The Disempowered Development of Tibet in China: A Study in the Economics of Marginalization 9780739134375, 9780739134382, 9780739134399, 2013032370

Series: Studies in Modern Tibetan Culture, Lexington Books Weatherhead East Asian Institute, Columbia University Since

180 76 4MB

English Pages [467] Year 2013

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Table of Contents
List of Figures, Maps, and Tables
List of Abbreviations
Note on Tibetan Spelling
Glossary (a Note on Terminology)
Acknowledgments
Preface: Reflections on Researching Development in Tibet
Chapter 1:
Introduction: The Disempowered Development of Tibet in China
Chapter 2:
The Inception of Modern Development in Tibet under PRC Rule
Chapter 3:
Population Foundations of Marginalization in Tibet
Chapter 4: Instituting Economic Growth and Marginalization in Tibet
Chapter 5: The Great Transformation of Tibet? Rapid Labor Transitions, Polarization, and the Emerging Fault Lines of Stratification in Urban Tibet
Chapter 6: 6The Education-Employment Nexus of Urban Exclusion in Tibet
Chaptre 7: Subsistence Capacity and the Material Foundations of Resistance
Chapter 8: Boycotts and Religious Networks: Counter-Strategies of Integration into the Heart of the New China
Chapter 9: Conclusion: From Polarization to Protest in Contemporary Tibet
Bibliography
Index
About the Author
Recommend Papers

The Disempowered Development of Tibet in China: A Study in the Economics of Marginalization
 9780739134375, 9780739134382, 9780739134399, 2013032370

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

The Disempowered Development of Tibet in China

Studies in Modern Tibetan Culture Series Editor Gray Tuttle, Columbia University Advisory Board Lauran Hartley, Columbia University (literature) Isabelle Henrion-Dourcy, Université Laval (anthropology) Kurtis Schaeffer, University of Virginia (religion) Emily Yeh, University of Colorado at Boulder (human geography) The Studies in Modern Tibetan Culture series focuses on Tibetan culture and society from the early modern period of the seventeenth century to the present. The first series on modern Tibetan studies by a scholarly press, it explores how modernity manifests in a wide range of fields, not only religion, but also literature, history, economy, anthropology, media, and politics. It seeks to bring rarely heard and important Tibetan perspectives to a wider audience by publishing fresh analyses of yet unexplored source materials ranging from census and yearbook databases to auto/biographies and ethnographic fieldwork, as well as original translations of poetry, biography, and history. Titles in Series The Nine-Eyed Agate: Poems and Stories by Jangbu and translated by Heather Stoddard Labrang Monastery: A Tibetan Buddhist Community on the Inner Asian Borderlands, 1709–1958 by Paul Kocot Nietupski The Hidden Life of the Sixth Dalai Lama by Ngawang Lhundrup Dargyé and translated by Simon Wickham-Smith The Disempowered Development of Tibet in China: A Study in the Economics of Marginalization by Andrew Martin Fischer

Studies of the Weatherhead East Asian Institute Columbia University Selected Titles (Complete list at: http://www.columbia.edu/cu/weai/weatherhead-studies.html) Beyond the Metropolis: Second Cities and Modern Life in Interwar Japan, by Louise Young. University of California Press, 2013. Imperial Eclipse: Japan’s Strategic Thinking about Continental Asia before August 1945, by Yukiko Koshiro. Cornell University Press, 2013. The Nature of the Beasts: Empire and Exhibition at the Tokyo Imperial Zoo, by Ian J. Miller. University of California Press, 2013. Reconstructing Bodies: Biomedicine, Health, and Nation-Building in South Korea since 1945, by John P. DiMoia. Stanford University Press, 2013. Tyranny of the Weak: North Korea and the World, 1950–1992, by Charles Armstrong. Cornell University Press, 2013 The Art of Censorship in Postwar Japan, by Kirsten Cather. University of Hawai‘i Press, 2012. Asia for the Asians: China in the Lives of Five Meiji Japanese, by Paula Harrell. MerwinAsia, 2012. Lin Shu, Inc.: Translation and the Making of Modern Chinese Culture, by Michael Gibbs Hill. Oxford University Press, 2012. Occupying Power: Sex Workers and Servicemen in Postwar Japan, by Sarah Kovner. Stanford University Press, 2012. Redacted: The Archives of Censorship in Postwar Japan, by Jonathan E. Abel. University of California Press, 2012. Empire of Dogs: Canines, Japan, and the Making of the Modern Imperial World, by Aaron Herald Skabelund. Cornell University Press, 2011. Planning for Empire: Reform Bureaucrats and the Japanese Wartime State, by Janis Mimura. Cornell University Press, 2011. Realms of Literacy: Early Japan and the History of Writing, by David Lurie. Harvard University Asia Center, 2011. Russo-Japanese Relations, 1905–17: From Enemies to Allies, by Peter Berton. Routledge, 2011. Behind the Gate: Inventing Students in Beijing, by Fabio Lanza. Columbia University Press, 2010. Imperial Japan at Its Zenith: The Wartime Celebration of the Empire’s 2,600th Anniversary, by Kenneth J. Ruoff. Cornell University Press, 2010. Passage to Manhood: Youth Migration, Heroin, and AIDS in Southwest China, by Shaohua Liu. Stanford University Press, 2010. Postwar History Education in Japan and the Germanys: Guilty Lessons, by Julian Dierkes. Routledge, 2010. The Aesthetics of Japanese Fascism, by Alan Tansman. University of California Press, 2009. The Growth Idea: Purpose and Prosperity in Postwar Japan, by Scott O’Bryan. University of Hawai‘i Press, 2009. Leprosy in China: A History, by Angela Ki Che Leung. Columbia University Press, 2008. National History and the World of Nations: Capital, State, and the Rhetoric of History in Japan, France, and the United States, by Christopher Hill. Duke University Press, 2008.

The Disempowered Development of Tibet in China A Study in the Economics of Marginalization Andrew Martin Fischer Studies of the Weatherhead East Asian Institute, Columbia University The Studies of the Weatherhead East Asian Institute of Columbia University were inaugurated in 1962 to bring to a wider public the results of significant new research on modern and contemporary East Asia.

LEXINGTON BOOKS

Lanham • Boulder • New York • Toronto • Plymouth, UK

Published by Lexington Books A wholly owned subsidiary of Rowman & Littlefield 4501 Forbes Boulevard, Suite 200, Lanham, Maryland 20706 www.rowman.com 10 Thornbury Road, Plymouth PL6 7PP, United Kingdom Copyright © 2014 by Lexington Books All rights reserved. No part of this book may be reproduced in any form or by any electronic or mechanical means, including information storage and retrieval systems, without written permission from the publisher, except by a reviewer who may quote passages in a review. British Library Cataloguing in Publication Information Available Library of Congress Cataloging-in-Publication Data Fischer, Andrew Martin, 1967The disempowered development of Tibet in China : a study in the economics of marginalization / Andrew Martin Fischer. pages cm. -- (Studies of the weatherhead East Asian institute) (Studies in modern Tibetan culture) Includes bibliographical references and index. ISBN 978-0-7391-3437-5 (cloth : alk. paper) -- ISBN 978-0-7391-3438-2 (pbk. : alk. paper) -- ISBN 978-0-7391-3439-9 (electronic) 1. Tibet Autonomous Region (China)--Economic conditions. 2. Tibet Autonomous Region (China)--Social conditions. 3. Economic development--China--Tibet Autonomous Region. 4. Marginality, Social--China--Tibet Autonomous Region. I. Title. HC428.T48F567 2013 338.951’5--dc23 2013032370

™ The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI/NISO Z39.48-1992.

`

Printed in the United States of America

To my teachers, both spiritual and secular May I humbly transmit their lineages To my mother, Phyllis Mary, my muse and Amiya and Éva, our delight and devotion

Table of Contents

List of Figures, Maps, and Tables

xi

List of Abbreviations

xv

Note on Tibetan Spelling

xvii

Glossary (a Note on Terminology)

xix

Acknowledgments: Reflections on Researching Development in Tibet xxv Preface xxix  1  Introduction: The Disempowered Development of Tibet in China

1

 2  The Inception of Modern Development in Tibet under PRC Rule

47

 3  Population Foundations of Marginalization in Tibet

83

  4  Instituting Economic Growth and Marginalization in Tibet

127

 5   The Great Transformation of Tibet? Rapid Labor Transitions, Polarization, and the Emerging Fault Lines of Stratification in Urban Tibet

191

 6  The Education-Employment Nexus of Exclusion in Tibet

247

  7  Subsistence Capacity and the Material Foundations of Resistance

291

 8  Boycotts and Religious Networks: Counter-Strategies of Integration into the Heart of the New China

335

ix

x

Table of Contents

 9  Conclusion: From Polarization to Protest in Contemporary Tibet

371

Bibliography 393 Index 411 About the Author

425

List of Figures, Maps, and Tables

Figures Figure 2.1: Real per capita GDP, selected provinces, 1991-2000 70 Figure 2.2: Real per capita GDP of the TAR, 1978-1998 72 Figure 3.1: Tibetan population of China, all provinces, age and sex, 2000 census 89 Figure 3.2: Birth, death, and natural increase rates, China and TAR, 1970-2010 91 Figure 4.1: Real per capita GDP growth rates, selected provinces, 1995-2010 134 Figure 4.2: Real per capita GDP, selected provinces, 1995-2010 135 Figure 4.3: Provincial to national real per capita GDP, 1995-2010 137 Figure 4.4: Structure of the TAR GDP, 1995-2010 139 Figure 4.5: Primary sector share of GDP, selected provinces, 1995-2010 141 Figure 4.6a: Industry and mining GDP share, selected provinces, 1995-2010 142 Figure 4.6b: Construction GDP share, selected provinces, 1995-2010 142 Figure 4.7: Tertiary sector share of GDP, selected provinces, 1995-2010 143 Figure 4.8: Local government deficit as a proportion of local government expenditure, selected provinces and provincial total, 1990-2010 152 Figure 4.9: Local government deficit as a proportion of provincial GDP, selected provinces and provincial total, 1990-2010 153 xi

xii

List of Figures, Maps, and Tables

Figure 4.10: Local government deficit per capita, constant 2010 rmb, selected provinces and provincial total, 1990-2010 156 Figure 4.11: Investment as ratio of GDP, selected provinces, 1990-2010 167 Figure 4.12: Gross capital formation as ratio of GDP, selected provinces 167 Figure 4.13: Trade balance (net exports) as percentage of provincial GDP, 1993-2010 174 Figure 5.1: Primary sector employment share, 1990-2010 196 Figure 5.2: Share of the total employment classified as rural, 1993-2010 199 Figure 5.3: Percentage difference between rural and primary shares, 1993-2010 202 Figure 5.4: TVE, rural private enterprise and self-employed individual employment as a share of total employment, 1993-2010 203 Figure 5.5: Secondary sector employment share, 1990-2010 204 Figure 5.6: Tertiary sector employment share, 1990-2010 206 Figure 5.7: Real per capita rural household net income, 1990-2010 216 Figure 5.8: Urban-rural household income inequality, 1990-2010 219 Figure 5.9: Real per capita urban household disposable income, 1990-2010 220 Figure 5.10: Urban wage inequality, 1998-2010 230 Figure 5.11: Average wages of staff and workers, 1998-2010 231 Figure 6.1: Illiteracy rates, two-period moving averages, 1990-2010 255 Figure 7.1: Original value of productive fixed assets per rural household, 2010 300

Maps Map 1:

Tibetan cultural areas and western provinces of the PRC

xxiv

Tables Table 3.1: Distribution of population by selected nationality, 2000 census Table 3.2: Selected province-level population statistics, c. 2000 Table 3.3: Changes in Han and minority populations, 1990 and 2000 censuses Table 3.4: Net population inflow or outflow from 2000 to 2010

97 103 106 110



List of Figures, Maps, and Tables xiii

Table 4.1: Expenditure, costs, and infrastructure inputs in education, selected provinces, 2000 and 2010 159 Table 4.2: Ratio of change in nominal GDP over change in deficit/subsidies, various provinces and provincial totals 163 Table 4.3: Composition of investment in fixed assets, 2010 169 Table 5.1: Relative and nominal GDP/labor ratios, 2000 and 2010 210 Table 5.2: Ratios of top and bottom urban income shares, 2000 and 2010 223 Table 5.3: Composition of urban employment, 2010 226 Table 6.1: Illiteracy rates among population aged 15 and older, 2009 survey 257 Table 6.2: Illiteracy rates among population aged 15 and older, by sex, 2009 survey 259 Table 6.3: Schooling levels among population aged 6 and older, 2000, 2009, and 2010 261 Table 6.4: Schooling levels among population aged 6 and older, rural and city, 2009 265 Table 7.1: Agricultural productivity in selected provinces and national average, 2010 297 Table 7.2: Major productive fixed assets per 100 rural households, 2010 300

List of Abbreviations

CPC

Communist Party of China

CPI

consumer price index

CSY

China Statistical Yearbook

FDI

foreign direct investment

GDP gross domestic product ILO

International Labour Organisation

INGO international non-governmental organization OECD Organisation for Economic Co-operation and Development OWC Open the West Campaign (Ch. xibu da kaifa) PRC

People’s Republic of China

Rmb

renminbi = yuan

TAC

Tibetan Autonomous County (see appendix 1)

TAP

TibetanAutonomousPrefecture (see appendix 1)

TAR

Tibet Autonomous Region (see appendix 1)

TVE

township and village enterprise

UNDP United Nations Development Programme

xv

Note on Tibetan Spelling

As this is a work of socio-economic analysis, not a work of philology, linguistics, religious studies, or even anthropology, the Tibetan spelling adopted in this book is based on a simplified approach of using phonetic spelling according to the spoken pronunciation in the Lhasa dialect, following the standard conventions of phoneticization as developed by Jeffrey Hopkins. Wiley standards of Tibetan transliteration are not used. In order to avoid confusion with regard to place names, Chinese names are also provided where appropriate.

xvii

Glossary (a Note on Terminology)

In this book, “China” refers to the People’s Republic of China (PRC) and “Tibet” refers to all of the Tibetan areas in China, including the Tibet Autonomous Region (TAR) and the Tibetan areas that are incorporated into the provinces of Qinghai, Gansu, Sichuan, and Yunnan. Often known as “greater Tibet,” “cultural Tibet,” or “ethnographic Tibet,” this region is about the size of western Europe or about one quarter of China. This larger understanding of Tibet is actually not controversial given that it conforms to administrative definitions in China, which identify Tibetan (Ch. Zang) autonomous areas at various levels of jurisdiction. The TAR is equivalent to a province, whereas Tibetan areas incorporated into the other provinces are designated as either autonomous prefectures (TAPs) or counties (TACs), and even as Tibetan Autonomous Townships in some cases. With the exception of some disagreements on the borderlands of eastern Tibet, the Chinese administrative definitions are almost identical to the definition of Tibet used by the Tibetan Exile Government and they conform with indigenous self-definition, insofar as they include most of the regions that have been traditionally inhabited by people who have come to be called (or have come to call themselves) Tibetan and their territory Tibet. However, administrative divisions largely ignore indigenously defined subregions. The indigenously defined Tibetan regions of Utsang (central Tibet), Jangtang, and Ngari (Western Tibet) are all located in TAR. Kham (eastern Tibet) is divided between the TAR, Qinghai, Sichuan, and Yunnan. Amdo (north eastern Tibet) is divided between Qinghai, Gansu and Sichuan (see map 1). Most of these highland areas, at average altitudes of well over 3,000 meters, are also clearly differentiated from non-Tibetan lowlands by xix

xx

Glossary (a Note on Terminology)

topography and population density. While the TAR is the administrative area that the PRC government and most of the western media usually imply when they refer to “Tibet,” it only accounts for just over half of the Tibetan autonomous areas in China and less than half of the total Tibetan population in China. The boundaries of this region were determined by the territory controlled by Lhasa at the time of the PRC invasion in 1950. Amdo The indigenously defined region of northeast Tibet, currently subdivided between Qinghai, Gansu, and Sichuan. Chinese and/or Han This book uses the terms “Chinese” and “Han” more or less synonymously. In the statistical sections, the official terminology of Han is used, whereas in the qualitative sections, the term “Chinese” is used to designate Han Chinese “nationality.” This usage is officially contested in China given that “Chinese” is supposed to designate nation-state nationality rather than ethnicity. In other words, Tibetans, Uyghurs, Mongolians and other distinct minorities of China, along with the Han, are all officially “Chinese,” that is, citizens of China, similar to the usage of “American” in the United States. However, this official usage is often contested in the minority areas for obvious reasons. Tibetans also largely ignore this official differentiation, at least outside officially sanctioned discourse, and they use the terms “Han” and “Chinese” more or less synonymously (usually referred to as gyami in Tibetan). It is for this reason that the indigenous usage is used in the qualitative sections. Interestingly, the two terms are also implicitly synonymous in Chinese, given that one can refer to Chinese language as zhongwen (zhong as in Middle Kingdom, i.e., China), or as hanyu (language of the Han), both of which refer to the same language. The Han are thus the only nationality in China to enjoy the privilege of synonymy with the nation. Hence, rather than playing around with nationalist semantics, I prefer to call a spade a spade. Chinese Muslims (i.e., Hui, Salar, etc.) The term “Chinese Muslims” is often used to differentiate Muslims who are racially close to the Han Chinese from the Turkic Uyghur Muslims of Xinjiang. Unlike the Uyghurs, most of these Muslims are Chinese speaking and have been settled in China for most of the second millennium. The dominant group among the Chinese Muslims is the Hui, who are spread throughout the country and are actually quite diverse. Nonetheless, an



Glossary (a Note on Terminology) xxi

important concentration of Hui is based in the region that surrounds Amdo, split into three northwest provinces—the Ningxia Hui Autonomous Region, Gansu, and the northeast corner of Qinghai. The Hui population in this northwest region is supplemented by several other small Chinese Muslim groups, such as the Dongxiang in Gansu and the Salar in Qinghai. This book will refer to Chinese Muslims as Muslims and the Uyghur Muslims as Uyghurs. On Muslims in Tibet and China, see Gladney (1991), Cabezon (1997), Lipman (1997), and Dillon (1999). Kham The indigenously defined region of eastern Tibet, currently subdivided between the TAR, Sichuan, Yunnan, and Qinghai. Lama The term lama is used in this book according to its actual meaning, as understood by religiously educated Tibetans, and not by its typical usage in many official Chinese publications. The word literally means “one who possesses sublime qualities” and it is a translation of the Sanskrit word guru. In practice, it is used to designate one’s spiritual teacher if used with personal reference. If used with a more social or generic reference, it designates religious teachers who have been recognized for their exceptional teaching, scholarship, or meditation, although in colloquial usage it is used for any religious teacher, even from other religions (e.g., kache lama would be a Muslim imam). Unfortunately, most official Chinese publications (and many Western journalists) use the term lama to refer to monks and also designate Tibetan Buddhist monasteries as “lamaseries.” Tibetan Buddhism is also sometimes referred to as “Lamaism,” although this would be equivalent to calling Catholicism “Priestism” or Islam “Imamism.” There is some evidence to suggest that this use of English terminology by the PRC was possibly inherited from earlier English colonial scholarship. Similarly, rimpoche, which means “the precious one,” is an honorific title given to a person who has been recognized as the reincarnation of a previous lama. A trulku (see below) is the current incarnation of a lineage of recognized reincarnated lamas, and these lineages are institutionalized within the political and religious hierarchy. The word trulku translates as “emanation body,” implying that the lama is an emanation of a Buddha (i.e., an enlightened mind), appearing in ordinary human form (i.e., a human with faults). This understanding is derived from a disciple’s personal practice of guru devotion. It is not necessarily implied that the trulku is definitely, for all people, a Buddha, although certain high lamas

xxii

Glossary (a Note on Terminology)

such as the Dalai Lama, the Panchen Lama, or the Karmapa are definitely considered by religious Tibetans to be Buddhas. The Chinese usually translate trulku as huofo, which translates into English as “Living Buddha.” Thus, Tibetan trulkus are usually called “Living Buddhas” in official English publications of the PRC and even by many Western journalists. Panchen Lama The Panchen Lama lineage, with its seat in Tashi Lhunpoe in Shigatse, TAR, is considered to be the second-highest authority within the politicoreligious hierarchy of central Tibet up to the Chinese invasion in 1950, although it is considered to be higher than the Dalai Lama lineage in terms of pure religious or hagiographic hierarchy. The Tenth Panchen Lama was a contemporary of the current Fourteenth Dalai Lama, and both were born in the region of Amdo, near Xining (Tib. Siling) City, the capital of Qinghai Province. Unlike the Fourteenth Dalai Lama, who escaped into exile in 1959, the Tenth Panchen Lama remained in Tibet and China in an attempt to collaborate with the Chinese. He was imprisoned for almost ten years and then spent another five under house arrest in Beijing, after which he was rehabilitated in 1982 (see Barnett 1998). He passed away in 1989. The selection of the current Eleventh Panchen Lama in 1995 was the source of intense dispute between the PRC and the Tibetan Government in Exile that continues to the present (see Hilton 1999). TAC Tibetan Autonomous County, the lower level of autonomous status accorded to Tibetan (and other minority nationality) areas outside the TAR, incorporated into prefectures, or sometimes self-standing. Other minority autonomous areas are similarly designated, such as HAC for Hui Autonomous County, TMAP for Tibetan-Mongolian Autonomous Prefecture, and so forth. TAP Tibetan Autonomous Prefecture, the higher level of autonomous status given to Tibetan (and other minority nationality) areas outside the TAR and incorporated into the other four provinces containing Tibetan areas, namely Qinghai, Gansu, Sichuan, and Yunnan. TAR Tibet Autonomous Region, the provincial-level jurisdiction that the Chinese often refer to as “Tibet” but that encompasses just over half of



Glossary (a Note on Terminology) xxiii

the traditional Tibetan areas. The boundaries of this region were more or less determined by the territory controlled by Lhasa in the first half of the twentieth century. Some scholars (such as Melvyn Goldstein) to refer to it as “political Tibet,” meaning that it was the only region of Tibet prior to Chinese invasion in 1950 that had a centralized state governing the entire region, thereby substantiating claims of independent statehood as effectively practiced prior to the Chinese invasion. Tibetan This book refers to Tibetans as those people who call themselves Bodpa (Tibetans) or some recognized sub-group of Tibetans (i.e., Amdowa, Khampa), speak various dialects derived from the written language referred to as bod-yig (Tibetan), and/or refer to their indigenous territory as Bod (Tibet) or some recognized sub-division thereof (i.e., Amdo, Kham, etc.). Trulku Trulku lineages originate from the recognition of a past lama as having attained a state of high realization through outstanding scholarship and meditative achievement. This state of realization is then believed to be passed on from life to life because the lama in question is able to control the process of death and rebirth through meditation. The trulku institution is generally considered to have emerged in the thirteenth century as an innovation by the Sakya order of Tibetan Buddhism. Utsang The indigenously defined region of central Tibet; U is the area around Lhasa and Lhoka, the main seat of the Dalai Lamas, whereas tsang is the area around Shigatse, the main seat of the Panchen Lamas.

