end of cheap Chinese labor.” In: The Journal of Economic Perspectives 26.4, pp. 57–74. Li, Lianjiang (2010b). “Rights consciousness and rules consciousness in contemporary China.” In: The China Journal 64, pp. 47–68. Li, Lianjiang and Kevin O’Brien (2008). “Protest leadership in rural China.” In: The China Quarterly 193, pp. 1–23. Li, Qiang et al. (2013). “China’s labor transition and the future of China’s rural wages and employment.” In: China & World Economy 21.3, pp. 4–24. Li, Xiaoying and Richard B Freeman (2015). “How does China’s new labour contract law affect floating workers?” In: British Journal of Industrial Relations 53.4, pp. 711–735. Li, Yao (2018). Playing by the Informal Rules. Cambridge University Press. Liang, Z (1999). “Foreign investment, economic growth, and temporary migration: The case of Shenzhen special economic zone, China.” In: Development and Society 28.1, pp. 115–137. Lichbach, Mark Irving (1998). The Rebel’s Dilemma. University of Michigan Press. Liebman, Benjamin L (2010). “Changing media, changing courts.” In: Changing Media, Changing China. Ed. by Susan L Shirk. Oxford University Press, pp. 150–174. Lin, Justin Yifu, Fang Cai, and Zhou Li (2003). The China Miracle. Development Strategy and Economic Reform. Chinese University Press. Lipsky, Michael (1968). “Protest as a political resource.” In: The American Political Science Review, pp. 1144–1158. Lipton, M (1985). Capitalism and Apartheid 1910–1984. Rowman & Littlefield Publishers. Liu, Linping and Zhijian Guo (2004). “Firm character, government absence, collective bargaining and migrant female workers’ right protection.” In: Sociology Research 6, pp. 64–75.

References

247

Liu, Ye, Zhigang Li, and Werner Breitung (2012). “The social networks of new-generation migrants in China’s urbanized villages: A case study of Guangzhou.” In: Habitat International 36.1, pp. 192–200. Liu, Ying (2005). “690 yuan/month: Shenzhen Considers Substantially Increasing the Minimum Wage (690 yuan/yue: shenzhen lunzheng dafudu tiaozheng zuidigongzibiaozhun).” In: Chinese Journal of Sociology 6, pp. 64–75. Lohmann, Susanne (1994a). “Information aggregation through costly political action.” In: American Economic Review 84.3, pp. 518–530. (1994b). “The dynamics of informational cascades: The monday demonstrations in Leipzig, East Germany, 1989–91.” In: World Politics 47.1, pp. 42– 101. (2000). “Collective action cascades: An informational rationale for the power in numbers.” In: Journal of Economic Surveys 14.5, pp. 655–684. Longgang District Government Labor Bureau (2008). “Shenzhen longgang district government 2008 key point report.” Loong-Yu, Au (2019). The Jasic Mobilisation: A High Tide for the Chinese Labour Movement?, Made in China Journal, https://madeinchinajourn al.com/2019/01/12/the-jasic-mobilisation-a-high-tide-for-the-chinese-labourmovement/. Lorentzen, Peter (2013). “Regularizing rioting: Permitting public protest in an authoritarian regime.” In: International Quarterly Journal of Political Science 8.2, pp. 127–158. (2014). “China’s strategic censorship.” In: American Journal of Political Science 58.2, pp. 402–414. (2017). “Designing contentious politics in post-1989 China.” In: Modern China 43.5, pp. 459–493. Lorentzen, Peter and Suzanne Scoggins (2015). “Understanding China’s rising rights consciousness.” In: The China Quarterly 223, pp. 638–657. Loury, Glenn C and G C Loury (2009). The Anatomy of Racial Inequality. Harvard University Press. Lu, Chuang, Siyuan Tang, and Guanmin Liao (2015). “Labor protection, labor intensity and investment efficiency.” In: Accounting Research 6, pp. 42–47. Lu, Feng (2012). “Wage trends among migrant workers in China: 1979–2010 (zhongguo nongmingong gongzi zoushi).” In: Chinese Sociology 7, pp. 47– 68. Lü, Xiaobo (2014). “Social policy and regime legitimacy: The effects of education reform in China.” In: The American Political Science Review 108.2, pp. 423– 437. Lüthje, Boy, Hao Zhang, and Siqi Luo (2010). “Interview with Chen Weiguang, Chairman of the Guangzhou Federation of Trade Unions.” In: Global Labour Column. Ma, Changbo (2009). “The large scale of ‘double relocation’ only just started (daguimode ‘shuangzhuanyi’ caiganggang kaishi).” In: Southern Weekly. Mansfield, Edward D. and Diana C. Mutz (2009). “Support for free trade: Selfinterest, sociotropic politics, and out-group anxiety.” en. In: International Organization 63.3, pp. 425–457.