The five shades of gray distinguish the indigenously designated Tibetan regions of Amdo, Kham, Utsang, Jangtang, and Ngari. Source: The Tibetan and Himalayan Library, www.thlib.org.

Map 1: Tibetan cultural areas and western provinces of the PRC

Acknowledgments

I am always uncomfortable writing an acknowledgment for work on the politically sensitive region of Tibet, particularly on matters that are politically sensitive, because those to whom I am most indebted invariably cannot be mentioned. Doing so would risk complicating their lives, especially since the widespread protests of 2008 and, again, since the recent protests centerd on the wave of self-immolations that started in 2011 and continue at the time of writing. Suffice it to say that if and when they would ever read this text, they would know who they are. I extend my deepest gratitude to them, in particular the numerous Tibetans, Chinese, Chinese Muslims, Mongolians, and Mongours whom I met in the course of my fieldwork, including scholars, development workers, officials, monks and nuns, lamas, teachers, students, farmers, pastoralists, businesspeople, workers, and unemployed. Special mention goes to the numerous refugees, from all regions of Tibet, with whom I had intimate contact during my time living in India and Nepal from 1995 to 2001. They provided me with many initial insights into the state of marginalization faced by Tibetans in the Tibetan regions of China. They also fueled my initial inspiration to return to the world of academic research, after escaping it in youthful disillusion for about ten years. Among those who can be named, many deserve to be mentioned. These include, in alphabetical order: Andre Alexander, Anders Anderson, Bob Ankerson, Robbie Barnett, Ken Bauer, Beimatso, Andrew Bunbury, Geoff Childs, Sue Costello and her family, Sienna Craig, Thierry Dodin, Philippe Dufourg, David Germano, Mark Giffard-Lindsay, Ethan Goldings, Melvyn Goldstein, Nancy Harris, Arthur Holcombe, Charlotte Hyvernaud, Iida, xxv

xxvi

Acknowledgments

Erika Jacobson, Tanzen Lhundup, Li Tao, Gelek Lobsang, Ma Cheng Jun, Anna Morcom, Françoise Robin, Ma Rong, Damien Morgan, Padmatso, Fernanda Pirie, Losang Rabgey, Tashi Rabgey, Carol Sandup, Kate Saunders, Kevin Stuart and his team of teachers and students in Xining, Lobsang Tashi, Gray Tuttle, Daniel Winkler, Yang Wenkui and family, Yangzhongtar, Emily Yeh, Yi Lin, and Yi Qing. I would also like to thank Adrian Zenz, whom I discovered at the last minute of finalizing this book when I was asked to review his book by editors at Brill. Finally, Mary Zsamboky is beyond thanking; words do not suffice for the immense inspiration and help that she brought to the waning moments of this book project. Special thanks go to Gray Tuttle, the editor of this series published by Lexington. He approached me to publish in the series in 2007, before the events of 2008, and he has since been incredibly patient and encouraging with my perpetual delays in finishing the book. I could not imagine a more supportive series editor and I hope that I have done justice to his trust and confidence. Great thanks are also due to the anonymous reviewers of the book, as well as to Emily Yeh for her invaluable input, comments and suggestions, despite living and working in one of the few nations left on earth than does not have any meaningful maternity leave. The editors at Rowman & Littlefield have also shown much patience since first negotiating the book contract with me in 2007, particularly given that the preparation of this book has passed through at least three of them, to the best of my knowledge, each of whom has been very helpful: Patrick Dillon, Justin Race, and Sabah Ghulamali. The most patience was shown by Justin Race and the most work has been done by Sabah Ghulamali, who took over the book just as we were starting to enter into the production phase. Thanks also to Laura Reiter, the production editor of the book, whose helpfulnes and effectiveness have impressed me. Much of the research of this book was conducted during my PhD studies at the London School of Economics from 2002 to 2007, and several people from this period deserve special thanks. Athar Hussain and Stephen Feuchtwang both served as my supervisors in the first year of my PhD, Athar formally and Stephen informally. James Putzel took over from Tim Allen as my supervisor at the Development Studies Institute at LSE in my second year and was an exemplary guide and has been an incredible source of support and friendship ever since. Tim Dyson, Jo Beall, and John Harriss also provided enthusiastic support throughout the course of my years at LSE. Finally, I would like to thank my two PhD thesis examiners, Peter Townsend and Sarah Cook, whose meticulous examination of my thesis improved it immensely. I continue to be guided by the inspiration of several of my early academic teachers from McGill University in Montreal, namely, the late Tom Asimakopoulos, Tom Kompas, Tom Naylor, Kari Polanyi-Levitt, Charles



Acknowledgments xxvii

Taylor, and James Tully. I also pay my respect and devotion to my spiritual teachers, in particular the trinity of my root gurus: His Holiness the Dalai Lama, Thubten Zopa Rimpoche, and the late Khalkha Jetsun Dhampa Rimpoche, whom I had the great fortune of serving for three years. Several generous contributions made the early phases of this research possible. These deserve mentioning, even though I have had no financial support since 2006 (besides my own research funds provided by my current institute, the Institute of Social Studies in the Hague, the Netherlands, part of Erasmus University Rotterdam since 2009). These include: The Québec Government (Fonds québecois de la recherche sur la société et la culture), the UK Government (Overseas Research Student Award), the London School of Economics, and the Canadian Section of Amnesty International, all of whom provided general PhD funding. In addition, the Crisis States Research Centre at the LSE and the Central Research Fund of the University of London provided field research funding in 2003 and 2004. More recently, my staff group in my current place of work—the Institute of Social Studies—contributed some of the typesetting costs during the final stage of producing this book. However, this journey would never have left the ground had it not been for the vital assistance of my mother, Phyllis Fischer, at the start of my degree, when I had been left financially stranded by all other funding sources. Without her support, I would probably be working in an INGO right now rather than pontificating in academia. While I am still not sure whether the latter will end out being more beneficial to society, her faith in me never ceases to amaze me, just as her commitment to social justice has never ceased to awe me. I hope to repay her by living in her example. Many people have contributed to this book. Whatever merit is contained within is surely in large part the result of their collective wisdom. The opinions expressed are my own, and any erroneous meanderings are due entirely to my own ignorance or arrogance, having not listened enough to those who know better.

Preface: Reflections on Researching Development in Tibet

Writing about the development of Tibet in China is fraught with controversy. The most obvious reason is that development serves as a primary discourse for the Communist Party of China (CPC) to legitimize its incorporation and integration of Tibet into China since the 1950s. Hence, it is also a conflicted subject for Tibetan exiles, to the extent that some reject the possibility that whatever has occurred in Tibet over the last sixty years can or should be considered as development given that it has occurred under occupation. Development in Tibet is similarly a controversial subject from a wider geopolitical perspective given that China has traditionally been considered a leader of Third Worldist struggles for self-determination. It continues to inspire many—including some of my own mentors of development economics—as a bulwark and possible alternative to the increasingly predatory behavior of US power in the global economy. From this perspective, the topic of Tibet is often perceived as an annoying rhetorical tactic used by the United States and other Western powers to undermine the legitimacy of China’s developmentalist project. These sensitive and politicized associations that the subject of Tibet conjures often create an awkward space within which to conduct research. I found myself in this position following an experience of living with Tibetan refugees in India for about seven years and acquiring a decent degree of fluency in Tibetan (which has since deteriorated due to the tolls of contemporary academic work). There is little doubt in my mind—and in the minds of most observers of the Tibetan situation, however their awareness might be subsequently rationalized—that the incorporation of Tibet into China since the 1950s has been an experience of profound xxix

xxx

Preface: Reflections on Researching Development in Tibet

disempowerment in a broad political sense. It has involved periods of oppression and has resulted in a situation that might be called “effective occupation.” Whether or not we decide to call Tibet an “occupied region,” it effectively functions as such due to its disempowerment at the level of regional and national politics. By “disempowerment,” I do not mean that Tibetans have not been able to exercise any agency. They definitely have. Some of the most striking forms of agency have been exhibited in the widespread protests that spread across the entire region in 2008, or in the ongoing wave of self-immolations since 2011. Other subaltern manners of agency include flight into exile, or even return migration from exile back into Tibet. Agency is also expressed in more collaborative ways, such as by joining the Communist Party (the option of many elite Tibetans), or the People’s Liberation Army (the option of many poor Tibetans), or else in using positions of local influence as government employees or teachers to produce much benefit to local populations, such as with the impressive development of Tibetan-medium education in Amdo/Qinghai since the beginning of the reform period and even earlier under Maoism. Instead, by disempowerment, I mean that the process of integrating Tibet into China and, in particular, the development policies that have driven rapid economic growth and structural change over the past twenty years have, for the most part, been determined by non-Tibetans and by the ebbs and flows of political dynamics occurring outside the Tibetan areas in ways that have borne little or no relation to actual conditions within Tibet. Past and present policies determining the fate of the region and the Tibetan people within it have, for the most part, been deliberated by outsiders who speak (or spoke) little or no Tibetan and who perceive (or perceived) Tibetans in derogatory ways that are comparable to the ways that European colonialists perceived the various non-European peoples that they colonized. Even though Tibetans have been able to exercise various degrees of influence within Tibetan society under these constraints, including privileged Tibetan elites serving the interests of the PRC, they have had little or no control over the main policy levers driving economic development, the development of urban areas and of infrastructure, or the broad shifts in labor regimes and land tenure systems from the mid to late 1950s onward (the timing depends on which part of Tibet one considers). Many aspects of this disempowerment have been oppressive, particularly in the political realm and also, during certain periods such as the Maoist one, in the realms of culture, religion, and society. While such oppression might have since morphed into more ordinary forms and practices of authoritarianism, from a broadly held Tibetan worldview the sources of power have remained an unbroken and alien continuum. This perspective serves as the ontological starting point of the present book.



Preface: Reflections on Researching Development in Tibet xxxi

There is no doubt that development has taken place in this context. By “development,” I do not refer to projects, interventions, and/or practices, as is commonly understood in much of the post-development literature, or to pseudo-experimental impact evaluations of poverty reduction as has become popular in mainstream development economics. Rather, I refer to a classical understanding of development as processes of long-term and (thus far) irreversible social and economic structural transformations closely associated with (sociological) modernity and capitalism. This includes demographic transitions driven by sustained declines in mortality and leading to population increase and urbanization, or else the emergence of “industrial civilization” in fundamentally shaping patterns of power and wealth in local and global economies. Similarly, capitalism is not understood here as the free market economy, as with recent neoliberal cooptations of the term, but as the relatively recent emergence of economic systems that are perpetually driven to increase labor productivity through economic rather than non-economic regimes of compulsion and discipline. The use of these terms does not imply moral judgment or teleology; these transformations are not all necessarily good, and it is definite that significant swaths of populations are negatively impacted by processes associated with capitalist development. Nor does it imply that each society inevitably passes through these transformations in a roughly similar manner, or that people are necessarily happier or more enlightened as a result. In this sense, the notion of modernity needs to be distinguished from the much more simplistic ideas associated with modernization theory. Nonetheless, we are living in a global social system today that has certain fundamental differences with all systems of the past. Most of our children are now likely to outlive us, our consumption is more and more likely to have passed through various industrial processes in one way or another, and our movements and communications are more and more likely to take place through industrial technologies powered by inorganic sources of energy. A meaningful study of development requires locating our understanding of changing societies within these underlying structural transformations. Transposing this understanding of development onto the context of Tibet has served as my means to seek a bridge—or as an exercise of translation—between many of the valid (and some invalid)1 grievances of Tibetans (both in exile and in Tibet), and the leftist developmentalist perspective from which much of the Chinese social science scholarship and leadership views Tibet. Since the 1950s, the former grievances have come to be increasingly expressed through the language of human rights and liberal democracy, which Chinese Communist interlocutors have tended to interpret as a disguised and disingenuous expression of Western imperialism, or at least of US foreign policy interests. In many cases this interpretation is no exaggeration. Indeed, one of the ironies of Tibet is that

xxxii

Preface: Reflections on Researching Development in Tibet

throughout much of the post-imperial history of Tibet (i.e., throughout the second millennium), various Tibetan elites have tended to associate and ally themselves with various outside imperial forces, be they Mongol, Manchu, or British. With no exception, exiled Tibetan elites entered the post-World War II period of decolonization firmly aligned with the United States, rather than with the emerging Non-Aligned Movement in support of decolonization, even despite the leadership of India—the host country of the exiles—within this movement. This was understandable because Tibet was in the process of being occupied by one of the leaders of the socialist anti-imperialist struggle against the West and a clear enemy of the United States. Nonetheless, Tibetan elite affinity with Western imperial powers long pre-dated this specific conjuncture. There is, however, potential to turn this propensity on its head by interpreting Tibetan grievances through the lens of a structuralist development economics tradition, with which social science and economics scholars in China are familiar (albeit through a particular Marxist lens). Indeed, this idea was implied by Rong Ma (1998), in his contribution to a volume on development in Tibet, edited by Graham Clarke and based on papers presented at the first panel ever convened on the subject of development at an international Tibet studies conference, in this case the seventh seminar of the International Association for Tibetan Studies in Graz in 1995. In his contribution, Ma proposed applying a framework put forward by Hechter (1975) regarding national development as a function of political and ethnic as well as economic differences, building on the “core-periphery” framework in the version of dependency theory elaborated by Frank (1971) for application to Latin America. The spirit of this book is similar, albeit informed by the earlier and more nuanced contributions from Latin American scholars who were largely unknown to Northern academia given that they wrote and published in Spanish or Portuguese, in contrast to the more dichotomous and mechanical neo-Marxist approach of Frank, who unfortunately remains the predominant reference point for dependency theory in the Northern development studies literature.2 This bridging endeavour is especially pertinent given that the political stalemate on Tibet is unlikely to be resolved in the foreseeable future, whereas in the meantime the sheer speed of socio-economic transformation currently under way in Tibet has been and continues to be tumultuous. The potential marginalization faced by Tibetans in these transformations, particularly due to their severe lags in schooling compared to the Han Chinese majority, makes it paramount that we strive to address the character of development in Tibet now, as it is evolving, regardless of the absence of political resolution. Indeed, even a future hypothetically independent Tibet would most likely remain severely dependent on China given the current



Preface: Reflections on Researching Development in Tibet xxxiii

state of human resources in Tibet, particularly in more technical fields. As a corollary, the arena to influence decision making on the present and future of Tibet for better or worse lies in China, particularly in Beijing. Through this approach, I have had some success in opening up spaces of dialogue with Chinese scholars and policy makers, although to a certain extent these encounters might have also been part of a broader government strategy of engagement with Western scholars of Tibet and have probably had little substantive effect on Tibet policy. I was invited several times to present my research on Tibet in Beijing (although not since 2008). The presentations were always politely received. In particular, my presentation at the Fourth International Seminar on Tibetan Studies in Beijing organized by the China Tibetology Research Centre (CTRC) in October 2008, about six months after the spring 2008 protests, garnered much attention and some controversy, and proved popular with Tibetan scholars as well as with many Chinese scholars, especially the younger ones, even though I was explicitly critical of government policy during my presentation. I was subsequently told that the presentation (and my work more generally) had an influence on the groundbreaking report on the causes of unrest in Tibet in 2008 published by the Gonmeng group in China in 2009 (the “Open Constitution Initiative,” a think tank founded in 2003 by some of China’s most prominent liberal lawyers and university professors). This included my suggestion during my seminar presentation that the government has ironically re-created strong class divisions in the TAR, if not a new aristocracy, through its subsidization and related policies. Indeed, my 2005 book, which had been translated into Chinese by scholars at the CTRC and then distributed as a “closed publication,” was cited extensively in the Gonmeng report. However, following my talk, there was some murmuring that I had written in one of my publications that China had “invaded” Tibet in 1950. This was deemed unacceptable, a political faux pas in China that transgressed the limits of suitable public discourse on matters related to Tibet. One of the conference organizers asked me with much consternation whether I had indeed written such words. It took me a while to realize which publication, exactly, they were referring to, as I had written several commentaries and opinion pieces in the Western media on the Tibet protests that were strongly critical of Beijing. To their credit, the conference organizers gave me about ten minutes to make a public explanation in our panel at the end of the conference. I explained that, yes, indeed, I had used the word “invade,” although I was simply citing the famous work of Melvyn Goldstein (1989b), The Decline of the Lamaist State, in which one of the chapters was titled “The Invasion of Chamdo.” Notably, Goldstein is highly respected in these circles for his evenhanded intellectual rigor and pioneering work on so many issues. Luckily, I recalled from a conversation

xxxiv

Preface: Reflections on Researching Development in Tibet

with him in 2007 that when his seminal history was translated into Chinese in China, the censors changed the title of this chapter to “The Peaceful Liberation of Chamdo.” So, I explained to the audience that the consternation and confusion was the product of their own censorship, given that I had simply been following the authority of Goldstein on this matter. The irony was greeted with humor. Even a senior Chinese scholar-official who had been politically correcting my arguments and opinions throughout the conference smiled and accepted that the academic standards to which we both had to comply were quite different—I would be ridiculed by my own colleagues were I to adopt the standards of China, which he nonetheless deemed the (officially sanctioned) correct ones. On the other hand, my bridging exercise has, in a few cases, been ferociously denounced by nationalist Tibetan exiles and their supporters; this has been ironic given that my work has been generally well received by the Exile Government and Tibet Support Groups (TSGs). In April 2004, following a presentation that I gave in Victoria, Canada, an elderly and aristocratic member of the exile community and former minister of the exile government stormed out of the lecture hall in disgust during the question period. Before leaving, he criticized me, saying that even the idea that there could be “development” in Tibet under Chinese rule, or else that official statistics could be treated with any degree of legitimacy whatsoever (even if these would be used to show negative developments), is to appease China. A year later, during the launch of my first book in May 2005 at the London School of Economics, a US journalist and seasoned China expert living in London and well known for his ardent criticisms of China’s human rights abuses took the floor to denounce my book as being essentially complicit with China’s occupation of Tibet. His criticism was that I did not question the occupation but accepted the de facto status quo as my starting point, and that I made almost no mention of the horrors of the Maoist period and the Cultural Revolution in Tibet (which was not entirely true). Discussing development in Tibet without these qualifications is deplorable, he alleged. Similarly, I somehow earned the honor of being blog attacked by Jamyang Norbu in June 2008, who lumped me together with several other people of various persuasions and differing opinions, and called us all “barefoot experts,” doing more harm than good despite our good intentions but poor investigations. Admittedly, he was incensed by a somewhat provocative comment I wrote in the Guardian following a series of protests during the Olympic torch relay through London and Paris. I had made an appeal to seek a more moderate approach that would support talks between the exile government and Beijing (note that the Dalai Lama had openly supported the Olympics and thus the torch relay protests were making him look either untrustworthy or else not in control of the Tibet support movement in the eyes of Chinese leaders). I had also appealed to avoid alienating broader