248

References

Marwell, Gerald and Pamela Oliver (1993). The Critical Mass in Collective Action. Cambridge University Press. Mayhew, David R. (1974). Congress: The Electoral Connection. Yale University Press. McAdam, Doug (1982). Political Process and the Development of Black Insurgency, 1930–1970. The University of Chicago Press. (1988). “Micromobilization contexts and recruitment to activism.” In: International Social Movement Research 1, pp. 125–154. McCarthy, John D, Clark McPhail, and Jackie Smith (1996). “Images of protest: Dimensions of selection bias in media coverage of Washington demonstrations, 1982 and 1991.” In: American Sociological Review 61.3, pp. 478– 499. McCubbins, Mathew D and Thomas Schwartz (1984). “Congressional oversight overlooked: Police patrols versus fire alarms.” In: American Journal of Political Science 28.1, p. 165. McGraw, Kathleen M (1990). “Avoiding blame: An experimental investigation of political excuses and justifications.” In: British Journal of Political Science 20.1, pp. 119–131. Meng, Xin and Nansheng Bai (2007). “How much have the wages of unskilled workers in China increased?” In: China: Linking Markets for Growth. Ed. by Ross Garnaut and Ligang Song. ANU E Press, pp. 151–175. “Migrant workers contribute 16 percent GDP growth” (2006). In: China Daily. “Migrants make up seventy percent of labor dispute cases (laodong zhengyi anjian nongmingong zhan qicheng)” (2011). In: Beijing Youth Daily. Monroe, Kristen R. (1979). “Econometric analyses of electoral behavior: A critical review.” In: Political Behavior 1.2, pp. 137–173. Montinola, Gabriella, Yingyi Qian, and Barry R Weingast (1995). “Federalism, Chinese style: The political basis for economic success in China.” In: World Politics 48.01, pp. 50–81. Moore, Barrington (1966). Social Origins of Dictatorship and Democracy. Beacon Press. Mosley, Layna and Saika Uno (2007). “Racing to the bottom or climbing to the top? Economic globalization and collective labor rights.” In: Comparative Political Studies 40.8, pp. 923–948. Mouzelis, Nicos (1985). “On the concept of populism: Populist and clientelist modes of incorporation in semiperipheral polities.” In: Politics & Society 14.3, pp. 329–348. Muller, Edward N, Henry A Dietz, and Steven E Finkel (1991). “Discontent and the expected utility of rebellion: The case of Peru.” In: American Political Science Review 85.4, pp. 1261–1282. Nang, Leung Pak and Ngai Pun (2009). “The radicalisation of the new Chinese working class: A case study of collective action in the gemstone industry.” In: Third World Quarterly 30.3, pp. 551–565. Naughton, B J (2006). The Chinese Economy: Transitions and Growth. The MIT Press.

References

249

Ngok, Kinglun (2008). “The changes of Chinese labor policy and labor legislation in the context of market transition.” In: International Labor and WorkingClass History 73.1, pp. 45–64. (2011). China in Transition Labor Issues and Labor Policies (zhuanxingqide zhongguo laodong wenti yu laodong zhengce). Oriental Publishing Center. Ngok, Kinglun and Joseph YS Cheng (2010). “The role of migrant labour in China’s labour policy.” In: The Journal of Comparative Asian Development 9.2, pp. 347–368. Ngok, Kinglun and Wenjia Zhuang (2009). “Labor inspection regime in transition in contemporary China.” In: Journal of Public Administration Review 5, pp. 92–120. O’Brien, Kevin and Lianjiang Li (1999). “Selective policy implementation in rural China.” In: Comparative Politics 31.2, p. 167. (2006). Rightful Resistance in Rural China. Cambridge University Press. O’Donnell, Guillermo A (1988). Bureaucratic Authoritarianism: Argentina, 1966–1973, in Comparative Perspective. University of California Press. Oi, Jean C (1992). “Fiscal reform and the economic foundations of local state corporatism in China.” In: World Politics 45.1, pp. 99–126. (1999). Rural China Takes Off. Institutional Foundations of Economic Reform. University of California Press. Oliver, Pamela, Gerald Marwell, and Ruy Teixeira (1985). “A theory of the critical mass. I. Interdependence, group heterogeneity, and the production of collective action.” In: American Journal of Sociology 91.3, pp. 522–556. Oliver, Pamela and Daniel J Myers (1999). “How events enter the public sphere: Conflict, location, and sponsorship in local newspaper coverage of public events 1.” In: American Journal of Sociology 105.1, pp. 38–87. Olson, Mancur (1965). The Logic of Collective Action. Public Goods and the Theory of Groups. Harvard University Press. Opp, Karl-Dieter (1998). “Does antiregime action under communist rule affect political protest after the fall?: Results of a panel study in east Germany.” In: The Sociological Quarterly 39.2, pp. 189–213. (2009). Theories of Political Protest and Social Movements: A Multidisciplinary Introduction, Critique, and Synthesis. Routledge. Ouhai District (2012). “Ouhai district on how to deal with labor mass incidents.” Page, Benjamin I and Richard A Brody (1972). “Policy voting and the electoral process: The Vietnam war issue.” In: American Political Science Review 66.3, pp. 979–995. Palley, Thomas I (2005). “Labour standards, democracy and wages: some crosscountry evidence.” In: Journal of International Development 17.7, pp. 883– 898. Pei, M (2010). “Rights and resistance: The changing contexts of the dissident movement.” In: Chinese Society: Change, Conflict and Resistance. Ed. by Elizabeth J. Perry and Mark Selden. Routledge, pp. 49–74. Peng, Thomas (2011). “The impact of citizenship on labour process: State, capital and labour control in South China.” In: Work, Employment and Society 25.4, pp. 726–741.

250

References

Perry, Elizabeth J (1994). “Casting a Chinese ‘Democracy’ Movement: The roles of students, workers, and entrepreneurs.” In: Popular Protest and Political Culture in Modern China. Ed. by Jeffrey N Wasserstrom and Elizabeth J Perry. Westview Press. (2002). Challenging the Mandate of Heaven: Social Protest and State Power in China. M.E. Sharpe. (2008). “Chinese conceptions of ‘rights’: from Mencius to Mao – and now.” In: Perspectives on Politics 6.01, pp. 37–50. (2010). “Popular protest in China: Playing by the rules.” In: China Today, China Tomorrow: Domestic Politics, Economy and Society. Ed. by Joseph Fewsmith. Rowman & Littlefield, pp. 11–28. Piven, Frances Fox and Richard Cloward (1979). Poor People’s Movements. Vintage Books. PRC Bureau of Statistics (2011). Size, Structure, and Characteristics of the New Generation of Peasant Workers (xinshengdai nongmingongde shuliang jiegou he tedian). Tech. rep. (2014). “2013 Report on Peasant Worker Monitoring Investigation.” (2019). “2019 Report on Peasant Worker Monitoring Investigation.” Pribble, Jennifer (2010). “Worlds apart: Social policy regimes in Latin America.” In: Studies in Comparative International Development 46.2, pp. 191–216. Pringle, Tim (2011). Trade Unions in China : The Challenge of Labour Unrest. 1st ed. The Challenge of Labour Unrest. Routledge. Przeworski, Adam and Michael E Alvarez (2000). Democracy and Development. Cambridge University Press. Pun, Ngai (1999). “Becoming Dagongmei (Working Girls): The politics of identity and difference in reform China.” en. In: The China Journal 42, pp. 1–18. (2005). Made in China. Women Factory Workers in a Global Workplace. Duke University Press. Pun, Ngai and Jenny Chan (2013). “The spatial politics of labor in China: Life, labor, and a new generation of migrant workers.” In: South Atlantic Quarterly 112.1, pp. 179–190. Pun, Ngai and Lu Huilin (2010). “Unfinished proletarianization: Self, anger, and class action among the second generation of peasant-workers in present-day China.” In: Modern China 36.5, pp. 493–519. Pun, Ngai and Chris Smith (2007). “Putting transnational labour process in its place: the dormitory labour regime in post-socialist China.” In: Work, Employment and Society 21.1, pp. 27–45. Pye, Lucian W (1990). “Tiananmen and Chinese political culture: The escalation of confrontation from moralizing to revenge.” In: Asian Survey 30.4, pp. 331–347. Qian, Xiaolei and Russell Smyth (2008). “Measuring regional inequality of education in China: widening coast–inland gap or widening rural–urban gap?” In: Journal of International Development 20.2, pp. 132–144. Ramakrishnan, S. Karthick (2013). “Incorporation versus assimilation: The need for conceptual differentiation.” In: Outsiders No More Models of Immigrant Political Incorporation. Ed. by Jennifer Hochschild et al. Oxford University Press, pp. 27–42.