Preface: Reflections on Researching Development in Tibet xxxv

Chinese public opinion, which contained elements that were potentially quite sympathetic to Tibetans. Norbu compared my argument to that of pro-Nazi appeasers in Britain in the 1930s who argued that attacking Hitler or criticizing German rearmament weakened the position of “moderate” Nazis who wanted to negotiate with Britain. Obviously, these accusations and their emotion and fervor stem from a particular worldview that equates the CPC with the Nazis, particularly post-Tiananmen when nationalism has increasingly displaced socialism as the legitimizing ideology of power in China (but also pre-Tiananmen, in a Cold War context of ardent anticommunism). My response to such accusations had usually been to evoke the instructions of the Dalai Lama himself, that we should seek friendship and dialogue with China. However, any possibility of such dialogue is foreclosed if we base our starting position on a high moral ground of human rights (specifically, civil and political rights conceived in liberal terms). The foreclosure is not because of an imagined orientalist aversion to such rights, but because the posturing of such rights discourses has come to be associated with the rhetoric of US foreign policy and hence as an expression of US interests.3 The high moral ground appears especially hypocritical when asymmetrically and exclusively applied to China by people based in Washington or London in an era of resurgent and intransigent US and UK militarism, with a blind eye turned to the egregious abuses of the enemies of one’s enemies or else to the prolonged and bloody occupations of Afghanistan and Iraq. Hence, dialogue with China requires that certain sensitive issues be put aside, at the very least so that attempts at dialogue are not censored at the outset. This allows for addressing the urgency of the present, that is, the extreme rapidity of changes facing Tibetans now and their marginalization within these changes. I have intentionally chosen a language of development rather than human rights to address this urgency precisely because this is the language adopted by Chinese scholars and policy makers in China, and also because it is an appropriate language from which to discuss and analyze the structural and institutional dimensions of rapid change and potential marginalization in Tibet. Similarly, I avoid reference to common terms used by the exile government and TSGs such as “cultural genocide.” While powerfully evocative, the use of the term “genocide” is not at all helpful for cultivating dialogue with Beijing and, more importantly, it is not particularly representative of realities in Tibet, where the government has financed numerous cultural and literary outlets for decades. I realized this even before going to Tibet for my research, during my time living in the exile community in India, and particularly once I obtained a moderate degree of fluency in Tibetan. Most of the institutions in the exile community at the time (the late 1990s

xxxvi

Preface: Reflections on Researching Development in Tibet

and early 2000s) that required a high level of competency in contemporary written Tibetan (including literary institutes as well as newspapers) were effectively being run by recently arrived Tibetans from Tibet, that is, Tibetans who grew up and were educated in Tibet, and had arrived in India for the most part during the 1990s. These leaders of Tibetan literacy in the exile community were disproportionately from parts of Amdo in Qinghai, where a strong educational and literary tradition in Tibetan language had been maintained throughout the Maoist and reform periods despite being on the supposed frontline of assimilation and Chinese inmigration. This raised my awareness of the degree of cultural and linguistic preservation and revival that was happening in this region, much of it funded by the central or various provincial governments in China, or at the very least permitted by them. This perception was later confirmed by the excellent groundbreaking edited volume by Goldstein and Kapstein (1998), Buddhism in Contemporary Tibet: Religious Revival and Ethnic Identity, as well as by my own fieldwork.4 Indeed, some of the best texts for studying modern spoken Tibetan as a second language were produced by state-owned printing presses in Lhasa, although these required knowledge of Chinese to be able to read the instructions. They also provided fascinating insights into the norms and boundaries of state-sanctioned discourse in Tibet in the 1990s. These were the preferred teaching texts of my two Tibetan teachers, one an ex-monk from Tashi Lhunpoe monastery in Shigatse and the other a journalist from Lhasa, both bona fide political exiles (unlike the majority of newly arriving exiles in India at the time). The former was teaching Tibetan to Tibetan high school graduates in the exile community with the objective of bringing their written Tibetan up to a level of proficiency that was sufficient for working in the exile government. Some years previously, the exile government had shifted their policy from writing official documentation in English to writing in Tibetan, with the effect that there was a severe shortage of Tibetans raised and educated in the exile community who were able to meet the task given that their standards of written Tibetan were substandard for such purposes, even after twelve years of exile schooling. By far the purest and most proficient Tibetan was undoubtedly to be learned from educated recent arrivals, before their own Tibetan became mixed with the exile dialect, replete with various English and Hindi hybridizations. The research of this book started from a deep appreciation for this linguistic and cultural vitality within Tibet, much of it state sponsored, as well as for the incredible power of language as a tool of both oppression and resistance. Indeed, language has come to play a central role in my understanding of exclusion and marginalization, but not because of linguistic degeneration as implied by the idea of “cultural genocide.” Rather, despite enormous efforts and much success by Tibetans to preserve



Preface: Reflections on Researching Development in Tibet xxxvii

and modernize their linguistic traditions, language plays a crucial role as a tool of political and economic dominance and subordination by the Han Chinese in Tibet, serving as a key mode of bias operating in most aspects of political and economic life outside the rural areas and small towns. It is in this dialectical tension between revival and subordination that the ethnicized dynamics of development in Tibet need to be understood. The ways that language shapes even fundamental perceptions of reality or, at least, modes of communication and hence representations of reality were poignantly demonstrated to me during the abovementioned conference in Beijing in October 2008. I was having dinner at a banquet table at the end of this conference with eight leading Chinese scholars of Tibet studies, four of whom were from Taiwan and four from the PRC, as well as one senior Tibetan government official from the TAR. I was invited to the table by the Tibetan official, who had attended my presentation and was quite enthusiastic about it. He was senior and secure enough that he was perhaps the only among dozens of Tibetan and Chinese participants to have actually publicly mentioned during the conference the protests that occurred earlier that year (which everyone else had been avoiding as a taboo). In the course of the panel discussion, he reminded the participants that there were serious problems in the TAR and other Tibetan areas that were not being addressed (he made this comment before my presentation). His status as a veteran official probably allowed him to speak candidly without censor—he was from the older generation of Tibetan leadership, with hardy complexion and eloquent Lhasa Tibetan, and he was definitely more comfortable in Tibetan than in Chinese. As I sat down beside him, we started conversing in Tibetan. In the midst of trying to explain to him that I preferred researching in the Tibetan areas of Qinghai rather than in the TAR because Qinghai was much more relaxed and less tightly controlled, I was searching for the words “tightly controlled” in Tibetan. I opted for the words “tight law” (trim tampo). He immediately rebutted (in Tibetan), “There is no law in Tibet! There is none at all! There are only police and soldiers, but no law.” I asked, “Don’t you have to be careful saying such things?” He answered, while scanning the people around him, “Yes, you have to be very careful.” Perhaps the most striking aspect of his response was not that he held such views and would express them in private, but that he was completely comfortable saying them out loud at the table because none of the Chinese scholars present actually spoke or understood Tibetan, at least not more than a few ceremonial words such as tashi delek, given that their scholarship on Tibet had been conducted entirely in Chinese. Indeed, I was introduced at this conference to some of the leading Chinese scholars of Tibetan studies in China—the grandfathers of the field, so to speak, who had

xxxviii

Preface: Reflections on Researching Development in Tibet

taught over three decades’ worth of students—and few of them spoke even a modicum of Tibetan despite having dedicated their entire careers to the study of this people. The dinner experience was a striking demonstration of how very different realities are recounted depending on whether one converses in Tibetan or Chinese. One invariably receives a different account from Tibetans if one speaks to them in Chinese, if only because others (Han Chinese) might be able to understand them, whereas Tibetans talking among themselves in Tibetan will be much more forthright, in part because most Chinese people cannot understand them. It was partly in this spirit that I was told by many educated Tibetans in Qinghai that even though they might maintain cordial relations with many Han Chinese people, there were certain subjects that they would simply avoid (such as politics) because of the fundamental gulf of understanding that their Chinese interlocutors very rarely if ever proved themselves capable of overcoming. It was definitely in this spirit that I conducted most of my fieldwork in Tibet in Tibetan or else, in the case of dialects that I could not fully understand (such as Golok or Amdo dialects), through an interpreter from Tibetan to English, but rarely through Chinese (except in the case of interviewing Chinese or Muslim respondents). I have often suggested that conducting fieldwork in Tibet in Chinese is akin to doing fieldwork in Palestine in Hebrew. This linguistic divide and the associated gulf in worldviews help to clarify the fundamental tension in China’s Tibet policy between autonomy and assimilation. Many Chinese scholars rightly recognize the linguistic biases and disadvantages faced by Tibetans, but then most (whom I have met) conceive of the solution in terms of stronger and implicitly more assimilationist integration (or else “acculturation”), especially through better Chinese-medium education, perhaps starting as early as primary school, so as to prepare Tibetans for the de facto domination of the Chinese language in the wider context into which they are integrating. The option of strengthening the use of Tibetan in Tibetan areas, through the education system and also through employment practices, is rarely considered by Chinese as realistic, even though it has been shown on several occasions that Tibetans (and most linguistic minorities) perform much better in all subjects including Chinese if they are taught in their mother tongue (i.e., teaching Chinese as a second language). In this respect, there is even a rift in the understanding of “bilingual education.” Tibetans usually use the term to refer to strengthening the use of Tibetan-medium in schools, ideally across all subjects, so that Tibetans can be guaranteed full competence in their own language as the basis for learning other subjects and languages. In contrast, Chinese usually use the term to refer to bringing in more Chinese-medium earlier in the minority schooling system in order to improve the Chinese language



Preface: Reflections on Researching Development in Tibet xxxix

skills of Tibetan students, even if this means eroding the quality or the coverage of their Tibetan-medium education or even their Tibetan literacy (which is generally quite poor in the current system, unless students specifically specialize in linguistic or cultural subjects or else attend Tibetan-medium schools that have been strong in this respect). Even if there is an appreciation for the heritage of the Tibetan language, this latter option is usually advocated as a modern and inevitable necessity to help Tibetans to function and perform in the linguistic mainstream of China, much as all US citizens must seek fluency in English as a primary credential of their citizenship. The problem, of course, is that this common Chinese perspective simply does not grapple with the indigeneity of Tibetans in Tibet and with the very real historical reality that prior to 1950 most Tibetan areas and definitely the area now incorporated as the TAR were essentially self-governed by Tibetans and in the Tibetan language. Moreover, the sudden and radical shift in the 1950s to rule by outsiders who neither spoke Tibetan nor understood much about the people they had subjugated was forced and was embedded within an ideological animosity toward multiculturalism (and religious culture) until the 1980s. It is also wishful thinking that assimilation—linguistic or otherwise—can solve deeply entrenched systems of discrimination and disadvantage. Rather, it usually involves the reproduction of discrimination and disadvantage on a new assimilated footing, given that these are more fundamentally rooted on asymmetrical power relations, vastly unequal in the case of Tibet. As my critics from the Tibetan exile side rightly point out, it will only be when these power relations are seriously and sincerely addressed that we will begin to be able to conceive of a more just path of development for Tibetans in Tibet. It is my hope that this book might encourage steps in that direction. In the meantime, development in Tibet continues to proceed in a broadly disempowered manner. We can only hope that the marginalizing tendencies will be somehow mediated, first and foremost by the initiatives of Tibetans themselves, or else through benevolence, oversight, or even neglect from the side of the state.

Notes  1. For instance, in one rural Tibetan village I visited in Qinghai, locals complained of being taxed by the local government at the rate of one sheep per year per family. Considering that families typically had twenty to forty sheep, this amounted to an annual tax rate of 2 to 4 percent of their hoofed wealth. While the perception of injustice was partly related to the fact that the authorities were perceived as lacking legitimacy, and also that the tax was regressive, rural Tibetans in Tibet admittedly have an attitude toward taxation that could be roughly compared to that of the Tea Party in the United States. Indeed, my commitment to this people

xl

Preface: Reflections on Researching Development in Tibet

has always had to sit uncomfortably with the widespread popularity of George W. Bush among Tibetans in both Tibet and exile.   2.  The eurocentric confusion was most poignantly described in the classic article by Fernando Cardoso (1977) entitled “The Consumption of Dependency Theory in the United States.” He argued that US academics were not aware of the contributions of dependency theorists when they were writing in the 1960s (and even 1950s) because most of them were writing in Spanish and Portuguese. Northern awareness of the concept of dependency either stemmed from the work of Paul Baran or else began when a few Latin American articles were translated in the 1970s. However, even then, most attention was directed toward Frank and Wallerstein (who were not Latin American) because they were writing in English. Due to this focus, Cardoso argued that the understanding of dependency theory was reduced to that of structural-mechanical formal models aimed at finding “laws of motion” based on patterns of integration into the world economy. This, he contended, completely misread the original purpose of dependency theory, which aimed to elucidate a dialectic focused on internal processes of exploitation in conjunction with external processes of dependency. The combination could, in certain periods, lead to growth, even lots of growth, as it did in various periods of Latin American history, but this growth would tend to lead to an intensification of the internal processes of exploitation. However, even his insights were lost as Anglo-American academia continued to assert a grossly simplified and misconstrued understanding of this rich body of Latin American thought, ironically drawing mostly from non-Latin American authors. Of course, even Cardoso expressed these ideas through a Marxist sociological language, although he argued that analysis must be based on concepts of social class that are rooted in indigenous realities rather than being simply imported from European social thought. For a balanced retrospective presentation of this whole tradition of Latin American thought and a powerful and more recent reformulation of the eurocentricity criticism, see the seminal book by Kay (1989).   3.  See Adams (1998) for a similar and interesting reflection on the problems of emphasizing human rights in the Tibetan context.   4.  Despite being a target of some of his criticisms, I largely agree and greatly appreciate the arguments of Zenz (2013) in this regard.

1 Introduction: The Disempowered Development of Tibet in China

In March 2008, a wave of large-scale demonstrations quickly fanned out from Lhasa to the Tibetan areas of Qinghai, Sichuan and Gansu over the course of about three weeks.1 Despite premonitions of mounting tensions and brewing discontent that had been evident for years, including demonstrations in Amdo that had preceded those of Lhasa,2 the usual developmental alibis were nonetheless presented through various state organs in China even as the demonstrations were still ongoing. The government argued that the “riots” (in reference to the riot that took place in Lhasa on 14 March) were due to political meddling and manipulation from abroad, particularly from the Tibetan exile community and their Western supporters. Given that the region had been experiencing ample development and rising prosperity, the government argued that local Tibetans had no valid cause for grievance, certainly not that they were being discriminated against. If anything, Tibetans in the Tibet Autonomous Region (TAR) have been far more privileged than any other minority group in China, to the extent that many in China view their privilege as unfair. This line was supported by many Chinese and some Western scholars and commentators who, at the extreme, tended to interpret the events as the covert handiwork of US neocons, with a reasoning that Emily Yeh (2009a) astutely identifies as “radical reductionism.” When the events and their aftermath calmed down, some of these scholars offered more nuanced messages, admitting that there had been some problems with recent development in Tibet, in particular, rising inequalities, corruption, and competition with migrants, albeit with the corollary that these problems were no different from those elsewhere in China, 1

2

Chapter 1

hence requiring no special treatment for Tibetans beyond more and better development. The calming also allowed for a more prominent voicing of critical perspectives within the government and Chinese academia. Perhaps the highest profile to date (at the time of writing) was the presentation by Jin Wei, director of Ethnic Religious Studies at the Central Party School of the Communist Party of China in Beijing, at a forum on ethnic minority development hosted by Minzu University in Beijing on 15 December 2012 (see Lan 2012). In her presentation, based on a study of 987 central government aid projects to the TAR between 1984 and 2005, Jin called for a policy overhaul of aid programs in the TAR. She noted in particular that many programs had failed to consider cultural contexts, among other concerns such as poor efficiency.3 A number of other scholars at the forum followed suit by stating that aid policies to ethnic regions need to adapt to local conditions and include participation from local communities. From this perspective, it might well be the case that ongoing protests in Tibet have been creating the space for serious reevaluations of Tibet policy in Beijing, although as analyzed in this book, such soul-searching has nonetheless taken place in a policy context of much more of the same, given that central government subsidies to the TAR surged to record-high levels in the two years following the 2008 protests. At the surface, many signs do suggest that Beijing has been prioritizing the TAR and, we are led to assume by implication, Tibetans. To start with, the TAR has always been privileged with the highest levels of direct budgetary subsidies in China since the 1960s, with subsidies per person ranging in many multiples of what other provinces receive. Even in the early 1990s, at the depth of the collapse of regional fiscal redistribution in China, direct budgetary subsidies from the central government (not including subsidized investments and other indirect subsidies) consistently covered well over 90 percent of local TAR government expenditures, which in turn never fell much below 50 percent of GDP, even at their lowest point in 1992. In contrast, most other western provinces were largely forced into self-reliance until the mid-1990s. In Qinghai, the next most subsidized province, expenditure was less than half subsidized by the central government by 1993 and this expenditure amounted to only 10 percent of GDP, despite similar levels of poverty and human development needs as in the TAR. In Gansu, the second-poorest province in China (poorer than the TAR according to per capita GDP), only 18 percent of local government expenditure was subsidized by Beijing in 1993, accounting for only 3 percent of GDP. Since this trough, subsidies have increased in all western provinces as a result of successive waves of western development strategies since the mid-1990s. Subsidies have increased both in real per person values and as proportions of government expenditure and GDP. The TAR has nonetheless maintained its relatively privileged priority, to the extreme and somewhat



Introduction 3

perplexing extent that direct budgetary subsidies from the central government exceeded 100 percent of the GDP of the TAR for the first time in 2010. This was higher than even the heights of intensive subsidization during the Maoist period. Such priority has simply not been observed in any other province of China—the next-highest levels were observed in Qinghai, where direct budgetary subsidies reached 47 percent of GDP in 2010. In Gansu, still the second-poorest province of China, they reached 27 percent of GDP in the same year. If this is our meter, then the TAR has definitely maintained its distinction of fiscal privilege. These headline figures must be seen in light of the evolving context of regional development strategies in China since the mid-1990s. In response to economic lagging in its western hinterland (much of it comprising of Tibetan autonomous areas), the central government started several major policy initiatives in the early 1990s aimed at correcting the increasing disparities between the economic takeoff of coastal China and the stagnation and even recession in large parts of western China. These initiatives started with a major poverty alleviation campaign (the “8-7 strategy”) in 1994, which disproportionately focused on central and western China by implication of the fact that these regions were the poorest. Most of the counties of the TAR were targeted by this campaign, although the province was not necessarily among the poorest in western China, whether in terms of per capita GDP or even average rural household incomes. Most of the Tibetan autonomous areas in other provinces were similarly targeted, although with less intensity than the TAR due to the symbolic political importance of the latter. Indeed, the national plans were complemented by the Third Tibet Work Forum of 1994, which focused exclusively on the TAR. Poverty alleviation campaigns were followed up with various western development strategies under the ninth and subsequent Five-Year Plans. The Ninth (1996-2000) prioritized the western region for the first time, as had been explicitly planned as early as the Seventh Five-Year Plan in the mid-1980s. This initiative was stepped up further with the announcement of the “Open the West Campaign” in 1999 (OWC; xibu da kaifa),4 in conjunction with the Tenth Five-Year Plan starting in 2000 and the Fourth Tibet Work Forum specifically for the TAR in 2001. More recently, the Eleventh Five-Year Plan (2006-2010) continued the commitment to balanced development across regions, including a prioritization of social development in addition to economic growth, or what the government refers to as a “people-first” approach to development. As noted by Goldstein et al. (2010), the TAR specification of the plan explicitly called for an allocation of fiscal resources to projects that engage directly with village households as a means to improve rural livelihoods and quality of life, although this observation is representative of broader policy trends throughout China and was not particular to the TAR. The Twelfth Five-Year Plan (2011-2015)

4

Chapter 1

maintains this approach, now framed by the discourse of domestic rebalancing, which implies a shifting emphasis from investment toward consumption and from development in urban and coastal areas toward rural and inland areas. This was again complemented by the Fifth Tibet Work Forum in January 2010, although for the first time in the history of Tibet Work Forums, the focus of the Fifth was extended to include all Tibetan autonomous areas in China, not just those of the TAR. These successive plans, strategies, and forums have served to maintain the commitment to western development high on the government agenda, including the prioritized subsidization of the TAR and increasingly the other Tibetan areas outside the TAR as well. Indeed, the extension of Tibet policy focus in the Fifth Tibet Work Forum to all Tibetan autonomous areas inside and outside the TAR was probably in recognition of the fact that the 2008 protests occurred throughout these areas. Eastern Tibet has also become an increasing priority for the central government due to its perceived importance for downstream ecological concerns in China. As a result, subsidies to the TAR and these other Tibetan areas have never been higher, both in per capita terms and relative to the local economies of these areas. This high degree of priority nonetheless amounts to a very small share of the total subsidies that the central government directs to the western provinces, most of which are focused on matters not necessarily connected to Tibet policy, such as the development of inland industrial clusters around the major western cities of Chongqing, Chengdu, Lanzhou, and Xi’an. The Tibet-specific subsidies are thus sustainable from a central budgetary perspective. The combined efforts of the first wave of western development strategies in the mid-1990s brought a tidal reversal in regional fiscal flows, back toward the subsidization of western China by coastal China via the fiscal intermediation of the central government, which regained control of over 50 percent of national fiscal revenue through its tax reform in 1994 (up from less than one-quarter in 1993). The pattern of regional fiscal redistribution that had been established in the Maoist era but then collapsed in the early years of the reform period in the 1980s was thereby restored. In the early reform years, fiscal decentralization allowed provinces to retain most of their surplus revenue in excess of a quota for local spending and investment, which left the erstwhile heavily subsidized and deficit western provinces floundering in a fiscal drought in the midst of the gradually unfettering takeoff of coastal China. The reversal of this trend, crucially supported by the tax reform in 1994, was a welcome respite for provincial governments in western China. The resurgent subsidies became a key factor in facilitating these provinces to join the national takeoff, in tandem with other policy changes occurring in the early to mid-1990s. The two subsequent Five-Year Plans in the first decade of the twenty-first century



Introduction 5

maintained and enhanced this redistributive fiscal transfer from east to west. Given that the TAR and Qinghai (which together account for about three-quarters of Tibetans and Tibetan areas in China) have been the most subsidized provinces in China, both provinces have been particularly susceptible to these trends, at least in terms of the sectors that have been dependent on fiscal transfers. The resurgent subsidies resuscitated growth in western China and especially in the TAR and Qinghai, where growth accelerated to very rapid, above national-average rates in the late 1990s. The speed of economic growth in the TAR and Qinghai over this period was phenomenal, even by Chinese standards. For instance, the nominal gross domestic product (GDP) of both provinces grew at a rate about one-third faster than the national economy from 1997 to 2010, even though the national experience has been perhaps the fastest (and definitely the largest) experience of sustained rapid economic growth the world has ever seen. Indeed, from 1997 to 2007 (the year before the outbreak of large-scale protests), the GDP of the TAR quadrupled, versus a tripling of the GDP of China as a whole. The pace of change has been astonishing, as has been the extent of subsidization driving this change. Growth has undoubtedly led to trickle down, especially since 2003 when an increasing portion of subsidies has been directed at (Tibetan) rural areas and households, as per the principles of “harmonious” and “people-first” development. As a result, there is no longer any doubt that macroeconomic growth has been reflected as rising prosperity for many Tibetans. Poverty rates have been falling, average household incomes rising, and levels of material consumption rising, whether in butter and meat consumption or in motorcycle and mobile phone ownership. Levels of educational attainment have been moderately improving, as well as various aspects of social service provisioning (albeit these remain paltry in most Tibetan rural areas). Were it not for some restrictions on cultural or religious practice, tight control over the political realm, and persistent political repression, or else the worrisome rise in inequalities (which has nonetheless been a problem throughout China), it might indeed appear that the government has been tending well to the disadvantaged position of Tibetans in China. Many Chinese have gone so far as to conclude that Tibetans (referring to the TAR), are lavishly pandered by central government subsidies to a far greater extent than any other minority nationality or western province in China and are quite simply spoilted, ungratefully complaining with their stomachs full. Some Western scholars have also stood up for Beijing along these lines. For instance, Barry Sautman, one of the more prominent non-Chinese scholars writing on these issues in defense of Beijing and with an apparent obsession to discredit the Dalai Lama, argued along these lines soon after the 2008 protests occurred (see Sautman 2008). He suggested that

6

Chapter 1

even rural Tibetans (in the TAR) receive much more state subsidies than other minorities in China (including, we must presume, Tibetans outside the TAR). His charge more or less repeated arguments he made with Irene Eng seven years earlier. In their previous article, Sautman and Eng (2001) elaborated an urban bias thesis to interpret the situation in Tibet, arguing that much of the apparent disparity between Tibetans and Han Chinese observed in the TAR in the 1990s was due to the urban-rural disparities that are inherent to the imbalanced processes of development throughout China. Inequality between these nationalities, they argued, is largely a reflection of the fact that most Tibetans are rural and, unlike in any other province in China, rural populations in the TAR are almost entirely Tibetan (which is true). They contended that the severity of the Han-Tibetan disparity is thus a reflection of the poverty of these rural areas, in contrast to the emerging subsidized affluence of the urban areas of the TAR, which was not without problems but was nonetheless a sign of benevolence and preferentiality on the part of the government. In this sense, ethnic inequalities should not be seen as alarming as such, particularly in light of the urban-led development that has been leading more and more Tibetans away from rural poverty and into the emerging so-called “middle-class” of the urban areas. Sautman reiterated the essence of this argument in a 2012 article in which he suggests that the self-immolations in eastern Tibet in 2011 and 2012 were isolated events concentrated around Kirti Monastery in Ngawa County, Sichuan, as a result of resistance to patriotic education campaigns and the subsequent expulsion of a large number of monks. However, he stresses that “there is no repression of Tibetans simply for being Tibetan.” Instead, the hard-line government responses to these Tibetan self-immolation protests are actually responses to threats of separatism, which he argues the government has the international right to render illegal: . . . the Tibetan areas, like the rest of China, have plenty of human rights problems. Economic development benefits urban areas—where Han Chinese tend to congregate—rather than the largely Tibetan countryside. There is also unrestricted migration to Tibetan areas, although the percentage of Tibetans in the population has remained largely stable over the past several decades. In some parts of the Tibetan Plateau, there are also clumsy efforts to force Tibetan monks to disavow the Dalai Lama as a political leader. Most protests in Tibetan areas have focused on these economic and social issues, rather than the political status of Tibet, but separatists have sometimes raised their slogans during these protests, and this has brought on repression.

Moreover, he writes that the expelled monks “have few alternative life opportunities,” implying, it would seem, that the actions of the self-immolators were a response to the prospects of destitution. It is disturbing how a scholar who has often denigrated religious nationalism in Tibet would take



Introduction 7

it upon himself to offer an interpretation of what motivates these religious people to commit “suicide” (his choice of words to describe self-immolation). In this opinion piece, he also refers to the one day of rioting in Lhasa on 14 March 2008 as “a mini-pogrom against Han Chinese.” Recalling that a “pogrom” refers to “an organized, often officially encouraged massacre or persecution of a minority group, especially one conducted against Jews” (per the American Heritage Dictionary of the English Language, 4th ed., 2000), his comments are remarkably in line with accusations published almost simultaneously by China Tibet Online, a state-run website, and soon thereafter by the Xinhua News Agency, that the Dalai Lama has been advocating “Nazi-style racial segregation ideas” (AP 2012). More generally, Sautman’s perspective here is similar to the standard official approach in China of viewing the problems of Tibet in terms of a cocktail of “ecological poverty” (meaning chronic poverty in remote, hard-to-access mountainous regions) mixed with separatist agitation organized by small groups that take advantage of such poverty despite sincere government attempts to address it. Indeed, this logic is comparable to the global phobia regarding the potential of poverty to fertilize terrorism. The underlying logic of these arguments is also echoed in Hu and Salazar (2008), who argue that urban-rural prejudices among both local Tibetans and migrant Han in the TAR are a much greater driving force for social stratification and status than inter-ethnic prejudices. Moreover, they argue that Han prejudices toward Tibetans have no “bite” for urban Tibetans (17). Rather, they characterize urban permanent resident Tibetans as elites (and many from the past aristocracy), who look down on both newly arriving rural Han migrants as well as rural Tibetans, and they claim on the basis of an unavailable Lhasa government website that only 30 percent of government cadres in the TAR were ethnic Han in 2005 (10). On this count, they completely ignore official and publicly available statistics in TSY (2004) showing a sharp increase—both proportionate and nominal— in non-Tibetan cadre employment in the TAR in 2003, and a corresponding collapse in Tibetan cadre employment, the share of which fell below 50 percent for the first time since 1980. The government subsequently stopped publishing this ethnic breakdown of public employment data, although as analyzed in this book, it is highly probable that the decline in the Tibetan share of cadre and other forms of public employment in the TAR continued after 2003. Presumably, the Lhasa government website data used by Hu and Salazar was in reference to the situation up to 2000, when Tibetans did indeed account for about 70 percent of cadres in the TAR. In this sense, their characterization of urban TAR (mostly Lhasa) might have had some relevance for the 1990s, but not to the more recent decade that they claim to analyze. Irrespective of the sloppy and tenuous use of evidence, and the complete absence of consideration of scholarly contributions presenting

8

Chapter 1

evidence to the contrary, they suggest that even in the case of the riots that erupted in Lhasa on 14 March 2008, which displayed “the single most intense expression of ethnic differentiation,” “the structural effect of the urban-rural divide is evident and important” (18). Notably, the theorization of Hu and Salazar is exceptionally Lhasacentric. It offers little insight into the widespread (and mostly peaceful) protests that spread throughout eastern Tibet and where the urban-rural distinction among Tibetans is much less pronounced. Indeed, Barnett (2009) argues that, in comparison to earlier protests, the 2008 protests were not primarily urban but involved new cross-sections of the population, including substantial participation from rural areas, which he claims had not been involved in previous protests during the reform period (such as the protests in and around Lhasa in the late 1980s, which were predominantly urban). My own reading differs somewhat from Barnett’s, in that the recent protests have been more specifically related to urbanizing rural areas (as well as urban centers) or to populations that have been increasingly in spatial flux in the midst of a crucial moment within rapid development transitions. Moreover, while the social origins of the recent and ongoing protests have been definitely wider than in the protests of the 1980s—and hence more rural in a sense—they have not been more rural in comparison to widespread resistance in the 1950s, which was predominantly rural in Kham and Amdo. This being said, these more subtle debates and distinctions share a common dismissal of the simplistic urban-rural dichotomies offered by authors such as Sautman, Hu, and Salazar as explanations for Tibetan grievance. In particular, one conceptual problem within these explanations is that grievance is inversely associated with living standards. Indeed, given the raft of indicators showing rising incomes and other conventional signs of development progress, it is understandable how those in the central and provincial governments (and related or supportive scholars) without an ear to the ground would have been so surprised by the explosion of protests in 2008. Obviously, even an ear to the ground would have needed to be fluent in Tibetan, which excludes most Chinese cadres serving in the Tibetan areas as well as most Chinese scholars of Tibet in China (in my experience, as discussed in the preface). Instead, up to the 2008 protests, many had assumed that the strategy of dealing with this conflict-prone region of China through the force of subsidies, growth, and improving livelihoods was bearing fruit and gradually winning over the population yuan by yuan. The protests required an aggressive domestic propaganda campaign to address the dissonance, although official government pronouncements continued to assert a message of development success and broad Tibetan support marred with scattered attempts at destabilization by a handful of separatists organized from abroad. Such narratives have rung increasingly hollow



Introduction 9

in the face of ongoing and increasingly broad protests, particularly those in eastern Tibet, even despite the efforts of scholars such as Sautman (2012) to provide intellectual rationalization. The dissonance between development success and mounting grievance points to the fact that absolute and even relative measures of living standards probably offer little insight into the reasons why so many Tibetans might have come to be so aggrieved amid the economic and human development of their homeland.5 Rather, contemporary development in Tibet—including the character of modern poverty and social differentiation—needs to be understood as deeply shaped by sixty years of radical and more or less forced incorporation and then integration into the vicissitudes of revolutionary and post-revolutionary nation-state consolidation under the People’s Republic of China (PRC). Within this process, marginalization is not simply a condition of backwardness due to a late or obstructed start along a roughly linear development trajectory. Marginalization is better understood as being continuously reproduced within development through the manner by which peripheral and disadvantaged populations are integrated into broader national or even global processes. The above-mentioned scholars such as Sautman might accept this perspective on marginalization with respect to Chinese development more generally, or else with respect to the challenges faced by de-colonizing nations to develop in the post-war period from their disadvantaged and subordinated position within the global economy.6 However, judgment seems to be abandoned when the Sino-centric gaze is turned toward the Tibetan highlands, as pointed out by Peter Hansen (2003) with regard to the reluctance of leading post-colonialist scholars in India and the West to apply their analyses to Beijing’s efforts to incorporate Tibet into China. The analogy of “kicking away the ladder,” to quote Ha-Joon Chang (2002)—that is, whereby all major “now-developed countries” try to forbid contemporary developing countries from using the same interventionist economic policies that they used in the past—serves well to characterize this asymmetry of perspective regarding sub-national development in the minority areas of China. The refusal to consider Tibet in such a light seems to be partly due to annoyance with the international attention that Tibet receives, particularly from the centers of capitalist power in the West, themselves involved in the invasion and occupation of countries. However, it is also because the obvious preferential treatment of the TAR by Beijing would seem to render such analogies invalid. Nonetheless, the rejection of this perspective does a disservice to our understanding of development in Tibet. It even does a disservice to our understanding of development in China given the particular value of examining centers through the lens of their peripheries.7 In this regard, the conviction of the central government that grievances and nationalism among Tibetans can be resolved through the force of eco-

10

Chapter 1

nomic growth and improved livelihoods is problematic if not simplistic given that it implies that the “Tibet problem” is a result of poverty and developmental “backwardness.” In reality, nationalism and grievance in Tibet—and in other parts of western China such as Xinjiang (East Turkestan)—have not subsided alongside growth and improved livelihoods. Rather than blaming bogeymen, the more plausible implication is that poverty is ill suited to explain these dynamics. Hence, part of the project of explaining the conflictive repercussions of economic growth and development in Tibet (and other regions) requires a careful differentiation of processes of marginalization and exclusion from those of poverty.8 There are obviously other reasons that are central to understanding why Beijing’s strategy to resolve the “Tibet Problem” might be doomed to failure (short of a “total solution”), such as political repression, patriotic education campaigns, particularly in monasteries, and various other forms of cultural and religious insensitivity or prejudice. However, the issue of development requires careful and special attention, in part because it is so central to the government’s strategy and discourse. This requires an analysis of the structural and institutional dynamics underlying rapid economic growth in the Tibetan areas in order to evaluate whether these dynamics might in fact be exacerbating discriminatory biases and accentuating certain aspects of subordination and marginalization within development processes even despite improvements in living standards, thereby contributing to social tensions rather than resolving them as claimed by Beijing. The aim of this book—and the original contribution of my work more generally—has been to provide a substantiated analysis of these structural and institutional dynamics of subordination and marginalization occurring within contemporary development in Tibet, as a complement to our wider understanding of the current situation in Tibet provided by other scholars, including political, legal, cultural, linguistic, religious, and historical perspectives. The period of study principally concerns the reform period in China and, more specifically, the mid-1990s onward, when western development strategies started to be implemented in earnest by the central government. The statistical work of the book ends in 2010, the last year of available data at the time of final editing. Of course, the timeline of the analysis could be taken back further, covering the whole sixty-year period since most Tibetan areas were incorporated into the People’s Republic of China in 1949-1951. Indeed, some chapters deal with some aspects of this longer perspective, although the earlier period of incorporation is largely left to the pursuits of other scholars. Here, the main focus is not on the status of Tibet as an “annexed,” “occupied,” or “liberated” part of China (however the reader wishes to consider its status, in relation to the TAR or to other regions of Tibet). Rather, the focus is on economic developments in the more contemporary period from



Introduction 11

1990 onward, when rapid economic growth occurred in tandem with a particularly assimilationist (or “acculturationist”) mode of integration into China following the imposition of martial law in the TAR in March 1989 and, almost three months later, in Beijing in May. This period in Tibet logically needs to be understood in the context of developments in China more broadly, especially the ideological transition from a more Socialist to a more nationalist orientation. Within this context, the particularly assimilationist mode of advancing the economic integration of the Tibetan areas into China has been publicly presented (and apparently broadly accepted by much of the Chinese public), despite considerable dispute and contestation within the Communist Party of China (CPC) itself, as if it were the only legitimate path of regional development plausible within the national strategies of China. The alternative that has been largely silenced since 1989 draws from the erstwhile more moderate and minority-tolerant approaches of the 1950s and 1980s, which were based on recognition of the benefits of some substantive local-level autonomy in Tibet and other minority areas. Similar to the earlier phase of radicalizing Maoism in the 1960s, these alternatives were eclipsed by the current assimilationist phase following 1989. In essence, the argument of this book is that the intensified economic integration of Tibet into regional and national development strategies on these assimilationist terms has, in turn, intensified various dynamics of subordination and marginalization faced by Tibetans of all social strata and despite the evident material improvements in their living standards. These dynamics are partly—although not entirely—reflected by rapidly rising inequalities within Tibetan areas that have accompanied different phases of rapid growth, some aspects of which had reached levels much higher than anywhere else in China in the 2000s. Taking the TAR as our exemplar, urban-rural household income inequality first reached an exceptionally high level as growth took off in the late 1990s. Following the shift in policy priority in 2003 toward raising rural incomes, this aspect of inequality quickly subsided, at least back down to the upper range of urban-rural inequalities observed across other western provinces in China. However, this aspect of inequality was replaced by rapidly rising inequality within rural areas and especially within urban areas from the early 2000s onward. Hence, to a certain extent, the tempering of interregional disparity (between provinces) has been achieved at the cost of rising intra-regional (within province) disparities. Nonetheless, such trends can be observed in other western provinces as well—they are not merely a Tibetan affair, as often noted by Chinese (and some foreign) scholars. However, the dynamics of inequality in the Tibetan areas have been qualitatively different from elsewhere in western China. They have been much more directly instituted by government policies—particularly subsidy and

12

Chapter 1

wage policies—than anywhere else in China given the preponderant degree to which such policies dominate the local economy and determine the places and the beneficiaries of affluence. Inequalities in other parts of China have been based much more on the dynamics of capitalist productive accumulation, including fairly classical political economy tensions between wages, profits, and taxes. In contrast, most activities outside traditional farming and herding (and the booming trade in caterpillar fungus) in the TAR and, to a lesser extent, in other Tibetan areas have been by and large the construct of subsidization policies. Even changes in the Tibetan rural areas have become increasingly dependent on subsidization, such as the subsidies driving investment into chicken production in rural parts of the TAR near heavily subsidized towns and cities, as discussed by Childs et al. (2011). Hence, privilege and polarization are driven much more by the hierarchy of position within the flow of these subsidies rather than by incentives and compulsions for primitive and productive accumulation, to the extreme extent in the TAR where changes in urban-rural and intraurban inequality are largely reflective of changes in the salaries of statesector employees (who accounted for about 12 percent of the labor force in 2010, much of this including Han Chinese migrants). These state-sector employees in the TAR benefited from among the highest salaries in China since the beginning of the OWC, neck and neck with those of Beijing and Shanghai for several years, which has had nothing to do with productivity or overall prosperity considerations in the TAR. It is related to the so-called “hardship allowances” of those posted to work in the TAR, although even these allowances rose sharply at the beginning of the OWC relative to other “hardship” posts in western China. Even among those people working outside the state sector, including the increasing influx of Han Chinese and Chinese Muslim migrants into Tibetan areas, the benefits of growth have been disproportionately channeled through institutional processes that reward positions of access to the highly externalized flows of wealth passing through the region with increasing velocity. There is no doubt that many Tibetans have benefited handsomely from this situation, albeit they represent a minority of the local population. They are members of an upper stratum that includes a minority of Tibetans and a large proportion of non-Tibetan migrants, both mostly concentrated in urban areas and well positioned within these state-centered channels of privilege. There has been significant trickle down as well, as noted above. However, the particular cocktail of restrictive state-sponsored privilege and intensified competition for urban employment within a context of rapid urbanization and open migration, combined in a strongly assimilationist atmosphere that has reinforced cultural and linguistic biases in the economy, needs to be understood if one is to come to grips with the widespread intensity of discontent that development has brewed in Tibet.