References

251

Rapoza, Kenneth (2011). “In coastal China, a labor shortage.” In: Forbes. Repnikova, Maria (2017a). “Media openings and political transitions: Glasnost versus Yulun Jiandu.” In: Problems of Post-Communism 64.3–4, pp. 141– 151. (2017b). Media Politics in China. Improvising Power under Authoritarianism. Cambridge University Press. Rho, Sungmin and Michael Tomz (2017). “Why don’t trade preferences reflect economic self-interest?” en. In: International Organization 71.S1, pp. 85– 108. Roberts, Dexter (2014). “China’s officials muzzle labor activists.” In: Bloomberg Businessweek, 9 May. Roberts, Margaret (2018). Censored. Distraction and Diversion Inside China’s Great Firewall. Princeton University Press. Robertson, G B and E Teitelbaum (2011). “Foreign direct investment, regime type, and labor protest in developing countries.” In: American Journal of Political Science 55.3, pp. 665–677. Robertson, Graeme B. (2010). The Politics of Protest in Hybrid Regimes: Managing Dissent in Post-Communist Russia. Cambridge University Press. Rodrik, Dani (1999). “Democracies pay higher wages.” In: The Quarterly Journal of Economics 114.3, pp. 707–738. Rueschemeyer, Dietrich, Evelyne Huber Stephens, and John D Stephens (1992). Capitalist Development and Democracy. Polity Press. Schatz, Edward (2009). “The soft authoritarian tool kit: Agenda-setting power in Kazakhstan and Kyrgyzstan.” In: Comparative Politics 41.2, pp. 203–222. Schmitter, Philippe C (1971). Interest Conflict and Political Change in Brazil. Stanford University Press. Scott, James C. (1985). Weapons of the Weak: Everyday Forms of Peasant Resistance. Yale University Press. Seidman, Gay W (1994). Manufacturing Militance. University of California Press. Sevastopulo, Demetri (2014). China Charges Labour Activist after Yue Yuen Shoe Factory Strike. Sharma, Yojana (2014). “What do you do with millions of extra graduates?” In: BBC News. Shenzhen Daily (2005). “Guangdong faces 1-million labor shortage.” Shenzhen Foreign Invested Firm Association (2012). “Additional report on the effect of protest, strike, and production halt that were caused by diaoyu island problem on Japanese firms.” In: Foreign Business Reflect (waishangfanying) 156.3. Shenzhen Government Research Team (2005). “Empirical estimation and analysis of foreign export benefits in Shenzhen (shenzhen waimaochukou xiaoyi shizhengcesuan yu fenxi).” In: Special Zone Economy (tequ jingji) 7, pp. 12–15. (2013). Report on 10-Year Development of Finance and Tax Economy in Shenzhen (2001–2010). China Financial Economy Publishing House. “Shenzhen wants to drive away labor-intensive firms (Shenzhen yao bizou laodongmijiqiye)” (2005). In: Southern Metropolis Daily.

252

References

Sherkat, Darren E and T Jean Blocker (1997). “Explaining the political and personal consequences of protest.” In: Social Forces 75.3, pp. 1049–1070. Shirk, Susan (2007). “Changing media, changing foreign policy in China.” In: Japanese Journal of Political Science 8.01, pp. 43–70. (2010). Changing Media, Changing China. Oxford University Press. Shirouzu, Norihiko (2010). “Chinese workers challenge Beijing’s authority.” In: The Wall Street Journal, June 2010. Shorter, Edward and Charles Tilly (1974). Strikes in France 1830–1968. CUP Archive. Silver, Beverly (2003). Forces of Labor. Workers’ Movements and Globalization Since 1870. Cambridge University Press. Siu, Helen F (1990). “The politics of migration in a market town.” In: Chinese Society on the Eve of Tiananmen. Ed. by Deborah Davis and Ezra F Vogel. Council on East Asian Studies. Smith, Chris and Ngai Pun (2006). “The dormitory labour regime in China as a site for control and resistance1.” In: The International Journal of Human Resource Management 17.8, pp. 1456–1470. Snow, David A and Pamela Oliver (1995). “Social movements and collective behavior: Social psychological dimensions and considerations.” In: Sociological Perspectives on Social Psychology. Ed. by Karen S Cook, Gary Alan Fine, and James S House, pp. 571–599. Snyder, David and William R Kelly (1977). “Conflict intensity, media sensitivity and the validity of newspaper data.” In: American Sociological Review 42.1, pp. 105–123. Solinger, Dorothy J (1999). Contesting Citizenship in Urban China: Peasant Migrants, the State, and the Logic of the Market. University of California Press. Song, Ho-keun (1990). “Labor politics and the development of opposition movement in authoritarian capitalism.” In: Asian Culture 6, pp. 37–68. Steel, R. P. and N. K. Ovalle (1984). “A review and meta-analysis of research on the relationship between behavioral intentions and employee turnover.” In: Journal of applied psychology 69.4, pp. 673–686. Stepan, Alfred C (1978). The State and Society: Peru in Comparative Perspective. Princeton University Press. Stockmann, Daniela (2012). Media Commercialization and Authoritarian Rule in China. Cambridge University Press. Stockmann, Daniela and Mary Gallagher (2011). “Remote control: How the media sustain authoritarian rule in China.” In: Comparative Political Studies 44.4, pp. 436–467. Su, Yang and Xin He (2010). “Street as courtroom: State accommodation of labor protest in South China.” In: Law and Society Review 44.1, pp. 157–184. Sun, Zhongwei and Binbin Shu (2011). “The standard of the minimum wage, and the wages of peasant workers (zuidigongzibiaozhun yu nongmingong gongzi).” In: Management World 8, pp. 45–56. Tang, Wenfang (2016). Populist Authoritarianism. Oxford University Press.