Introduction 13

Prior to the events in spring 2008, I had regularly predicted such tensions on the basis of this understanding (as elaborated further below). These were not particularly original or groundbreaking insights for those of us seasoned by fieldwork inside Tibet and western China. Tensions were obvious for Tibetans and even for Chinese scholars and officials with intimate knowledge of this region and its people, although neither Tibetans nor Chinese in China had been permitted to express such insights publicly, or else only at significant risk. Indeed, when I interviewed four senior Chinese Sichuan government officials in 2007 who were responsible for researching and advising the Sichuan government on developments in the Tibetan areas of this province, they admitted then that although their economic strategies had been thus far a success (from a government perspective), the political strategies of creating “stability” had been a dismal failure. Instead, in this and other provinces containing Tibetan autonomous areas, ongoing nationalist agitation against the state had been occurring alongside rapid development and change, as well as other forms of contestation around various social and economic issues, including the worrisome emergence (or re-emergence) of conflicts between Tibetans and Muslims in many parts of Tibet. The protests of 2008 were testament to this failure. The underlying tensions have continued more or less unabated in a subverted manner despite the severe clampdown and reinvigoration of patriotic education campaigns that followed the 2008 protests, re-emerging in 2011 and 2012 as a wave of self-immolations and related protests in eastern Tibet (specifically in Northern Kham and Southern Amdo, on both sides of the QinghaiSichuan border and including parts of Gansu and the TAR).

The Economics of Marginalization within Peripheral Development There is no doubt that development has nonetheless occurred in this context. However, given the resistance among some exiles to use the word “development” to refer to changes occurring in Tibet under Chinese rule, this statement requires some clarification. By development, I refer to long-run social, political, technological, and economic structural transformations that are distinctly modern (although not necessarily conforming to modernization theory). Development implies a fundamental transformation from a pre-transitional condition to some form of “modernity,” whether beneficial, benign, or perverse, without necessarily implying that there is a pre-destined teleological path or outcome within this transformation or that later “developers” will necessarily follow the same path as those that preceded them. Nonetheless, it does imply some discernible underlying structural features of transition. Tim Dyson (2001) argues that the most

14

Chapter 1

essential features are found in health, fertility, and related demographic and sociological transitions, including urbanization. Political and institutional aspects of modernity include the emergence of nation-states as the relatively recent although quasi-total modality of territorial sovereignty, together with bureaucratized forms of state governance, even though it is debatable whether such political modernity necessarily leads to democratization or, as argued by Karl Polanyi (2001 [1944]), to fascism when economic freedom is allowed to supersede social and political freedom. Technological and economic aspects of modernity essentially refer to the emergence of what Celso Furtado (1983) called “industrial civilization.” This is the structuring of an entire economy, including the consumption patterns of its poorer and more marginalized members, around industrial processes and technologies, even if large parts of the economy are not involved in industrial production per se and/or become progressively marginalized from the wealth generated through industrial processes. In many respects, integration into modern “industrial civilization” happens through consumption, not production, and is related to underlying processes of monetization (the increasing use of money in the exchange of goods and services, and hence an economy that is increasingly monetary) and commoditization (an increasing share of consumption and production based on buying goods and services through market transactions rather than producing these goods and services oneself or accessing them through non-market forms of exchange such as barter). Similar to Karl Marx, Polanyi (2001[1944]) argued that the feature that distinguishes the modern capitalist era from all others is not the existence of markets, which have always existed, but the commodification of “fictitious commodities,” namely, land, labor, and money, and the creation of markets in these fictitious commodities “as the inevitable consequence of the introduction of the factory system in a commercial society” (78). These elements form the basis of what he called a complex market society. Within this broad frame, a meaningful understanding of modern (capitalist) economic development would be an increasing amount of valueadded9 produced per person, achieved through increasing labor productivity (output per unit of labor time rather than simply people working more or more people working) and sustained by capital accumulation. Capital accumulation refers to the accumulation of produced means of production—for example, machines and also infrastructure—rather than simply an increase of inputs such as labor, land, natural assets, or money.10 At a national level, increasing labor productivity usually results from a synergy of increasing degrees of specialization and scales of production, as well as increasing applications of technology in production processes (including social organization, which might be seen as a social form of technology). A corollary of increasing labor productivity in physical production (food,



Introduction 15

manufacturing, mining, and construction) is that labor eventually shifts proportionately into tertiary (service sector) activities because fewer workers are needed in physical production, whereas more workers are needed to manage the output of physical production and to provide other services in an increasingly specialized and, hence, complex economy. However, productivity is difficult, if not impossible, to measure in service sectors, where output is non-physical by definition. Hence, value-added in service sectors is mostly a reflection of prices rather than output (whereas in physical production, value-added reflects both).11 Labor productivity in this sense is a somewhat problematic concept when applied at an aggregate level to complex modern economies, although it primarily derives its meaning in physical production, particularly in agriculture, where it can be measured in comparable physical terms. The precise combinations of these elements at sub-national levels vary depending on regional patterns of integration into national and global economies. Indeed, some regional economies might possess few or none of the attributes of increasing productivity in agriculture or manufacturing, but benefit nonetheless from a rise in national labor productivity, in particular through the supply of cheap mass-produced goods, the circulation of monetary aggregate demand, and the dissemination of wage norms throughout the national economy supported by government fiscal policy. Furtado (1973) referred to this as modernization without economic development, in the sense that modern goods are inserted into the consumption basket of an economy without the capital accumulation or the adoption of more effective productive processes that are associated with the production of these modern goods. He posited that such new forms of consumption could be afforded by increases of income earned through either the depletion of natural resources or else through static reallocations and specializations of production according to comparative advantage, both of which do not necessarily entail economic development. Along these lines, aid transfers might also be understood as another way to raise incomes and hence modern consumption without economic development. However, from a systemic perspective, it is difficult to know where to draw the regional boundaries of economic development in this respect, that is, the boundaries within which a region can be considered to be developing economically on the basis of sectoral, regional, national, or even global capitalist processes. Notably, there are usually wide swaths of territory and large proportions of population even in advanced industrial countries that are based mostly on the consumption of modern goods rather than their production (increasingly so as employment continues to shift into tertiary activities). In these terms, modern development in China can be said to have definitively taken off from the 1950s onward, following half a century

16

Chapter 1

of volatility in the transition from imperial state to a modern form of statehood rooted in nationalism. Obviously, industrialization and modern state formation had been in course since the nineteenth century, but viewed in terms of structural sociological aspects, the starting point is really located in the 1950s. For instance, life expectancy at birth in China only rose above forty years in the early 1950s, only slightly ahead of India. Similarly, industrialization and broad-based national infrastructure expansion only started to deepen significantly from the 1950s onward, versus previous attempts that would have remained quite insulated from the broader economy in a structural sense, with little impact on nationally aggregated labor productivity. In Tibet, modern development also definitively started in the 1950s, primarily in a demographic sense (as discussed in chapter 3). Despite efforts by the Thirteenth Dalai Lama to modernize the state and some aspects of the socio-economic system in the 1920s and 1930s, he mostly failed (see Goldstein 1989b). Hence, modern development in Tibet did more or less start with the arrival of (Han) Chinese Communists in Tibet, mostly under the guise of the People’s Liberation Army (PLA) in the early years, and bringing with them an entire apparatus of modern statecraft, including modern health care practices as well as modern weaponry and transport technologies that had been previously sidetracked for several decades by war with Japan and then civil war with the Nationalists. Whether equivalent development would have happened without the takeover by Communist China is an academic counterfactual that we will never be able to answer, although very similar socio-economic achievements have been made in Bhutan over a similar time period, albeit Bhutan has also been heavily aided by India. As these are remote and peripheral regions, some degree of integration and dependence on their respective regional superpowers has been inevitable in both cases. From occupation onward, political and economic development in Tibet followed a distorted trajectory based on the forced incorporation of this region into the consolidating Chinese nation-state and the abortion of any substantive sense of local state autonomy (in contrast to Bhutan). The specialized and peripheral structuring of the Tibetan areas according to the broader strategies of national development in China implied that the region never really industrialized outside of certain enclave ventures. It was integrated into the broader national economy, which was industrializing, just not in the Tibetan hinterlands. The region did benefit from the extension of national physical and social infrastructures (to various degrees of depth and quality) and also the extension of a modern politicobureaucratic system of administration, including practices such as Five-Year economic planning. Modern development in Tibet needs to be understood in light of this regional pattern of integration into China.



Introduction 17

With respect to technological and economic aspects of modernity, remote peripheral rural economies generally transform through consumptiondriven integrations into “industrial civilization” without necessarily being involved in industrial modes of production, as discussed above. Such a situation would aptly describe rural Tibet as well as much of sub-Saharan Africa, outside of specific enclave sectors such as mining where industrial processes are definitely in use, although usually under the control and/or ownership of foreign (or non-local) corporations. Ignoring the role of fiscal and other transfers, the relative purchasing power of such a peripheral economy would be determined by the terms of trade between its non-industrial, mostly rural or artisanal output versus the industrial goods (and related services) that are consumed in the economy, mostly through import. The relative wealth that this purchasing power moderates is determined by underlying differences in labor productivity, or what Arthur Lewis (1954) called the “factoral terms of trade” (labor being a factor of production). For instance, even if prices for meat and hides have been increasing in the Tibetan areas relative to other goods such as instant noodles, motorbikes, and mobile phones, the disposable wealth of a herder is nonetheless limited by the number of sheep or yak he or she is able to sell in a year. The number remains limited so long as herding is based on household and non-industrial forms of organization (hence, the reason why the central government was intent for many years on making herders become ranchers as a means of poverty alleviation, until environmental concerns took precedence in the late 1990s). Conversely, the factory worker involved in the production of instant noodles, motorbikes, or mobile phones most likely contributes to the production of much more output and value-added in a year than the herder, particularly in a highly mechanized and automated factory. The worker’s income might not be better than the herder’s, not because of productivity considerations but because his or her wages are squeezed through the process of distributing the surplus (i.e., value-added) between profits, taxes, wages, and other income payments such as interest payments on the debt used to purchase factory machines or rent. Hence, it is not obvious that the herder would become better off by becoming a factory worker (or, along the same logic, a rancher), as is often blithely assumed by scholars and policy makers, although it is quite likely that, in economic terms, the worker produces more value-added than the herder (i.e., the worker has a higher degree of labor productivity), which is then circulated throughout the economy through various distributional mechanisms. Industrialization is thereby usually seen as a significant driver of economic growth whereas “peasant” production is not, even though the “peasant” might not necessarily become better off by becoming a worker within an industrial system of production. As is argued in chapter 7, the delicate balance can lead to many paradoxical dynamics within the labor transitions of such peripheral

18

Chapter 1

economies, that is, paradoxical from the perspective of our understanding of labor transitions from a classical European perspective. In this sense, it is important to break out of simplistic normative binaries regarding development. It is not as if negative attributes (such as persistent poverty) or even a lack of structural transformation in some sub-sections or sub-regions of an economic system (such as the persistence of low-productivity economic activities in rural areas) somehow invalidate the existence or possibility of development, as is sometimes implied in more crude versions of post-development thinking. Rather, from a broader perspective, it is important to consider how such negative attributes are the dialectical parts of a developmental whole, especially with regard to the integration of peripheral regions within wider economies. In particular, the ways that development manifests in peripheral settings can be quite different from the ways it manifests in more central locations. Whether or not such peripheral manifestations can be considered as capitalist by virtue of their association to broader capitalist systems is an arcane academic debate that is not dealt with here (albeit see some references to these debates in chapter 7). Admittedly, this perspective bears resemblance to what is often referred to as a “dependency theory” type of logic.12 More accurately, it draws inspiration from earlier scholarship by Latin American structuralist economists (particularly those of the CEPAL tradition)13 on processes of polarization and marginalization that were occurring alongside relatively rapid economic growth based on substantial degrees of foreign ownership in key sectors of leading Latin American economies in the early post-war era. As noted in the preface, this approach arguably applies quite well to the context of regional development in Tibet, regardless of the contentions that might be raised regarding more general applications to the national case of China (see Fischer 2010 for a suggestion of how it might nonetheless be applied to China). Indeed, Celso Furtado, one of the pioneers of the CEPAL structuralist approach, used his framework to analyze regional development in Brazil, given that his premise of domestic polarization applied both socio-economically across classes and spatially across regions (e.g., see Furtado 1967). The application here is based on the recognition that, despite the presumed exceptionalism of Tibet or else the unique aspects of development in China, development in Tibet is actually quite comparable to other situations of development in remote peripheral regions of consolidating nation-state entities. Accordingly, ownership and control by non-local Chinese corporations and related entities in Tibet would correspond to the idea of transnational corporate ownership in the earlier Latin American structuralist analyses. Some theoretical background is warranted here, given the major intellectual influence of this earlier approach on the research that has contributed to this book. Of particular interest for the present analysis is the work of



Introduction 19

Celso Furtado and Osvaldo Sunkel, which was grounded on the centerperiphery approach of Prebisch (see UN/Prebisch 1950). This approach aimed to elucidate the challenges posed by the historical integration of Latin America into the international economy, including the implications for regional development within countries. As outlined by Prebisch, the propagation of technical progress established the outward-directed, externally propelled development of peripheries and the four characteristics of peripheral capitalist economies: declining terms of trade; marginalization of disadvantaged populations in the peripheries; imitative metropolitan consumption patterns of periphery elites; and macroeconomic instability in the form of inflationary pressures and chronic foreign exchange gaps due to vulnerable economic structures.14 Furtado and Sunkel extended this CEPAL analysis to the patterns of industrialization that were already emerging in Latin America as early as the mid-1950s, including chronic balance of payments problems among the leading industrializers in the region. They criticized the initial optimism of Prebisch, Lewis, and other early development economists that foreign direct investment (FDI) could lead or finance an autonomous process of industrialization. Instead, they noted that the increasing dominance of FDI in key industrial sectors, particularly in Brazil, Chile, and Argentina, undermined national processes of capital accumulation in these countries. While steering clear of the more radical determinism of the neo-Marxist tangents of dependency theory,15 both brought particular attention to the structures of ownership and the resultant control over flows of wealth that ensued from these increasingly transnationalized forms of industrial organization. Furtado elaborated further on the understanding of technological dependence in the periphery. He argued (1973, 20) that “the ability of certain countries to control technical progress and to impose consumption patterns became the decisive factor in the structuring of the productive apparatus of other countries, which in consequence became ‘dependent’” (cited in Kay 2005, 1204). In turn, this dependence leads to dualism in domestic consumption and production structures. Upper strata within these structures are increasingly based on the absorption of technology and consumption patterns imported from the centers, and eventually finance and ownership as well, which reinforces external orientation and dependence. Lower strata remain grounded on more basic technologies and consumption patterns. Although the lower strata supply labor, technology adoption lowers the relative absorption of labor per unit of output before surplus labor is exhausted. Furtado made this last point in contention with the optimism of Arthur Lewis (1954) that industrialization in the peripheries would eventually lead to an exhaustion of unlimited supplies of labor and, as a result, to rising wages among these lower strata.16 Rather, he contended that this structural dualism explains the polarizing effects of peripheral and

20

Chapter 1

dependent growth within the Latin American context, arguing that dependence is reproduced through a combination of internal and external social relations, economic structures, and patterns of ownership. The synergy of internal processes of exploitation with external processes of dependence could, in certain periods, lead to substantial growth (as in Brazil in the late 1960s and 1970s), but this growth would tend to lead to an intensification of internal processes of exploitation, in contrast to the experiences of Europe and North America.17 Osvaldo Sunkel also elaborated the CEPAL structuralist analysis by examining the stratification of dependent economies in Latin America between transnationally linked versus excluded circuits of consumption and production. In Sunkel (1973, 518), he argued that “the massive penetration of foreign firms” accentuated the uneven nature of development in Latin America despite the acceleration of growth rates, especially industrial. Foreign investment in particular had some highly negative effects, including the distortion of local consumption, production, and investment patterns toward servicing the interests of the foreign-owned subsidiaries rather than the interests of a host country. He noted that a definite life cycle is associated with foreign subsidiary activities, with substantial initial contributions of capital, skilled personnel, technology, and management, followed at later stages by cash outflows exceeding inflows and with little educational effect given the policy of retaining a monopoly of skills and technology within the firm (525-27). Beyond these economic effects, he concluded that the socio-political consequences are of even far greater importance, such as the interruption of the formation of a local entrepreneurial class and the absorption of best talents into a new transnational managerial class (or “transnational technocracy”) servicing the foreign interests (527-28). Together with Furtado, he provided important insights into how foreign (or non-local) investment could undermine national (or local) processes of accumulation even in the midst of relatively rapid growth, particularly through its effect on ownership structures and on accentuating various financial mechanisms of wealth transmission from peripheries to centers. These insights of Furtado and Sunkel are relevant for the case of Tibet (and, arguably, to many other peripheral regions within nations) in terms of the extremely high level of external ownership within the local economy of Tibet and how this conditions flows of wealth and social and political economy dynamics within the wider context of regional development in western China. Some modification is required for the context of Tibet to take into account the fact that external ownership has been driven far more by a form of national state capitalism and associated subsidies than was the case in Latin America, where foreign ownership was mostly the result of private corporate investment flows. Modification also needs to be made for the fact that rapid growth in countries like Brazil was nonetheless based



Introduction 21

on intensive industrialization, even if much of it was foreign controlled, whereas rapid growth in Tibet (or the TAR at the very least) has been based mostly on subsidization and largely divorced from productive sectors. As noted previously, the Tibetan case would compare more appropriately to various peripheral regions within Brazil rather than to the national Brazilian case. Nonetheless, the broad relevance of their insights with respect to processes of polarization and marginalization is the same with respect to regional development as it is to international development. Notably, most of these structuralist analyses were built on the contention that peripheral economies tend to face contractionary pressures through processes of national (or international) economic integration, particularly if the tendency for outflows is not compensated for by strong regional transfer payments. In other words, the tendency in modern capitalist systems is for net outflows of labor, skills, and physical and financial resources from peripheral regions. In functional modern nation-states ruled by elites with a modicum of appreciation for redistribution, this tendency has usually been compensated for by regional transfer payments from richer regions to poorer regions, whether in northern Canada, northern Norway, Scotland, Nebraska, Sicily, or elsewhere. If well managed, such peripheral regions can thereby rise with the tide, so to speak, in the sense that national governments can offer the people remaining in these regions an approximation of living standards experienced elsewhere in the nation, albeit with the dialectical consequence of increased dependence on these transfer payments, often displacing more traditional forms of livelihood. In many cases, such dependence comes with various dynamics of paternalism, subordination, marginalization, disempowerment, and even appropriation of control over local resources and/or means of livelihood, such as with respect to mining. Indeed, my own earlier experience with aboriginal communities in Canada taught me to discern these tensions underlying regional transfer payments in a context of racist subordination, and the nefarious effects that they can have on local community development However, when there is little commitment to regional redistribution through the various fiscal means available to national governments, such peripheral regions can fall into economic stagnation or even contraction, particularly when a lack of redistribution is matched with strong dynamics of outflow. One might compare this dilemma to the automatic adjustment mechanism of the pre-war gold standard in Europe, whereby deficit countries would adjust through economic contraction, as famously elucidated by both John Maynard Keynes and Karl Polanyi (cf. Polanyi Levitt 2006). A similar dilemma faces Europe today in its neoliberal attempts to address the structural deficits of Greece and other parts of the Euro-periphery through severe austerity and economic contraction rather than through substantially increased fiscal transfer payments. Indeed, the international

22

Chapter 1

economy might serve as an extreme case of this dilemma, given that overseas development assistance (ODA) from rich to poor countries is paltry at best in comparison to the strong mechanisms of siphoning wealth from peripheries through the channels of international trade and finance (particularly considering that most ODA also probably ends up back in the donor country in any case, either directly or indirectly after a short circuit through the recipient country). This was precisely the point made by Gunnar Myrdal (1957)—one of the seminal pioneers of early development economics— in his theorization of processes of polarization and marginalization that take place as poor peripheral countries integrate into the international economy and how trade leads to greater inequality if the initial distribution of power is unequal. He called the potential impoverishment of these regions a “backwash effect,” which occurs despite the acclaimed benefits of comparative advantage in trade. In effect, the implications are comparable within nation-states, given that national currencies effectively serve as a gold standard (or common currency), imposing contraction on deficit regions in the absence of compensating transfer payments. This was precisely the dilemma of regional development in China in the 1980s, given that regional transfer payments collapsed in the first years of the reform period, leaving the subsidy-dependent western provinces languishing in the midst of a national (and especially coastal) takeoff, including the TAR, which was in a deep macroeconomic recession from the mid-1980s to the mid-1990s.

Integration, Growth, and Marginalization in Tibet Tibet has been much better endowed in this respect than other similarly integrated sub-national peripheral regions of the “developing world” given Beijing’s commitment to regional redistribution, particularly prior to the 1980s and from the mid-1990s onward. However, as mentioned above, this fiscal endowment implies, for better or worse, a much greater degree of dependence and loss of autonomy with regard to decision making and influence over the sources of subsidization located outside the region. The TAR presents an extreme case of disempowerment in terms of not possessing a form of local government run by and for Tibetans. Such a form of local government could have hypothetically taken many forms. It could have been a locally self-governing entity within a larger imperial system, as were most Tibetan regions under the Qing Empire. Or, it could have existed as an independent state, as the region now called the TAR was from 1913 to 1950. Or, it could have taken the form of a local government ruling with a certain degree of autonomy over local affairs within a larger nation-state entity, as was the case in the region now called the TAR under the 17-Point Agreement from 1951 to 1959



Introduction 23

or else to a very limited extent during the brief moments of opening that took place in the early reform period in the 1980s. This last form is the model suggested by the idea of minority nationality autonomy in China, although it was certainly not practiced in the TAR following the escape of the Dalai Lama in 1959 and throughout the Maoist period, despite the name given to the province in 1965. Since 1989, the notion of minority nationality autonomy has been practiced in very tenuous ways, especially in the politically sensitive regions of Tibet and East Turkestan (Xinjiang Uyghur Autonomous Region), where the broad scope of decision making has been, for the most part, deliberated and dictated by non-Tibetan outsiders, especially with respect to economic policy. Within these ebbs and flows of center-local relations, there has always been a margin of local control and influence for local Tibetan cadres, who have been able to tinker with implementation, explore wiggle room, subvert and even profit from local positions of influence, or else practice discrete subaltern forms of resistance. However, the overarching reality is that the policies that guide development in the TAR and other Tibetan areas have been essentially promulgated from Beijing as top-down dictate, following the trends of national development policy. Policies are then, effectively or ineffectively, implemented by local authorities, themselves appointed by Beijing in the case of the TAR or by provincial authorities in the case of Tibetan areas outside the TAR, with the assistance (in the case of the TAR) of a corps of professionals and cadres from around the country on terms of duty that usually last two to three years. Notably, no ethnic Tibetan has ever held the post of party secretary of the TAR, the TAR’s top leadership position. Ethnic Tibetans did make up about 70 percent of cadres in the TAR up to 2002, although this share subsequently fell sharply to less than 50 percent in 2003 (and the number employed also fell), precisely as the OWC was entering into full steam with the railway construction from Qinghai to Lhasa. The national leadership changed in 2002 and 2003 (recall that Hu Jintao was party secretary of the TAR in 1989, when martial law was declared in this province). As noted previously, these data on the ethnic composition of cadre and more general government employment in the TAR were no longer reported after TSY (2004), although there are strong indications that the shift away from Tibetan representation was maintained, if not worsened, in the subsequent years following 2003, especially following the protests in 2008, as discussed in chapters 5 and 6. In any case, due to the massive fiscal weight of Beijing and the political and security paranoia that grips the TAR, the remaining Tibetan cadres mostly toe the line set out from above and they are in turn rewarded for their obedience with some of the highest salaries in China. A meaningful decentralization of political and economic decision making has simply not taken place in the TAR the way it has in most other provinces of China.

24

Chapter 1

The fact remains that local initiatives and locally generated investment and accumulation play a minute role in the overall processes of economic change in the Tibetan areas. In the tense political environment, in many cases they may have even been discouraged. The TAR economy has been changing rapidly, but the local Tibetan population has been rendered very marginal as agents causing the change at the aggregate level, even if they reap some benefits. Their contribution to the indigenous village-based economy is huge, no doubt, but this economy has shrunk rapidly as a source of value relative to the rest of the economy, and much of the surging activity within the rural economy is subsidized by the government in any case—for instance, most of the entrepreneurial activities researched by Childs et al. (2011) were the result of subsidized rural development policies or else responses to opportunities related to other subsidized activities, such as construction projects. In this sense, the agency and “ownership” of development (if we are to use these words)18 is located largely outside of Tibetan hands and this situation has been accentuated as the economy becomes progressively less agrarian and rural. The resulting economic structure in Tibet—including the broader political economy structure of entitlements, incentives, compulsions, and distributive conflicts—is effectively very similar to that of a colonial-type economy. Indeed, the degree of aid dependence in the TAR is far greater than even the most aid-dependent countries of Africa, and the degree of disempowerment is more or less on par with that of an occupied region. Whether or not we formally define Tibet as a de jure “occupied” territory, many of its developmental dynamics function as if it were occupied—hence my use of the term “effective occupation.” The closest comparisons at the international level (if we are to accept Tibet as occupied, rather than simply a region within a nation-state) would be the Israeli occupation of Palestine or the US occupation of Iraq. Obviously, China is not waging a military campaign in Tibet the way the United States is waging one in Iraq or Afghanistan (or Israel in Palestine) although, as was often pointed out to me by many Tibetans, China need not do so precisely because of the totality of its control in Tibet (in contrast to the 1950s and early 1960s, when China was required to wage a military campaign in order to subjugate many Tibetan areas, particularly in Kham and Amdo). Similarly, it might also be argued that the US-backed government in Iraq or the severely constrained government in Palestine possesses considerably more autonomy of local rule than the symbolic Tibetan political leaders possess in the TAR. Alternatively, comparisons might be made to the relatively autonomous soviets under the Soviet Union (Kazakhstan, etc.), which subsequently became independent following the demise of the Soviet Union. In those cases the economic structures of integration and dependence were similar to those of the TAR, except that local nationality leaders (Kazakhs, etc.)



Introduction 25

still managed to retain considerable power and autonomy within their soviets. As a result, independence was a relatively seamless political transition, without much significant change in leadership or local state structure, even though it led to considerable economic upheaval and turmoil due to the disintegration of a tightly integrated regional economy coordinated by Moscow. In comparison, autonomous regions in China—at least the TAR and the Xinjiang—have been much more assimilationist, to the extent that senior local leadership is dominated mostly by outsiders, as noted above. For most ostensible purposes, “autonomous region” in the case of the TAR would be best defined as “being ruled autonomously from Beijing,” that is, being ruled more or less under central command, unlike most other provinces of China. The fact that all other Tibetan areas in China were explicitly included in the Fifth Tibet Work Forum in 2010 (for the first time in the history of Tibet Work Forums) might be suggestive of a policy decision by the central government to start treating other Tibetan autonomous areas in a similar manner, which could be viewed as ominous and/or as an opportunity to start coordinating Tibet policy in a regional manner, rather than as it currently is—fragmented across five separate provinces. In this respect, the case of Tibet actually helps to advance the theoretical tension that Celso Furtado himself struggled with in his own theorization of growth in Brazil. This concerns the tension between his stagnationist predictions and the subsequent strong economic performance of the Brazilian economy in the late 1960s and 1970s, which led him to revise his thesis (as in Furtado 1973a). Concern with the co-existence of marginalization and rapid growth was subsequently eclipsed by the agenda of development studies after the 1970s given that protracted economic crises in Latin America and Africa gave priority to concerns about polarization and marginalization that were occurring in a context of austerity and recession, and also because of the denigration of structuralism by mainstream scholars—especially economists—alongside the rise of neoliberalism in the 1980s. Nonetheless, the experience of China since the late 1970s, and more recently that of India, has resurrected issues of equity within growth (versus equity within stabilization and structural adjustment), and the manner by which subordinated peoples and peripheral regions are integrated into national development strategies. The challenge of explaining the co-existence of marginalization and rapid economic growth in Tibet is full of potential precisely because both have evolved in synergy as two aspects of strong regional integration into the national economy. The implication, as noted previously, is that marginalization, subordination, or exclusion are not necessarily associated with absolute or even relative indicators of deprivation (even though they are typically treated in this way; see Fischer 2011a). Rather, they represent processes occurring throughout social hierarchies and across all social strata,

26

Chapter 1

not merely at the lowest and poorest social strata where deprivations are evident, but also among various middle and upper strata. Moreover, these processes need to be identified relative to the rapidly changing norms and contexts of social stratification within growth and development. In this sense, the Tibetan case begs us to return to an intellectual agenda of attempting to understand the structural and institutional dynamics of marginalization within a context of growth and improving living standards.

On Discrimination Closely related to the issue of marginalization is the question of discrimination. Indeed, my research started with the question of whether persistent poverty among Tibetans19 or interethnic inequalities between Tibetans and Han Chinese have been the result of deliberate discrimination through either formal policy or informal practices, or whether they have arisen from other variables that are related to the presence of Tibetans but that are not intentionally discriminatory and that may or may not be specific to their regions. This question was intended partly to address the arguments of Sautman and Eng (2001), as discussed above, as well as the types of arguments made by many Chinese scholars. The other variables in question could include issues specific to Tibetans, such as the spatial remoteness and resource poverty of many Tibetan areas, their cultural and demographic characteristics, or the unfortunate consequence of being located in the wrong place or wrong sectors. Variables could also include issues that apply throughout China, analyzed in terms of how each influences the particularities of various Tibetan areas, such as aspects of uneven development, fiscal and pricing policies, urban-rural inequalities and their exacerbation in the West, the impact of liberalization on peripheries, or the political economy of local state governance. The main challenge of this line of inquiry is to differentiate discrimination on two levels. The first is to differentiate discrimination from the other variables, such as remoteness or uneven development. The second is to differentiate practices of discrimination that are specific to Tibetans from those that occur generally throughout China (such as the hukou system of household registration). Discrimination is definitely evident in the field, to the extent that I have often felt that we need to start using concepts such as prejudice and even racism in our analyses of Tibet (although the conception of race is problematic here, as discussed below). The closest we can approach the issue in terms that are somewhat acceptable to the government is with reference to “Han chauvinism.” As discussed by Yeh (2007) and Childs (2008), prejudice can be in a soft form, whereby Tibetans are characterized in romanticized terms as cheerful and noble albeit somewhat naïve and in need of the



Introduction 27

wardship of the Chinese, or else in a hard form, whereby Tibetans are characterized in negative terms as backward, unruly, and even lazy barbarians in need of civilization. Neither of these two attitudes is contradictory with the degree of subsidization that the central government is providing to this region. Moreover, Tibetan perceptions of being discriminated against were consistently confirmed to me during fieldwork in Tibet, even by Tibetan government officials. We must nevertheless ask to what degree such prejudices shape the adverse structural and institutional trends of development in the Tibetan areas. Certainly, not all dimensions of the developmental challenges in Tibet can be associated with discrimination. Some of these challenges are common to all peripheral regions within modern capitalist systems, as discussed above.20 The study of discrimination in Tibet in this aggregate and evaluative manner is problematic for both practical and conceptual reasons. Regarding the practical, very little data is available to evaluate these questions. Statistical tests that are common in other parts of the world, such as returns on education in order to see if people with equivalent levels of education tend to experience different wage and employment outcomes based on their ethnicity, race, gender, or other innate characteristics, are simply not possible in Tibet. Controlled surveys are mostly not permitted, certainly not for these types of questions, and official Chinese statistical sources mostly do not divulge disaggregated data by nationality (even though this would be available in principle), besides in the population census data (which is very useful). Creative extrapolation can be used in order to circumvent these limitations and to tease out insights from the available official data, in particular by focusing on the TAR given we know that its rural areas are almost entirely populated by Tibetans (unlike the other four provinces containing Tibetan areas), although this offers only suggestive insights. Qualitative field observations remain the only direct way of evaluating of discrimination. Hence, claims of representativity or objectivity can always be contested by the government (or its associated and supportive scholars) by virtue of its privilege to withhold the information required to be able to test such claims. Conceptually, we might also question how to conceive discrimination or prejudice in the Chinese context. Han Chinese people in China do not necessarily discriminate against fairly similar minorities on racial grounds, such as formal racial discrimination in former Apartheid South Africa, or informal racial discrimination in parts of Latin America or the United States. In other words, if a Tibetan is fully sinicized, with an equivalent education to a Han Chinese person, it is unlikely that he or she will be discriminated against on the identity basis of being Tibetan (as argued by Sautman 2012).21 Rather, Han Chinese people in China generally appear to discriminate on cultural or linguistic grounds (or political/ideological

28

Chapter 1

grounds if the patriotism of a minority group is generally mistrusted). For instance, ideological loyalty might be distrusted by virtue of being Tibetan. This does not mean, however, that these alternative modes of bias wield any less potency. In the medium or generational term, they can also be quite innate. Cultural and linguistic modes of bias are also very powerful in labor markets, as discussed at some length in this book. These modes of bias also carry a strong assimilating force, which is arguably at the root of contestation in Tibet and East Turkestan (Xinjiang) and belies the procedural equality that is supposed to exist in theory between all citizens of the PRC. Hence, discerning discrimination in the Tibetan areas requires a more subtle analysis of these modes of bias rather than simply testing to see whether Tibetans receive worse outcomes on the basis of their ethnic status for equivalent educational or other qualifications. Moreover, as mentioned above with respect to marginalization, discrimination might be poorly reflected by data on relative deprivation and inequality in any case. This is also an underlying problem with the way that social exclusion is generally conceived and operationalized in the social policy or development studies literatures (e.g., see Fischer 2008a; 2009a; 2011a). It is not even clear whether Tibetans are necessarily the poorest in China, although they definitely have among the worst levels of formal schooling and among the worst mortality rates and longevity. As discussed in chapter 7, despite their apparent income poverty, rural Tibetans nonetheless seem to have higher wage and non-wage work expectations than Han or Muslim migrants, and also display more resistance to moving into stigmatized occupations such a manual wage labor or petty service jobs, especially in the case of drokpa (i.e., pastoralists, herders, or nomads). Moreover, while relative deprivation and inequality measures do show evidence of the hurdles faced by the poorer strata of a society to access the norms operating within an economy, much of the cultural, linguistic, and political biases experienced by Tibetans are faced not only by the poor, but also by higher social strata. Some of the strongest exclusionary pressures in Tibet are arguably observed at the middle and upper strata of the local labor hierarchy, such as among high school and university graduates competing for state-sector employment or else among more affluent sections of the rural population migrating to urban areas. In these cases, we would expect the strongest resistance to government policies to emerge from these middle and upper strata, even though their tangible deprivations might not qualify them as absolutely or even relatively poor. It is also in this sense that rising prosperity cannot be taken as a refutation of discrimination, as is typically proposed by the government and associated scholars. The more pressing question is whether the integration of Tibetans into the emerging structural and institutional patterns of development in Tibet has accentuated their disempowerment in the governance of their home-



Introduction 29

land and in the “ownership” of their development, whether or not this necessarily results in some form of deprivation (unless discrimination and disadvantage are themselves tautologically defined deprivations). I labeled this dilemma as “exclusionary growth” in Fischer (2005a), although this label generated some confusion given the standard ways of understanding social exclusion as poverty. In this book, I return to the more general concept of marginalization as a broader frame within which we can analyze processes of both inclusion and exclusion, and how these work together in a dialectical manner, in particular settings and throughout social hierarchies, to produce subordinated and disadvantaged modes of integration and disempowering forms of development for certain people. Intentionality aside, these processes can be seen to effectively discriminate against subordinated or disadvantaged groups if their negative tendencies are left unmitigated. It is perhaps less urgent to identify precise discriminatory conduct and its perpetuators, as would be the focus of much human rights work,22 rather than to understand more broadly the ways that development strategies designed in Beijing have effectively resulted in discriminatory outcomes toward Tibetans in Tibet, regardless of whether such discrimination was necessarily intended in the design or application of these strategies.

The Economics of Marginalization in Tibet In sum, within a context of continued political disempowerment of Tibetan locals, centrally directed development strategies since the mid-1990s have channeled massive amounts of subsidies and subsidized investments (relative to the local economy) through Han Chinese–dominated state structures, corporations, and other entities based outside the Tibetan areas, thereby accentuating the already highly externalized orientation of wealth flows in the local economy. This has resulted in a socio-economic structure that increasingly and disproportionately rewards a small upper stratum servicing and/or operationalizing the development strategies, which have remained excessively hinged on decision making in Beijing or other regional (non-Tibetan) political centers (namely, Chengdu, Xining, Lanzhou, and Kunming) to a far greater extent than any other region in China. The upper stratum includes a small minority of Tibetans and a large proportion of non-Tibetan migrants, concentrated mostly in urban areas and well positioned to access the flows of wealth as they pass through the region with increasing velocity. Whether or not the ongoing outcomes are intended to be discriminatory, these structural and institutional dynamics effectively accentuate the discriminatory, assimilationist, and disempowering characteristics of development, even while producing considerable improvements in the material consumption of local Tibetans. The long-

30

Chapter 1

term consequences of the resultant economic structures are problematic considering the rapid shift of the Tibetan workforce out of agriculture and toward the highly subsidy-dependent sectors of the economy, especially in urban areas. Subsidization strategies thus result in a dynamic that might be called “boomerang aid” (Fischer 2009b), for which the TAR represents a quintessential example. Subsidies largely return to their sender while debilitating indigenously oriented forms of wealth creation and accumulation in most sectors outside the rural economy, even while they have nonetheless stimulated rapid transitions out of agriculture and the rural economy, again, at a faster pace than observed nationally or in other western provinces. This reinforces a situation of extreme economic inefficiency and dependence, to the extent that an increasing intensity of subsidization over this period— well above the high level that was already established since the late 1960s— has been required in the TAR in order to sustain the rate of economic growth. The increasing intensity presumably has an upper limit, especially given that direct budgetary subsidies from Beijing to the TAR government surpassed 100 percent of GDP in 2010 for the first time in the history of the TAR and investment in fixed assets (most of them probably subsidized) also surpassed 90 percent of GDP in the same year. These heights were partly reflective of the national response to the global financial crisis in 2007-2009, which involved a further invigoration of western development and ever higher levels of investment in China as a whole. However, the continued extreme exceptionality of the TAR also indicates that the government’s response to the 2008 protests was a strategy of much more of the same subsidization strategies they had been following prior to the protests, including an intensification of large infrastructure projects heavily based on imported inputs from elsewhere in China or abroad. Indeed, the intensive import dependence of growth in the TAR explains much of the inefficiency given that a large portion of subsidies immediately leaves the TAR via the trade account, or never enters but exists only as accounting entries on the balance sheets of corporations involved in these large projects. The situation accentuates strong cultural, linguistic, and political biases stemming from characteristics of the dominant cultural and political group, such as Chinese fluency, Chinese work cultures, and connections to government or business networks in China proper. In turn, this intensifies the ethnically exclusionary dynamics of urban employment and marginalization among the middle and upper strata of the Tibetan labor force. These biases and dynamics are most evident in the context of competition over urban employment opportunities between local Tibetans and non-Tibetan migrants from other parts of China, and are exacerbated by an open migration policy, the introduction of competitive labor market reforms, a corresponding weakened protection of local Tibetan employment despite severe