References

253

Tarrow, Sidney and Charles Tilly (2009). “Contentious politics and social movements.” In: The Oxford Handbook of Comparative Politics. Ed. by Carles Boix and Susan C Stokes. Oxford University Press. Teitelbaum, Emmanuel (2010). “Mobilizing restraint: Economic reform and the politics of industrial protest in South Asia.” In: World Politics 62.4, pp. 676–713. (2011). Mobilizing Restraint. Cornell University Press. “The effect of the new labor contract law on labor costs in different types of firms (xinlaodonghetongfa dui butongqiue leixing gongzichengbende yingxiang)” (2008). In: China Business News. “There won’t be hollowing out in PRD (zhusanjiao buhui chuxian kongxinhua)” (2009). In: Yangcheng Evening News. Thomson, Henry (2018). “Grievances, mobilization, and mass opposition to authoritarian regimes.” In: Comparative Political Studies 51.12, pp. 1594– 1627. Tian, Feng and Kaixuan Lin (2020). Why Would Not Want to Go Back. Dolphin Books. Tian, Ming and Lei Xu (2015). “Investigating the job mobility of migrant workers in China.” In: Asian and Pacific Migration Journal 24.3, pp. 353–375. Tilley, James and Sara B Hobolt (2015). “Is the government to blame? An experimental test of how partisanship shapes perceptions of performance and responsibility.” In: The Journal of Politics 73.2, pp. 316–330. Tilly, Charles (1977). From Mobilization to Revolution. University of Michigan. Tong, Yanqi and Shaohua Lei (2013). Social Protest in Contemporary China, 2003–2010 : Transitional Pains and Regime Legitimacy. 1st ed. Transitional Pains and Regime Legitimacy. Routledge. Tse, Chun Wing (2016). “Urban residents’ prejudice and integration of rural migrants into urban China.” In: Journal of Contemporary China 25.100, pp. 579–595. Turner, Ralph H (1981). “Collective behavior and resource mobilization as approaches to social movements: Issues and continuities.” In: Research in Social Movements, Conflicts and Change 4. Ed. by Louis Kriesberg. JAI Greenwich, pp. 1–24. UNDP (2005). China Human Development Report 2005: Development with Equity. China Development Research Foundation. Unger, Jonathan and Anita Chan (1995). “China, corporatism, and the East Asian model.” In: The China Journal 33, pp. 29–53. Valenzuela, J Samuel (1989). “Labor movements in transitions to democracy: A framework for analysis.” In: Comparative Politics 21.4, pp. 445–472. Vanhuysse, Pieter (2006). Divide and Pacify: Strategic Social Policies and Political Protests in Post-communist Democracies. Central European University Press. Verba, S and G Almond (1963). The Civic Culture: Political Attitudes and Democracy in Five Nations. Volden, Craig and Alan E Wiseman (2007). “Bargaining in legislatures over particularistic and collective goods.” In: The American Political Science Review 101.1. Publisher: American Political Science Association, Cambridge University Press, pp. 79–92.

254

References

Walder, A G and Xiaoxia Gong (1993). “Workers in the Tiananmen protests: The politics of the Beijing Workers’ Autonomous Federation.” In: The China Journal 29, pp. 1–29. Walder, Andrew G (1988). Communist Neo-Traditionalism. University of California Press. Wallace, Jeremy (2014). Cities and Stability: Urbanization, Redistribution, and Regime Survival in China. Oxford University Press. (2015). “Cities, redistribution, and authoritarian regime survival.” In: The Journal of Politics 75.3, pp. 632–645. Wallace, Jeremy and Jessica Chen Weiss (2015). “The Political Geography of Nationalist Protest in China: Cities and the 2012 Anti-Japanese Protests.” In: The China Quarterly 222, pp. 403–429. Walsh, Megan (2017). “China’s migrant worker poetry.” In: The Asia-Pacific Journal 15.12, pp. 1–10. Wan, Xiangdong, Linping Liu, and Yonghong Zhang (2006). “Wages and benefits, protection of rights and interests, and the external environment: a comparative study of migrant workers in the Pearl River Delta and the Yangtze River Delta.” In: Management World 6, pp. 37–45. Wang, Feiling (2005). Organizing through Division and Exclusion: China’s Hukou System. Stanford University Press. Wang, Juan (2015). “Managing social stability: The perspective of a local government in China.” In: Journal of East Asian Studies 15.1, pp. 1–25. Wang, Meiyan (2010). “The rise of labor cost and the fall of labor input: Has China reached Lewis turning point?” In: China Economic Journal 3.2, pp. 137–153. Wang, Mi (2008). “Dongguan industrial restructuring is not driving factory and people out (dongguan chanyetiaozheng bingfei “ganchangganren”).” In: Information Times. Wang, Qingfeng (2021). “‘Lying flat’ is shameful. Where is the sense of justice?” In: Nanfang Daily. Weaver, R Kent (1986). “The politics of blame avoidance.” In: Journal of Public Policy 6.4, pp. 371–398. Wei, Shengmin, Jing Xiang, and Liping Wen (2016). “Guangdong economic change from the perspective of tax revenues: 2011–2015.” In: Guangdong Economy 0.4, pp. 83–89. Wen, Dongwei, Guoming Xian, and Jing Ma (2009). “FDI.” In: Management World 4, pp. 96–107. Wen, X (2012). “Transformation and government management of workers’ collective action (gongren quntixing shijiande zhuanxing yu zhengfuzhili).” In: China Human Resource Development zhongguorenliziyuankaifa 5, pp. 81–85. “Wen Prime Minister: Treat migrant workers as if you are treating your own children (wenzongli: xiangduidai zijide haiziyiyang duidai mingong)” (2010). In: Yicaiwang. White, Gordon, Jude A. Howell, and Shang Xiaoyuan (1996). In Search of Civil Society: Market Reform and Social Change in Contemporary China. Oxford University Press.