Introduction 31

educational disadvantages, and a broadly assimilationist social agenda. While labor market reforms have also generated problems elsewhere in China, such as high rates of unemployment among university graduates,23 the Tibetan areas manifest a unique structural asymmetry whereby the most educated category of local residents (urban Tibetan men) possess a much lower level of educational attainment on average than even the least educated category of interprovincial migrants competing in local urban labor markets (e.g., rural women from Sichuan). This asymmetry simply does not exist in any other region of China or among any other minority or majority nationality. Various Muslim groups in Qinghai and Gansu with equally poor levels of schooling reside in locations where outmigration is much stronger than in-migration and where local county governments have been permitted much more activist roles in protecting and promoting local employment than typically permitted in most Tibetan counties. Moreover, these intergroup comparisons do not reflect the disadvantages that even educated Tibetans face in competition for the types of employment that would correspond to their expectations relative to their respective levels of educational attainment. In other words, on top of facing an institutional environment where opportunities created in the local economy are dominated by state-centered and externally oriented networks of power and wealth, Tibetan urban “middle classes” and elites are also required to contend with the fact that they do not profit from natural educational advantages over the typical (usually Han) migrant coming from elsewhere in China, contrary to all other regions of China where urban residents usually have higher educational attainments than rural migrants and thus can legitimately dominate the more privileged strata of urban employment opportunities. Notably, this particular aspect of marginalization in Tibet pertains specifically to middle and upper social strata and would not necessarily be reflected by inequality measures, which generally reflect only the marginalizing dynamics occurring among poorer social strata. While the latter are important, the former provide particular insights into the grievances underlying recent tensions in Tibet. Indeed, grievances would be particularly rife among Tibetans in middle and upper social strata because they potentially face the greatest relative downward displacements through the process of marginalization. Tibetans in these social strata are also more likely to mobilize politically. In contrast, the relative position of Tibetans in lower strata would more likely remain unchanged through these processes (i.e., if they remain as farmers or herders) or even slightly improved (through increased income earning opportunities in low-skilled work), and such Tibetans would also tend to be less politically engaged. The combination of these disempowering dynamics of marginalization with the sheer speed of dislocating and disembedding social change across

32

Chapter 1

the Tibetan areas provides important insights into recent tensions given that it has accentuated insecurity and protective impulses within communities while at the same time restricting an ability to mitigate against the adversities generated by such rapid change. While direct attribution is often difficult to trace in many cases, these combined factors can be considered as fundamental underlying determinants that condition the social environment in which patterns of contestation and conflict emerge and manifest, whether or not the expressions of contestation necessarily involve explicit or direct links to these underlying factors. For instance, as I have suggested (Fischer 2012a), resettlements in eastern Tibet could well have served as an underlying conditioning factor contributing to the wave of self-immolations that started in this region in 2011. Even though none of the self-immolators had made an explicit connection of their acts to the resettlements up to that point (albeit almost none had left behind testaments or explanations of their act either), it is nonetheless clear that the resettlements (and other policies such as schooling reforms) have been profoundly disruptive to local communities in this region and would have provided the underlying conditions for other more direct and proximate causes (such as patriotic education campaigns) to drive people to such extreme acts of protest. This understanding of contestation and conflict is influenced by Karl Polanyi’s conception of “double movement” as the impulse for social protection that emerges in response to the dislocating and disembedding effects of the extension of “self-regulating” market forms of social organization (Polanyi 2001 [1944]). “Disembedding” in this sense refers to the capitalist subordination of other forms of social organization to the demands of economic (primarily market) forms of organization, in contrast to all previous social systems prior to the advent of capitalism in which economic activities were generally subordinated to other forms of social organization. As opposed to Marxist thought, Polanyi argued that the double movement in response to such disembedding did not carry any deterministic class logic but was an impulse that cut across class, involving business groups as well as workers’ movements. The double movement could therefore result in reactionary, reformist, or revolutionary protective responses, and Polanyi used this analysis to explain the emergence of Fascist, welfare, and Socialist states in the twentieth century. The case of Tibet is not exactly one of disembedding as understood in these Polanyian terms—in the sense of an expansion of self-regulatory market forms of social organization—although the framework is analogous. There has been a definite process since the mid-1990s of a rapid dislocation of the social structure from its traditional agrarian roots that had more or less survived both Maoism and the early reform period. This dislocation has been complemented by an increasing monetization of the economy and the rise of urban economies as explosive growth poles, even



Introduction 33

though these changes have been largely instituted through state rather than market-driven mechanisms. Despite the state-market distinction, disembedding and dislocation refer most fundamentally to the rapid transition of the local (mostly Tibetan) labor force out of the primary sector (mostly farming and herding).24 Again, these changes have been more rapid than changes occurring elsewhere in China albeit without the relative autonomy that local people and governments in other regions of China can rely on to mediate the consequences. In the TAR, for instance, the share of the local labor force considered as employed in the primary sector dropped from 74 percent in 2000 (the most agrarian labor force in China at the time) to 53 percent by 2010—a reduction of twenty-one percentage points in a decade. The speed of such transitions in the Tibetan areas outside the TAR might well be even faster given the implementation of large-scale resettlement schemes in pastoral areas (which have largely bypassed the TAR to date) and the closer integration of these areas into neighboring Han Chinese urban centers. The degree to which such labor transitions represent disembedding (per a Polanyian understanding) could be debated and is related to the degree to which they are reversible. Some of the decline in the Tibetan primary labor share probably reflects migratory workers who are still fairly well embedded in the rural economies from which they seasonally emigrate for part of the year in search of off-farm employment. In this case, the official data probably exaggerate the degree to which the local labor force has become disembedded from the rural economy, implying that these labor transitions could be reversible if urban employment opportunities were to become more austere, in the sense that these migrants could easily return to farming or herding. Nonetheless, such speculations do not necessarily lessen the sense of rapidity that the official data reflect regardless of their precise accuracy given that similar migratory considerations also apply in other parts of western China. Conversely, from a global demographic perspective, we can expect that, once started, these transitions will probably continue, in the broad structural sense that populations rarely move back into farming or herding once they have moved out of these activities (short of some massive traumatic event).25 Indeed, the migratory employment patterns discussed above are fairly typical in early to mid-stages of urbanization. Hence, the structural shifts observed in the employment data plausibly represent the unleashing of profound social transformations that, once started, are unlikely to reverse—even considering the rural embeddedness of migratory labor or else a hypothetical prospect of dire economic conditions in the urban areas. These transformations will obviously not spell the death of farming and herding in Tibet, but they will undoubtedly change the nature of farming and herding within the broader socio-economic system.

34

Chapter 1

The degree to which these changes have been experienced by rural Tibetans as positive or negative is also under debate. For instance, Goldstein et al. (2008; 2010) and Childs et al. (2011) have portrayed these changes as largely positive, based on the improved livelihoods of villagers in the three farming villages near Shigatse in the TAR that they surveyed. These improved livelihoods were based on the forms of labor migration described above and on emerging forms of local entrepreneurship, both of which reflected a shift away from more “subsistence” modes of production toward a new attitude of “going for income.” My own views have been more influenced by the situation outside of the TAR in eastern Tibet, where labor transitions have been much more associated with the appropriation—partial or whole—of land tenure by the state through resettlement policies, particularly in pastoral areas, or else limitations in the use of land more generally, hence undermining land-based livelihoods in more coerced ways than observed in the TAR. The rural labor transitions in the TAR, as described by the team of authors cited above, were not enforced by or did not otherwise entail any significant erosion of the land tenure of these rural dwellers. Rather, they were arguably induced by the deluge of subsidies flowing into the region, which created abundant and lucrative short-term off-farm opportunities, particularly in construction. The difference might explain the more explosive and protracted nature of protests currently occurring in eastern Tibet (at the time of writing) than in the TAR, although both regions nonetheless burst into widespread protests in 2008. Indeed, the commonalities are probably better identified in the exclusionary nature of urban employment dynamics in both regions rather than in the respective patterns of agrarian transition, which in either case has increasingly channeled rural dwellers to towns and cities, and thus to greater exposure to urban-centered exclusions. More generally, the speed and largely disempowered dynamics of these changes in both contexts have been deeply unsettling for the populations involved. To the extent that many of these socio-economic changes might be irreversible, they highlight a variety of concerns particular to the disempowered circumstances of Tibetan areas and to the role of government policies in mediating the pace and character of change. A major concern regards the dependence on massive levels of subsidization (relative to the local economy) that have been driving economic growth and structural change in Tibetan areas and on which many Tibetans have increasingly come to rely through the course of these labor transitions, despite the absence of sustainable productive economic foundations to support these changes as elsewhere in China. To the extent that urbanization becomes increasingly central to these changing employment patterns, the continuing, if not strengthening, dominance of Han Chinese in the urban economies of Tibet and the associated urban exclusionary pressures faced by Tibetans



Introduction 35

also become increasingly contentious. Similarly, the heightened state of disempowerment faced by Tibetans in the governance of their regions leaves them with little capability (relative to populations in other regions in China) to mediate these changes politically vis-à-vis the dominant sources of power determining subsidies and related regional development policies. One aspect of the dislocation resulting from these transitions can be characterized as the disruption or even reversal of traditional social and wealth hierarchies in the Tibetan areas, in particular, the emergence of Chinese Muslims in Tibetan areas (particularly Amdo) from a lower subordinated status in the pre-modern period to an increasing degree of influence through commercial success. In this light, recent Tibetan-Muslim conflicts can be understood as emerging from the impulse to increase stability and predictability within these dislocations, as Tibetans (particularly Tibetan elites) seek to innovate or negotiate alternative forms of integration into the changing matrices of wealth and power in China. In other words, many of the protests emerging in the Tibetan areas over this period can be understood as strategies to mediate the dislocation and disruption of hierarchical orderings and to reconstitute these orderings along terms that are deemed advantageous or dignified within a new socio-economic setting. The case for this interpretation is made in the final chapters of this book through a qualitative analysis along two themes: resistance to demotion in labor hierarchies and interethnic conflict or collaboration rooted in the legitimating authority of Tibetan Buddhism. There are obviously other forms of strategic protective impulse, such as political protest against the state or the porous population movements back and forth into the exile community in South Asia. However, the two chosen here serve to highlight indigenous strategies of pushing the boundaries of the possible and the permitted within the context of effective occupation in order to reconstitute hierarchical social orderings in their favor. The observations from this line of analysis are important because they imply that two critical dimensions of current Tibet policy need to be seriously addressed in order to lessen both the degree of marginalization faced by Tibetans in the context of contemporary development and the related social instability and political protest. On one hand, urban employment for locals requires protection and promotion at both lower and upper strata, ideally on the basis of more enhanced systems of locally autonomous governance. On the other hand, the linguistic and cultural disadvantages faced by Tibetans in urban employment need to be lessened, albeit in ways that do not undermine Tibetan language and culture. Both dimensions imply a shift back toward an understanding of minority autonomy as it was practiced in the early reform period, particularly in the 1980s. Indeed, the existing national minority laws in China—which formally survived from this earlier period, albeit not without serious challenges and modifications—already

36

Chapter 1

provide ways to resolve both issues if their full implications were to be put into practice, as exemplified by the recommendation by the Tenth Panchen Lama that public-sector employees working in Tibetan areas should have at least some working knowledge of Tibetan, supported by a strong promotion of bilingual education in the minority areas. The challenge, of course, is how to advance such policies that flourished at a time of relative fiscal neglect—particularly in Tibetan areas outside the TAR—now that the local economies of the Tibetan areas have been powerfully restructured toward outside interests and ownership. It is in this sense that the structural and institutional inertia of current development in Tibet provides little hope for a resolution of current contestations by Tibetans, at least not short of a strong and determined intention by the central government to re-rig its regional strategies on a footing that is much more respectful toward substantive practices of local autonomy.

Methodology The methodology of this study has been thoroughly interdisciplinary. The approach zooms from a macro quantitative perspective on population and economic transformations into a series of qualitative field studies. Both foci are backed up by extensive reference to secondary sources from disciplines ranging from history and religious studies to economics. In addition to my conviction that interdisciplinary and multi-method approaches are indispensable to a systemic understanding of social phenomena, this iteration of quantitative and qualitative research methods has been vital in the case of researching politically sensitive development issues in Tibet given limited access to data and the difficulties of conducting fieldwork in a tightly controlled setting characterized by repressed political dissent. In terms of both the quantitative and qualitative foci, the study includes most of the Tibetan areas in China, with particular focus on those of the TAR and Qinghai, as a means to highlight the broad patterns that are common and shared by Tibetans as a group and that function within a conception of Tibet as a cultural and social regional system.26 Although this approach sacrifices specificity for generality, the broad coverage aims to provide a macro framework for understanding the systemic evolution of the region, thereby complementing the growing body of location or sector-focused research in Tibet studies. Indeed, while the latter are used extensively here as secondary sources, a systemic approach offers insights into patterns that are otherwise not ascertainable from these more focused studies. It should be noted, however, that this systemic approach has also been formidable given that, when I started my research in 2001, there were few, if any, satisfactory macro-structural economic analyses of the Tibetan areas



Introduction 37

in China available in the Western scholarship on Tibet. There has been a tradition of such macro-structural research in China, as exemplified by the classic study by Wang and Bai (1991 [1986]), which is assessed in chapters 4 and 7; by Ma (1996; 1998); by several other books published in the 1990s and reviewed by Yan (2000); and more recently by Jin (2010). Many of these studies offer useful critiques of central government aid projects and subsidies in Tibet. However, with the exception of Ma (1996; 1998), they are mostly made from a dominant Beijing-centric perspective, in the sense of conceiving development in Tibet in terms of the necessity of integration into national development strategies rather than in terms of local indigenous dynamics and priorities. Perhaps due to political constraints or problems of reflexivity on such sensitive and highly charged issues, most of these Chinese authors tend to lack a more critical dimension of macro-structural analysis with respect to the disempowered and subordinated manner by which Tibetans have been integrated into the national economy. They are generally deficient with respect to Tibetan-centric perspectives on Tibetan experiences with development, whether from micro or macro perspectives.27 Otherwise, rigorous macroeconomic analysis has been lacking in most studies on the development of Tibet in the Western literature, and even these remain restricted to the TAR, ignoring the wider Tibetan region.28 My first book (Fischer 2005a) broke the trend, although it was only a preliminary analysis in many respects and only dealt with data up to 2001 (with some exceptions), thereby missing most of the subsequent impacts of the OWC. The method used in the quantitative part of this research derives from a structuralist development economics approach, focusing inductively on the evolution of aggregates, averages, structures, and compositions, rather than on statistical associations across individual and/or household characteristics as is typical in the more deductive methods of regression analysis. Quantitative data were obtained mostly from publicly available official statistical sources, such as the 2000 census tabulations, the China Population Statistics Yearbook from various years, and the China (and various provincial) Statistical Yearbook from various years. While the precise accuracy of these official sources is often doubted, the alternative of conducting independent household surveys was impossible given the political situation and would not have provided sufficient macroeconomic data in any case. However, in defense of these data sources, I have generally found that they corroborate with fieldwork in terms of representing broad structural trends over time. The issue of precise accuracy has not been an overwhelming obstacle for the inductive nature of quantitative analysis required for this study given that the accuracy of these data is arguably sufficient to tease out broad structural trends, obviously keeping in mind that all social statistical work must be approached interpretatively with an understanding of the institutional relativity of all statistics, whether in definition or in

38

Chapter 1

collection (for more discussion, see Fischer 2005a, 6-12). Moreover, the official statistics are all that we have to understand the broad nature of social and economic change in Tibet and thus it is urgent to exploit them as best we can. Indeed, the relevance of regression analysis is questionable without first establishing an understanding of the changing structural context from which various panel data are taken and to which they implicitly refer. This is the advantage of adopting an inductive structuralist economic method, particularly within the context of the Tibetan areas in China where limited and changing data availability, and often obscured data reporting, requires much interpretative creativity in quantitative analysis. However, this structuralist approach is not used to suggest a structurally deterministic understanding of the transitions studied (as with structuralism in sociology, for instance), nor a homogeneous experience among the individuals and groups represented by aggregates and averages. To the contrary, the sheer preponderance of subsidies in the Tibetan case points to the overwhelming role of institutional factors in influencing structural changes—particularly government spending in creating strong expansions in higher-value construction and urban tertiary sector activities, or else policy-enforced urbanizations via resettlements and so forth. Nonetheless, aggregate structural analysis, rooted in an institutionalist understanding of context, serves as a useful method for reflecting on the factors and forces shaping the broad socio-economic norms of people’s lives, within which people make decisions and act in a wide variety of ways. Institutionalist approaches in social sciences often refer to this as the sets of incentives and compulsions guiding individual behavior, or else (per the terminology often used in demography) the opportunities and constraints within which individuals exercise their choice and/or agency.29 Unfortunately, much of contemporary social statistical analysis (especially in mainstream economics) is mostly concerned with the range of individual choices and outcomes within a sample due to an obsession with applying increasingly sophisticated yet restrictive regression techniques to panel data. As a result, the (structural and institutional) forest is usually missed for the trees, even though the broad structural trends might in fact provide much more insight into the individual outcomes than the abstracted associations observed across the characteristics of these individuals. Hence, the exercise here is to flesh out the structural trends in Tibet as a means to reflect on the profound changes that are occurring to people’s lives and livelihoods in very real and rapid ways, and which are quickly transforming the landscape faced by the next generation of Tibetans. Accordingly, socio-economic structural transformations have been analyzed through a descriptive and comparative trend analysis of aggregate population, national accounting, employment and wage, and household



Introduction 39

survey data, comparing the TAR to several other provincial cases in western China and the national average. The inter-provincial comparison sheds light on the exceptional speed and characteristics of the transitions that have been induced in the Tibetan areas within a very short period of time, even after taking into account the very different socio-economic structure of the Tibetan areas. In much of this aggregate analysis, the TAR has been chosen as the basis of comparison because this province-level administrative entity is entirely composed of Tibetan areas and thus the aggregate structural data of the TAR—much of which is available only at a provincial level of disaggregation—represents the experience of an entirely Tibetan area in China. In contrast, the data (including rural data) for the other provinces containing Tibetan areas (Qinghai, Sichuan, Gansu, and Yunnan, accounting for over half of the Tibetan population in China) is overwhelmed by the population weight of non-Tibetans in the non-Tibetan areas of these provinces. Moreover, the data from the TAR arguably approximate broad trends in Tibetan rural areas outside the TAR given strong similarities across these Tibetan areas and stark differences between these areas and everywhere else in China in terms of topography, population density, patterns of land use and livelihood, levels and composition of rural household incomes, education levels, and health indices. This book focuses on the actual results of policies rather than the discourse, formulation, and implementation of policy. The latter are worked out through the opaque and complex negotiations and power bargaining between various levels of government jurisdiction, as analyzed in detail by the corporatist literature on China.30 Such an approach would be extremely difficult in the Tibetan areas given the political and security sensitivity in these areas. The constraint is circumvented here by focusing on ex-post inputs, outputs, and outcomes. Preference has been given to data not derived from institutionalized systems of rewards or penalties, given that these can skew the data considerably. To take the example of education, ex-post government expenditure data offer the closest indication of the reality of government rhetoric on their inputs to education. Per capita numbers of schools and teachers reveal the real output deriving from this expenditure. In terms of outcomes, indicators such as enrollment levels are highly susceptible to local manipulation given that they can be crucial for determining local budgetary or public employment allocations,31 whereas ex-post educational attainments derived from annual population surveys provide more accurate indications of actual education outcomes in the population. Obviously, within this broader research strategy, analysis has also been limited by what has and has not been divulged in official sources. Besides the population census data, this mostly precludes any ethnic disaggregation of data, although ethnic disaggregation is presumably available in the original unpublished data sets.