References

255

Whiting, Susan H (2004). “The cadre evaluation system at the grass roots.” In: Holding China Together. Ed. by Barry J Naughton and Dali Yang. Cambridge University Press, pp. 101–119. Whyte, Martin K (2010). Myth of the Social Volcano. Stanford University Press. Wintrobe, Ronald (1998). The Political Economy of Dictatorship. Cambridge University Press. Wolfsfeld, Gadi (1997). Media and Political Conflict: News from the Middle East. Cambridge University Press. Wolpe, Harold (1972). “Capitalism and cheap labour-power in South Africa: From segregation to apartheid 1.” In: Economy and Society 1.4, pp. 425– 456. Wong, Linda (2011). “Chinese migrant workers: Rights attainment deficits, rights consciousness and personal strategies.” In: The China Quarterly 208, pp. 870–892. Woo, W T (1996). “Chinese economic growth: Sources and prospects.” In: The Chinese Economy. Ed. by Michel Fouquin and Francoise Lemoine. Economica, pp. 17–34. Wood, Elisabeth Jean (2000). Forging Democracy from Below: Insurgent Transitions in South Africa and El Salvador. Cambridge University Press. Wood, Ellen Meiksins (1995). “The Separation of the ‘economic’ and ‘political’ in capitalism.” In: Democracy against Capitalism: Renewing Historical Materialism. Cambridge University Press, pp. 19–48. Wright, Erik Olin (2000). “Working-class power, capitalist-class interests, and class compromise.” In: American Journal of Sociology 105.4, pp. 957–1002. Wright, Teresa (2010). Accepting Authoritarianism. Stanford University Press. Xiang, Qing, Guoying Sun, and Yanling Qin (2005). “A retrospective analysis of local tax revenue of Shenzhen in the past ten years (shenzhen dishui shuishuoushouru shinian huigu yu pouxi).” In: International Taxation 6, pp. 39–43. Xu, Yi (2013). “Labor non-governmental organizations in China: Mobilizing rural migrant workers.” In: Journal of Industrial Relations 55.2, pp. 243– 259. Yang, Dennis Tao and Fang Cai (2003). “The political economy of China’s ruralurban divide.” In: How Far across the River? Ed. by Nicholas C Hope, Dennis Tao Yang, and Mu Yang Li. Stanford University Press. Yang, Lin (2011). “Difficulty of labor inspection (laodong jiancha zhikun).” In: Outlook Weekly. Ye, Linxiang, T H Gindling, and Shi Li (2015). “Compliance with legal minimum wages and overtime pay regulations in China.” In: IZA Journal of Labor & Development 4.1, p. 16. Yeung, Wei-Jun Jean (2014). “Higher education expansion and social stratification in China.” In: Chinese Sociological Review 45.4, pp. 54–80. Young, Joe (2002). “Hukou reform targets urban-rural divide.” In: The China Business Review 29.3, pp. 32–33. Yu, Jianrong (2010). “The origins of mass incidents lie in the conflicts between government and citizens (quntixing shijian zhuangjie zaiyu gongmin maodun).” In: Zhongguo Baodao 2, pp. 50–51.

256

References

Yu, Ya-ya and Guanmin Liao (2017). “Labor protection, inudstry downturn and firms’ bank laon default.” In: Journal of Central University of Finance & Economics 9, pp. 43–52. Zetka Jr, James R (1992). “Work organization and wildcat strikes in the US automobile industry, 1946 to 1963.” In: American Sociological Review 57.2, pp. 214–226. Zhan, Shaohua (2011). “What determines migrant workers’ life chances in contemporary China? Hukou, social exclusion, and the market.” In: Modern China 37.3, pp. 243–285. Zhang, Huafeng (2010). “The Hukou system’s constraints on migrant workers’ job mobility in Chinese cities.” In: China Economic Review 21.1, pp. 51–64. Zhang, Kevin Honglin and Shunfeng Song (2001). “Promoting exports: The role of inward FDI in China.” In: China Economic Review 11.4, pp. 385–396. Zhang, Li (2001). Strangers in the City. Stanford University Press. Zhang, Lu (2014). Inside China’s Automobile Factories. Cambridge University Press. Zhao, Yuezhi and Sun Wusan (2007). “Public opinion supervision: Possibilities and limits of the media in constraining local officials.” In: Grassroots Political Reform in Contemporary China. Ed. by Elizabeth J Perry and Merle Goldman. Harvard University Press, pp. 300–326. (2009). “Public opinion supervision: Possibilities and limits of the media in constraining local officials.” In: Grassroots Political Reform in Contemporary China. Ed. by Elizabeth J. Perry and Merle Goldman. Harvard University Press, pp. 300–326. Zhejiang Bureau of Statistics (2013). “Replacing People with Machines” Promotes Upgrading of Zhejiang Job Structure. Zheng, Gongcheng and Liruolian Huang (2007). Rural-Urban Migrant Workers in China (zhongguo nongmingongwenti yu shehuibaohu). Issue and Social Protection. Renmin Chubanshe. Zhou, Xueguang (1993). “Unorganized interests and collective action in communist China.” In: American Sociological Review 58, pp. 54–73. Zhu, Shengjun and John Pickles (2013). “Bring in, go up, go west, go out: Upgrading, regionalisation and delocalisation in China’s apparel production networks.” In: Journal of Contemporary Asia 44.1, pp. 36–63. Zhuang, Wenjia and Kinglun Ngok (2014). “Labour inspection in contemporary China: Like the Anglo-Saxon model, but different.” In: International Labour Review 153.4, pp. 561–585.

Index

All China Federation of Trade Unions (ACFTU), 9–10, 48–49 role in legal mobilization, 67, 163 urban bias, 63 Apartheid, 7, 26 comparisons to the hukou system, 32–34, 44, 84 atomized protest, see also protest, 10, 12 as a tool of material co-optation, 2, 57–58 changing nature of, 217–218 tolerance of, 1, 69–71 atomization, see also particularization, atomized Incorporation, repression, 3 as a labor control strategy, 50–51 challenges to, 219–227 cognitive implications of, 59 in China, 3, 55–56 atomized incorporation: definition, see atomization, incorporation, particularization, blame 3–4, 9, 54 advantages and limitations, 3–4, 11–12, 16–17, 54, 56, 156 as a strategy of blame avoidance, 102–105 characteristics, 11 comparisons to other strategies, 9 implications, 19, 227–229 political economy explanations for, 77–93 relationship between atomization and incorporation, 4, 74–75 risk, 61–62, 162

authoritarian regimes: information problem, see also co-optation, corporatism, repression, 16–17, 57–58 economic growth and resilience of, 6, 24–25 labor coercive growth in, 26–27 tolerance of protests, 1–2, 57–58 automation, 85, 200 anti-foreign protests, 207 Brazil, 24, 27, 44 blame attribution: internal, see also depoliticization; politicization, 100, 102, 107–109 firm-level pattens of, 182–184 of experienced workers, 110 of younger workers, 109–110 protest intention and, 151–154 survey data, 105–122 to the central government, 106–107 blame avoidance: strategy, 100–101 Cai, Yongshun, 72, 102, 128, 144 cellular activism, 2, 4 Chan, Chris, 60, 129, 209, 222 Chen, Xi, 18, 60, 186 China Labour Bulletin, 161, 217, 218 relations with mainland labor NGOs, 224–225 strike map data, 172–176, 179–181, 217–218 China’s rise, see also economic development, 6

257

258

Index

aftermath, 36–39, 43–44, 214 declining economic efficiency, 8, 77–78 its benefits, 6 marginalization of migrant workers, 6–7, 35 migrant workers’ contribution to, 31 co-optation: political, 52 institutional, 52 material, see material co-optation of unorganized groups, 52–54 collective action problems, see protest, 17 free riding, 125–126, 146, 155 in unorganized groups, 126–127 solutions, 127–128 collective action resources: mobilizational resources, see also mobilization, 127–129 collective bargaining: labor NGOs and, see also labor non-governmental organizations, 220–221 local governments’ attitudes toward, 221–222 corporatism: state corporatism, 47 exclusionary, 48 inclusionary, 48

effect on protest intention, 147, 150 political implications, 58 economic protest, see atomized protest efficacy, see individual efficacy, group efficacy environmental pollution: attitudes of local government, 86 exit: from the firm, see also vote with feet, Hirschman, labor turnover, 141 from the society, 229 export-oriented industrialization, 27 China, 7, 30, 36 declining exports, 81 employment of migrants, 30 labor coercion and, 35 labor shortages, 36–38 expressive benefits: definition, 144 for rank-and-file workers, 149 of joining protests, 145 extra-legal demands, see also legal demands, interest-based protests, 161 effect on protest tactics, 177–178, 199–202, 205 increase of, 164 local governments’ view of, 165–166

dagong. See migrant workers, 7, 42 frustrations of younger migrant workers about, 98, 112 democratization: role of labor, 76–77 depoliticization, see also politicization, 5–6, 58–61 as a strategy of blame avoidance, 96 Diaoyu-Senkaku dispute: labor protests and, 207 discursive opportunities: definition, 187 at anti-foreign protests, 210–211 bargaining power and, 186, 200–201 media and, 187, 195, 202 perceived group efficacy and, 193

fire alarm system, see also particularization, 11, 12, 67 firm embeddedness, 17 effect on job turnover intention, 142–143 effect on protest intention, 149 firm-specific training, 141 firm retribution, see repression firm tenure, 150, 151 protest intention and, 204–205 social network and, 203–204 firm-level patterns of protest: effect of ownership, 167–168, 172–176 effect of size, 169, 172–176 law-abiding firms and, 169–171 foreign direct investment, see also Opening up; foreign-owned firms, 31 labor coercion and, 28 labor disputes and, 167–168 foreign-owned firms: employment of migrant workers at, 31 higher incidence of labor disputes at, 167–168 interest-based protests at, 171–176, 208–210

economic development: in authoritarian regimes, see also export-oriented industrialization; labor-intensive production, 24, 26–27 authoritarian resilience and, 6, 24–25 sustainable development, 78 economic grievances: job-specific, see also social grievances; politicization, 13, 99 about low wages, 138–139 about wage arrears, 68 definition, 111

Index law-abiding behavior, 92, 167, 191–192, 200 reputational concerns of, 192–193, 200 Four Asian Tigers, 24, 44 Foxconn, 192 media reports of, 192 Gallagher, Mary, 68, 110, 167, 184 gender: effect on social grievances, 112–114 generalization: generalized framing, see also social grievances, politicization, particularization, 103 role of migration experience, 114–116 global value chains: labor coercion and, 27–28 collective action at a firm of, 199–202 reputational concerns, 191–193, 200 workers’ expectation of media exposure at, 196–198, 202 group efficacy: of SOE workers, 144 definition, 144 discursive opportunities and, 193–195 effect of protest experience on, 158 effect on protest intention, 149, 154 law-abiding firms and, 193 measurement of perception, 145–146 of migrant workers, 144–145 Henan, 20 Hirschman, 141 Honda Nanhai: labor strike at, 1, 210 Hong Kong; mainland labor activism and, 224 hukou system, see also Apartheid, migrants, 5, 7 changes, 63–64 consequence for migrants, 32–36, 97 history, 28–30, 32 incorporation, see also atomized incorporation, 54–55 industrial relocation: double relocation in Guangdong, see also industrial upgrading, 79–80 discouraging migration, 85–86 in Jiangsu, 80 in Zhejiang, 86 industrial upgrading, see also export-oriented industrialization; industrial relocation, 78 cases, 86–87, 89–90

259

effects on local tax revenues in Guangdong, 81–83 measures of local governments, 79–80 specific criteria for elimination and upgrading, 89–90 use of environmental policies, 88, 89 use of labor policies, 87–92 inequality: workers’ perception of, see also migrant workers, 111 information: protest as a source of, see also atomization, politicization, blame, generalization, protest, 2, 16–17, 19, 57–58, 120–122 about other protests, 201 atomization and, 59 blame attribution and, 100–101, 110, 114–116 control of, in factories, 131 interest-based protests, 18, 161, 164 anti-foreign sentiment and, 208 cases, 199–202, 208–210 for higher wages, 171–176 importance of leadership, 128–129 in Shenzhen, 164 local authorities’ handling of, 167 media reports of, 192–193, 200–202 police presence at, 179–181 political implications, 18, 161–162 strike map stats, 172 survey stats, 170–171 tactics of, 165–166, 177–178 Jasic incident, 75, 219, 225–227 job availability: of different workers, see also structural bargaining power, labor turnover, labor shortage, job hopping, 142 effect on protest intention, 149 increases in export-oriented regions, 36–37 job hopping: of different workers, 141–142 effect on social network, 143 job tenure: job turnover and, see also social network, job turnover, exit, 142 labor activists: repression of, 72–74, 216, 219–222, 224–225 external mobilization by, 221–227 student activists, 75, 225–226

260

Index

labor assertiveness, see also extra-legal demands, interest-based protests, structural bargaining power, 8 labor coercion: definition, 25 comparison to labor repression, 25–26 consequences, 43–45 Labor Contract Law, 65, 163, 200 labor control in authoritarian regimes, 47–51 labor grievances, 39–43 implications of, 6 labor law: adoption of pro-labor laws, see also labor rights, 64–65 bottom-up enforcement, 66–69, 165 consequences for labor costs, 65–66 limited top-down enforcement, 191 relations to industrial upgrading, 87–92 sectoral and regional variations of enforcement, 92 violations, 68 labor movement, 52 labor non-governmental organization: lack of independent organization in China, see also All China Federation of Trade Unions (ACFTU), 51–54 exposure of labor incidents and, 192 foreign-funded, 75 involvement in workers’ protests, 221–224 legal mobilization of, 163 movement oriented, 62, 220–221 state repression of, 71–76, 219–225 labor rights: legal rights consciousness, 67–68 individualization of, 69 labor shortage: of migrant workers, 7–8, 36–39 effect on protest intention, 149 labor turnover: of migrant workers, 7, 34, 39 consequences, 8 effect on politicization, 116 effect on protest intention, 149 turnover intention, 142 labor-intensive sectors: reliance on migrants, see also export-oriented industrialization; migrants; industrial upgrading, 32 declining profits, 8, 81, 85 declining tax contribution, 81, 86 governments’ changing attitudes to, 78–87

increasing labor costs, 8, 81 policies to discourage investment in, 86–93 land scarcity, 83 relations to industrial upgrading, 83, 87, 90 law-abiding firms, see also labor law, legal demands, 161 attribution of blame at, 182–184 concerns for reputation at, 191, 200 expected media exposure at, 195–196 perception of group efficacy at, 193 survey stats, 169–170 law-based protests, see also rightful resistance, 161 for overtime pay, 204 for payment of wage arrears, 171–176 for social insurance, 203–204 media attention to, 206 police presence at, 179–181, 205 reliance on state authorities, 165–166 Lee, Ching Kwan, see also cellular activism, 2, 60 legal demands: payment of wage arrears, see also extra-legal demands, 166, 171–172 about overtime pay, 203 about social insurance, 203 legal disputes, see also law-based protests, 66–69 judicialization of labor disputes, 163 labor dispute resolution system, 163 number of cases, 163, 170–171 Levi, Margaret, 99 Lewis Turning Point, 38 Lide protest, 222–224 line supervisor, see also protest leader, mobilizational resources, 17 capacity to demobilize, 130 experience of upward mobility, 107 internal attribution of blame, 107–109 migration background, 130 mobilizational resources, 130–133, 199–200, 203–204 perception of upward mobility, 111, 140 protest experience of, 157 protest intention, 133, 135, 140–141, 147 relationship with rank-and-file workers, 129–131

Index role in protests, 105, 129–130, 199–200, 203–204 views of the central government, 118–119 Lorentzen, Peter, 12, 57, 212 lying flat movement, 228–229 Mao era: the hukou system, 28–29 Maoist ideology, of student activists, 227 material co-optation, 12, 57–58 protests as a tool for material, 2, 57–58 media: reports of labor protests, see also mobilizational resources, mobilizational structures, protest leaders, 187–190, 200–201 changes in the Xi era, 218–219 commercialization, 190 depiction of migrant workers, 102, 104–105 foreign-invested firms and, 192–196 framing of protests; group efficacy and, 187, 193–195, 200–201 investigative journalism, 188–190 large-scale firms and exposure of, 196 perception of successful mobilization and, 149, 158 role during protest, 192–193, 200–201 workers’ expectation of reports, 194–196, 200–201 migrant workers: rural-to-urban workers, see also hukou system; economic grievances, social grievances, 5, 30–32 as protest participants, 8 as temporary residents, 30, 32–36 changing policies of, 63–66 exploitation of, 32–36 lack of social security, 84 media’s depiction of, 102, 104–105 new generation of, 39–42, 97, 103, 109–110 perceived benefits of, 36, 84–85 perceived social costs of, 84–86 perception of upward mobility, 97–98 public prejudice against, 85, 103 regional distribution of, 31 sectoral distribution of, 31 social and political exclusion of, 41–42, 63, 102 support for the status quo, 97, 102–103 the central government’s attitudes, 104–105 mobilization resources, 17, 133–135

261

discursive, 18, 186, 193–195, 200–201 external, 221–224 importance in overcoming collective action barriers, 127–128 legal knowledge, 67–69 of line supervisors, 129–135 use of anti-foreign slogans as, 208–210 mobilizational structures: definition, see also protest leader, collective action resources, line supervisors, social network, 133 importance of social network, 143 production units as, 133–135 nationalism: effect on labor protests, see also Asian Tigers, 208 display of, during labor protests, 208–210 newly industrialized countries: comparisons to China, 50 non-cooperation, subtle forms of, see also lying flat movement, sanhe youth, 228–229 OBrien, Kevin, see rightful resistance Olson, 127 Opening up (kaifang), 29, 30 particularization, see fire alarm system 60 particularistic bargaining, 11, 12, 16 particularistic framing, 59, 103 Pearl River Delta (PRD) region, 20, 31 attraction of migrants to, 31, 85 collective bargaining in, 221 declining industries, 81 increasing bankruptcy, 66 shortage of migrant labor in, 37 performance legitimacy, 95 personal efficacy: definition, see also group efficacy, 146 effect of protest experience on, 158 effect on protest intention, 146–147 measurement of perception, 147 political grievances, of migrants: debate about, 61 political grievances, of migrants: debate about, police see repression, 59 politicization: of economic grievances, see also depoliticization, generalization, structural attribution, 13, 59–96, 98–99 of social grievances, 15–16

262

Index

pro-democracy movement: involvement of labor, 2, 8–9 protest leader importance for unorganized groups, 128 in migrant protests, 199–200 in SOE protests, 128 protest tactics: disruptive and non-disruptive, 177, 205 local government response and, 177, 180 police response to, 181, 205 publicity, 192–193, 200–201 strike map data of, 177 protest: toleration of, in authoritarian intention of, 137–154 protest: toleration of, in authoritarian regimes, see also information, atomized protest, interest-based protest, law-based protest, 1–2 apolitical, 2, 4, 18, 57, 60, 183 as a source of information, 2, 57–58 as a tool for material co-optation, 2, 57–58 atomized, see atomized protest; defensive and offensive, 18, 129 blame attribution and intention of, 120–122 changing nature of, 164 debates about origins, 143–144 defensive and offensive, 18, 129 direct experience of, 120–121, 136, 156–159 indirect experience of, 120, 201 intention of, 135–136 media reports of, 187–190, 200–202 reasons for non-participation, 155 state authorities’ handling of, 165–167, 180–181 success rates of, 158, 171 survey stats of, 170 toleration of, in China, 1–3 Pun, Ngai, 7, 209 repression: differences from coercion, see also labor coercion, 25–26 during protests, 179–181 effect on protest intention, 149–150 in Tiananmen Square, 9–10, 51 limits of, 229 police, 146, 179–181, 205, 222 the 2015 crackdown, 219–225 the 2018 crackdown, 225–227 under the Xi leadership, 216

within firm, 146, 151, 200 workers’ expectation of, 146, 200 rightful resistance, 18, 164 safety valve, 2, 12, 17 sanhe youth, 228 social grievances, 42–43, 97–98 definition, 111 determinants of, 112–116 effect of job hopping on, 114–116 measurement, 111 passive resistance and, 228–229 politicization of, 116–119 social media: public exposure of protests and, 172 social movement: differences from protest, see also protest, 145 social network: in factories, see also mobilizational resources, 133–135 firm tenure and, 203–204 labor turnover and, 143 South Korea: economic growth in the military regimes, 26–28 labor activism during the military rule, 104 labor control, 50, 53, 61 labor disputes at Korean-owned firms, 168 student activists, 226 wage increase, 45 state-labor relations in authoritarian regimes, see labor coercion, atomization, corporatism, incorporation, co-optation 100 state-owned enterprises (SOEs): protest during SOE reform, 71–72 government’s framing of, 101 strategic agency of workers: debate in literature, 143–144 during anti-foreign protests, 208–210 in cases, 132, 200–201, 206 of line supervisors, 154, 199–201 use of media, 200–201 structural bargaining power, see also labor shortage, labor turnover, job availability, 8, 36–39, 124, 146 definition, 38 perceived job availability, 142 relationship with protest, 149 the hukou system and, 35

Index surplus labor: of migrants, 30, 37–38 absorption of, 38 survey data: using survey to study protest behavior, 136–137 system attribution, see also politicization; blame, 15 formation of causal beliefs and, 99 workers’ difficulty of, 98 Tilly, Charles, 127, 139 trade unions, see All-China Federation of Trade Unions upward mobility: lack of, within firm, see also social grievances, 17, 41, 140 determinants of, 112 effect on protest intention, 147–149 lack of, within society, 14, 41–42 of different workers, 111 politicization of, 118 workers’ perception of, 111, 113–114 urban bias: the hukou system and, 28–29 vote with feet, see also Hirshman, exit, 8 wage arrears, 166

263

media reports of, 188 protests for, 171, 172 wages: low wages of migrants, 35 increase of migrant wages, 8, 38 Wen Jiabao, 103, 188 Wintrobe, 26, 33 workplace disputes: about compensation on firm tenure, see also protest, economic grievances, 151, 175, 203 about overtime pay, 203 about social insurance, 150, 203 about wage arrears, 166, 171 about year-end bonuses, 150, 199 Xi era: increased repression, 216, 218–219 crackdown on labor NGOs and activists, 219–225 decreasing discursive resources, 218–219 decreasing interest-based protests, 217–218 Yangtze River Delta, 20, 31 Yue Yuen, 115, 140, 143 anti-foreign slogans in, 209 involvement of labor activists in, 222 media reports of, 188