40

Chapter 1

The qualitative methods of this study were also somewhat eclectic given the broad scale and political sensitivity of the issues researched. Formal surveys were not possible and the field methods were informal in practice. A wide variety of field sites was sampled, each with much less depth than would be accorded by ethnographic methods, for instance. This approach was taken partly to avoid spending too much time in each community for political reasons and partly as a means to trace links between events and regions in an effort to understand processes of circulation across the regional system of the Tibetan world. The resultant fieldwork was analyzed by drawing out and emphasizing the commonalities or similarities that appeared across various settings and respondents, or themes that were found to have a strong sense of consensus, as a means to tease out broad systemic patterns regarding social experiences of contemporary social and economic transformations. The resulting generalizations are meant to be understood in this perspective. While it is true that there are enormous variations of experience and perception across Tibet and even between two valleys, what I found interesting is that precisely despite the variations, there are still strong commonalities that can nonetheless be observed across this cultural and topographic region the size of western Europe, giving credence to indigenous notions of “Tibet” that transcend modern political boundaries. These commonalities are likely due to the existence of common processes of circulation, whether discursive, physical, or even financial, inside China and abroad.32 The fieldwork included fourteen months spent in Tibet and western China between June 2003 and January 2005, two subsequent one-month visits in December 2005 and September 2007, and regular visits to China since then. This was more generally informed by seven years of pre-doctoral field experience living in Tibetan refugee communities in India and Nepal from 1995 to 2001. The fieldwork in Tibet was conducted in all three of the major Tibetan regions (Utsang, Kham, and Amdo) and in four of the five Chinese provinces containing Tibetan areas (the TAR, Qinghai, Sichuan, and Gansu). The largest portion of time was concentrated in Amdo/Qinghai due to a range of concerns, from logistics and freedom of movement to research interest. Extensive visits were made to both farming and pastoral areas in Qinghai, Sichuan, and Gansu, including repeated and extended contact within eight rural communities. Limited rural travel was also undertaken in the TAR, although this was restricted due to the heavy controls over the movement of foreign researchers outside of Lhasa. Urban areas visited included (Tibetan names used for Tibetan towns): Lhasa and Shigatse, the two largest cities in the TAR and in Tibet overall; Xining and the main towns of Xunhua, Chentsa, Rebgong, Sogwo, Chabcha, Mangra, Jyeku, Dawu, and Darlak in Qinghai; Lanzhou, Linxia, and Labrang (Ch. Xiahe) in Gansu; Chengdu and the main towns of Sershul, Derge, Manikango, Kardze,



Introduction 41

Drango, Dawu, Tagong, Dartsedo (Ch. Kangding), and Ngawa in Sichuan; and Beijing. The urban areas were chosen because of their importance as regional centers for local government administration, education, business, off-farm employment, and migration. The period of intensive fieldwork from 2003 to 2005 entailed informal and unstructured interviews, conversations, focus groups, participant observation, and general living experience within households in several of these settings, in both rural and urban areas. Some interviews were conducted by me in Tibetan (mostly in Lhasa dialect), some were conducted by me in English with Tibetans or Chinese who spoke English, and about half were conducted with the assistance of either a Tibetan translator (in the case of interviews with Tibetans, particularly those from eastern Tibet) or a Chinese translator (in the case of interviews with Chinese or Muslims and some educated Tibetans). Interviewee selection was determined through snowball sampling given that the politicized nature of the research precluded representative sampling. The exact number of informants is difficult to quantify because in many cases contact was made in very fluid social settings.33 However, it is possible to enumerate roughly 228 key informants, in terms of contacts that resulted in significant and substantive field insights through time spent and questions asked in extensive interviews and discussions (all informal and unstructured). For reasons of security, identity and precise locations will be kept as vague as possible in most cases. The sampling was disproportionately weighted toward more elite informants, particularly local Tibetan, Chinese, or Muslim scholars who are a wealth of knowledge but are considerably more constrained than foreign researchers in the dissemination of their findings. These informants also included reincarnate lamas, monks and nuns, pilgrims; government officials (including various heads of various county and prefecture government departments); INGO and local NGO staff; professors, researchers, and MA- and BA-level university students; primary and secondary school teachers and numerous secondary school students; unemployed Tibetans; journalists; businessmen; farmers; nomads; traders; factory managers and owners; construction workers; returning refugees; and Tibetans, Han Chinese, Hui and Salar Muslims, Mongolians, Mongours, and one Manchurian.

Outline of the Book The argument of this book is presented in eight chapters. It starts with a chapter on the historical cleavages of nationality in the Tibetan areas as they emerged from the late nineteenth century onward as background to understand the differential impacts of recent development across the three broadly defined nationalities of Tibetans, Chinese Muslims, and

42

Chapter 1

Han Chinese. In chapter 3, the confluence of local population transitions among Tibetans (e.g., demographic transitions and urbanization) and the in-migration of Han Chinese and Chinese Muslims is discussed. Chapter 4 traces recent macroeconomic polarization in terms of the structure and sources of rapid economic growth, and the external orientation of the dominant processes of circulation and accumulation driving growth. This results in most of growth being concentrated within restricted channels in the urban areas, much more so than elsewhere in China. Sectoral polarization in turn sets the basis for labor, wage, and household income polarization, which is analyzed in chapter 5. Chapter 6 synthesizes the analysis by examining the resultant exclusionary dynamics that are catalyzed at both the lower and upper ends of the labor hierarchy, with a particular emphasis on underlying educational inequalities and linguistically and culturally articulated networks of political and economic power in the context of intensified migration. The last two chapters deal with various responses to polarization and exclusion. Chapter 7 introduces the concept of “subsistence capacity” as a means to examine how underlying conditions of subsistence wealth, as opposed to income poverty, provide the means from which people can act strategically in the economy, particularly with respect to various choices of integration into the off-farm employment. In this sense, subsistence capacity can be seen to form the material micro-foundations of the double movement, given that it provides the freedom to act strategically. This is explored further in chapter 8, which examines the conflictive or collaborative repercussions of exclusion as various aspects of double movement. It does this by first focusing on the ethnically exclusionary dynamics of de-industrialization and the deregulation of urban employment in Tibetan towns, and how this has fueled non-secessionist Tibetan nationalism as one logical response to the disjuncture between education and urban employment systems. The chapter then examines two different recent patterns of inter-ethnic relations: a rise in Tibetan economic activism directed against Muslims on one hand and a rise in collaborative private interactions between Tibetans and Han Chinese through religious channels on the other. On this note, the conclusion returns to the central argument of this book with regard to marginalization within peripheral and disempowered rapid economic growth.

Notes  1. For background on the protests and their aftermath, see Fischer (2008c), Shakya (2008) and Smith (2009).   2.  For instance, see Makley (2009) regarding a demonstration in Rebgong, Qinghai Province, that occurred before the protests in Lhasa.



Introduction 43

  3.  A similar analysis is presented more extensively in Jin (2010).  4. This campaign is usually translated by many—including me up until recently—as the “Western Development Strategy.” However, I have opted for “Open the West Campaign” after discussions with Lara Maconi, given that this offers a more accurate translation of the Chinese words xibu da kaifa, which convey a sense of opening and (resource) exploitation rather than development. In particular, the translation of the word kaifa as “development” is arguably inappropriate. According to Maconi (2008), the Chinese word used for development is generally rendered as fazhan, meaning “to develop, evolve, or progress.” For instance, this is how “development” is translated in the China Human Development Report. Kaifa is generally used to signify exploitation (as in mineral exploitation), adding value, or reclamation (as in opening up wasteland by pioneers). The two terms are not interchangeable. According to Maconi, fazhan is much more general and neutral whereas kaifa carries much more connotation, including political connotation, in the sense that it implies the “exploitation” of that which is primitive, uncultured, wild, and non-civilized. Hence, in the discourse of Beijing, East China is to be developed, as it is already considered to be evolved, whereas West China is to be opened up, given that it is considered to be particularly backward. Theoretically, kaifa signifies bringing progress to something that is completely untouched, or to bring civilization to where it does not exist. Obviously, in hindsight, my adoption of this translation has been long overdue given that xibu da kaifa had already been translated by Goodman (2004a) as the “Open Up the West” campaign.  5. “Absolute” here refers to its usage in poverty studies, referring to absolute measures that can be applied (in principle) as a meter for all, such as the income required to purchase food items equivalent to 2,100 calories per person per day as well as other essential non-food items. On the other hand, “relative” measures are applied with reference to context and social norms, similar to (or the same as, depending on the measure) inequality measures. For instance, six years of primary education can be considered an absolute measure of educational attainment, whereas it is relatively deprived in comparison to a minimum norm of nine years of compulsory schooling. Similarly, comparing average educational attainments across two or more groups gives an indication of relative disparities in educational attainment.   6.  Indeed, the co-authored work of Barry Sautman and Yan Hairong since 2007 on contesting dominant Western discourses suspicious of Chinese involvement in Africa is written very much along these lines.  7.  Melvyn Goldstein has made this point to me on several occasions with respect to the potential to revise our understanding of Mao Zedung from the perspective of his approach toward Tibet policy in the 1950s. This reveals a more conciliatory side than typically thought in that he was willing to allow the Tibetan government headed by the Dalai Lama to administer Tibet internally until it was ready to accept Socialist reforms, versus more radical and intransigent elements within the CPC who favored using force to implement reforms quickly. See Goldstein (2007).  8.  For a more conceptual and theoretical discussion of these points, see Fischer (2008a; 2011a).  9. The term “value-added” is a basic accounting term used in economics to denote the increase in value that is added to a combination of inputs in a production process. For instance, abstracting from all other input costs, if 10 kg of grain is

44

Chapter 1

planted and results in a harvest of 100 kg of grain, the value added is 90 kg of grain. In classical political economy (i.e., from Adam Smith to David Ricardo and Karl Marx), it would be roughly equivalent to the “surplus,” in the sense that it includes wages and profits, hence involving political economy conflicts over the distribution of the surplus between wages, profits, and rents, or between workers, capitalists, and rentiers (i.e., landlords). In modern national accounting, value-added is measured in monetary terms, that is, the money value of the output minus the money value of the inputs, and it is divided between wages, taxes, depreciation, profits, and other income payments such as net interest, rents, and royalties. Labor is not included as an “input” and wages are not treated as input costs but as part of the value added of the production process. Hence, higher wages effectively add value to the output of goods and services in modern national accounting. 10.  Standard mainstream (neoclassical) approaches to understanding economic growth and development (e.g., the Solow model and its subsequent variants, or else measures of total factor productivity) suffer from a seemingly intractable existential problem in theorizing economic growth because of an insistence to treat capital not as a produced means of production that enhances the productivity of labor, but instead as another factor input like labor, that can substitute labor as well as consumption goods, and also shares the qualities of a consumption good. If capital is not theorized in this way (i.e., if it is treated as something that is not substitutable with consumptions goods, such as tractors and wheat), then the whole theoretical corpus of neoclassical models is essentially rendered indeterminate and intractable. However, because of this simplistic reductionism, there is an underlying confusion as to whether capital accumulation refers to the accumulation of produced means of production (e.g., machines and infrastructure) or of any form of (financialized) wealth more generally. However, the accumulation of financial wealth in the absence of increasing productivity in the wider economy is usually predicated on financial means of appropriating the surpluses of others, not on capital accumulation. 11.  For further detail on this point, see Fischer (2011b). 12. Indeed, I was “accused” of making a Marxist argument by an anonymous reviewer of Fischer (2009b), although this reviewer mistakenly identified my structuralist analysis as Marxist, demonstrating a superficial understanding of structuralism (and dependency theory), as is quite typical in the mainstream economics and development studies literatures. Of course, structuralism and Marxism are not mutually exclusive. 13.  CEPAL is the Spanish acronym for the United Nations Economic Commission of Latin America and the Caribbean (ECLAC). 14.  See Polanyi Levitt (2005) for a synthesis of the thinking of Prebisch. 15. It is questionable to what degree Furtado or Sunkel should be associated with the subsequent school of “dependency theory.” Kay (2005, 1205) traces one of the earliest uses of the term “dependency” by a social scientist to Furtado (1956), while he was still working in CEPAL, although he also notes that Furtado was later criticized by Cardoso and other “dependency theorists” for having an exaggerated confidence in the ability of the state to lead an endogenous process of development (Kay 2005, 1204). I am also indebted to discussions with Kari Polanyi Levitt for insights on these matters.



Introduction 45

16. For those unfamiliar with this classic literature, Arthur Lewis was one of the most seminal pioneers of early development economics from St. Lucia in the Caribbean. He developed a theory of “economic growth with unlimited supplies of labor” in which industrialization would gradually exhaust reserves of surplus labor, only after which wages would start to rise with rising labor productivity. 17.  Paraphrased from a quote cited in Kay (2005, 1204); also see Furtado (1973) for a corrective to his earlier stagnationist hypothesis regarding Brazil. 18.  The principle of ownership has become common parlance in the so-called “Post-Washington Consensus” phase of mainstream development policy since the late 1990s, although it is usually used with reference to participatory principles in project management, or else as a technique to create more responsibility among domestic actors in the dissemination of policies, such as by devising ways to increase fiscal responsibility among donees. As such, it has often been used as a rhetorical guise to soften the fact that, after three decades of financial crises and structural adjustment in most of the developing world, effective policy space has been increasingly constrained by the dictates of international financial institutions. This is not the meaning of ownership used here. Rather, in accordance with a standard understanding of what ownership would mean to ordinary people (such as “owning” a house), I refer to ownership as the real ownership of assets and control over the main institutional levers of economic policy making, meaning that governments can decide policies rather than simply deciding how to deliver them. In terms of foreign (or non-local) involvement, ownership would imply that local authorities have control over decisions regarding who invests in what sectors and where, according to the needs and priorities determined by domestic constituencies rather than foreign non-elected technocrats. 19.  For perspective, when I started my research in 2001, real rural incomes in the TAR had been stagnant for about a decade from the early 1990s onward—see chapters 2, 5, and 7, and also TIN/Fischer (2003a). 20.  See Fischer (2005a, 1-6) for a more extensive discussion of these distinctions. 21.  This point was raised by my PhD supervisor, Athar Hussain. The Han are a classic “melting pot” case with a strong tendency throughout history to assimilate and absorb neighboring ethnicities. Deliberate attempts to assimilate minorities into the emerging modern Chinese nation-state clearly predated the PRC, as discussed by Tuttle (2005) with respect to the rise of Chinese nationalism in the late nineteenth and early twentieth centuries and its impact on Sino-Tibetan relations. For an excellent study of the historical evolution of Chinese discrimination toward Chinese Muslims, see Lipman (1997). 22.  For instance, Roth (2004, 63) argues that the “strength of organizations like Human Rights Watch is . . . their ability to investigate misconduct and expose it to public opprobrium. That methodology is most effective when there is relative clarity about violation, violator, and remedy. That clarity is best achieved when misconduct can be portrayed as arbitrary or discriminatory rather than a matter of purely distributive justice.” 23.  For analyses of these issues in China, see Qiang Li (2002), Guo and Chang (2006), Lulu, Li (2008), and Liu (2008). For an excellent survey, see Roulleau-Berger (2009). 24.  The “primary sector” is the national accounting term for economic activities related to the production of primary inputs, as opposed to the transformation

46

Chapter 1

of these inputs by secondary activities—construction and manufacturing—or else non-physical services in the tertiary sector. The primary sector is composed of farming, animal husbandry, forestry, and fishing (mining and quarrying are part of the secondary sector). In Tibetan areas, the primary sector is about half farming and half animal husbandry (pastoralism). 25.  See Dyson (2011) for an excellent discussion of urbanization from a global demographic (rather than economic) perspective. 26. The term “regional system” refers to the quantitative mapping of regional systems in China by Skinner (1994). According to his method, we might even suggest that the dualism of the Tibetan areas, whereby most Tibetans were quite peripheral to the modern urban economy until recently, actually allowed for overlapping regional systems, one centered on China and the other centered on local regional poles. 27.  Typically, even within China, most Tibetologists are not trained in economics, while most economists who are sent to study the economy of Tibet (invariably the TAR) are not trained in Tibet studies and do not speak Tibetan. 28.  Examples include Sautman and Eng (2001), who trumpet the emerging “Tibetan middle class” while more or less ignoring the problems of urban labor market exclusion, or else Dreyer (2003), who offers a decent albeit superficial survey of recent trends. 29.  See a discussion of the latter in Childs (2008). 30. A variety of different political economy theories of a corporatist lineage have been developed to explain local developmentalism within the Chinese polity, such as Blecher (1991), Blecher and Shue (1996), Cannon (2000), Oi (1992), Nee (1992), Solinger (1993), and Whiting (2001). 31.  For instance, a typical practice in one Tibetan town that I visited in Qinghai was to bring non-students into the classrooms during visits by officials from Xining. The concern was that teaching positions or funding would be cut if the surveyed enrollment was too low. Similarly, one INGO health worker in the TAR reported to me that maternal and infant mortality rates are underreported due to the fact that doctors are fined if either type of death takes place during a delivery. 32.  See Morcom (2011) for a fascinating ethno-musicological study that makes a similar point. 33.  For instance, many of my most informative moments with Tibetan teachers or students were in large social events, often lasting several days. A typical “interview” with farmers or nomads usually involved spending the better part of a day drinking butter tea and gossiping about the latest news from the Tibeto-Buddhist world, while interjecting a pertinent question for my research only once or twice an hour. While the host might have remained constant throughout such conversations, many people would usually visit and participate for shorter intervals. As a result, it is difficult to enumerate the exact number of informants interviewed during my fieldwork.

2 The Inception of Modern Development in Tibet under PRC Rule

There is no doubt that modern socio-economic development in Tibet started with the arrival of the Chinese Communists in the 1950s. Before this time, the starting point in at least Central Tibet could be characterized as a Tibetan form of serfdom, although there has been considerable debate as to the use of this term “serfdom” in Tibet Studies.1 As in most premodern settings around the world, life expectancy was low and health indicators were poor (in comparison to modern standards). Education levels were very low in modern forms of literacy and schooling given the absence of modern schools, leaving the monasteries as the mainstays of literacy and learning (Monasteries did nonetheless reproduce levels of literacy that were probably comparable to most parts of pre-modern Eurasia, and they were also the source of an impressive output of scholarly publications and knowledge.) Besides a few small pioneering efforts in the first half of the twentieth century, most aspects of a modern economy were absent. At the turn of the twenty-first century, five decades of Communist rule had resulted in heavily subsidized modern urban economies transposed onto rural areas where 85 percent of the Tibetan population resided (in the TAR as of the 2000 census). These rural areas remained largely subsistence based until the late 1990s, in the sense that rural households consumed much of what they produced and sold the surplus rather than conducting farming in a commercialized manner. However, even these rural areas had undergone profound transformations and in no way could be portrayed as “traditional”; traditional patterns of land ownership were replaced by collectivization from the mid-1950s outside the TAR and from the late 1960s in most of the TAR, and communes were in turn replaced by the 47

48

Chapter 2

reallocation of land to individual households in the early 1980s. Moreover, rural-urban distinctions became quickly blurred from the late 1990s onward given a rapid pace of labor migration and urbanization alongside rapid economic growth. As a result of the intensive subsidization that accompanied the integration of the Tibetan areas into Chinese regional economic planning, modern education and health care are now widespread in the Tibetan rural areas, although they are under-supplied and often of very poor quality compared to the rest of China. There are roads and other infrastructure that did not exist before, although some of the technology often referred to in official Chinese publications on Tibet (e.g., PRC 2001a), such as Internet and satellite television, did not exist anywhere in the world in the 1950s. Notably, following a huge government effort in the TAR from the mid-1990s onward, there is now solar-powered television reception in all townships and most villages, even those without electricity or running water (e.g., see PRC 2001a, question 61). This emergence of modernity in Tibet has been fundamentally experienced by Tibetans as an encounter with Han Chinese, in the sense that most of these structural and institutional transformations are difficult to disassociate from the incorporation of Tibet into the People’s Republic of China (PRC) from 1950 onward and the entry of Han Chinese into this region on a significant scale for the first time in history. As noted by Tuttle (2005), Han Chinese in fact played quite minor roles in the premodern era of most Tibetan areas, whose relations with imperial China were largely mediated by Mongols or Manchus from the thirteenth century onward (albeit perhaps less so in the eastern parts of Kham, bordering the Sichuan basin). Regardless of these past imperial relations and the various ways that they might be interpreted, the domineering involvement of the PRC in the direct governing of Tibetan areas—with a strong Han bias (or “chauvinism”) and an ideological animosity toward most aspects of traditional Tibetan society, especially religion—was an entirely new political and institutional landscape for Tibetans. The PRC definitely ushered in the subsequent structural transformations whether intentionally as a matter of policy or else as a matter of historical coincidence and conjuncture. Indeed, the PRC was able to establish its supremacy and direct rule over this remote and difficult-to-access region only through modern technologies of transport, communication, and weaponry, which prior to the twentieth century were simply not in existence. Similarly, as discussed in the next chapter in more detail, the fact that Chinese rule occurred alongside various aspects of demographic transition—such as falling death rates and increasing life expectancy—is arguably a historical coincidence given that the imposition of rule happened precisely at a time when the same demographic processes were occurring throughout the developing world (including in China and in most parts of Africa). However, the fact that these transformations took



The Inception of Modern Development in Tibet under PRC Rule 49

place within a radical reconfiguration of power relations with China during the tumultuous events of the twentieth century has resulted in a close association and even conflation within the politicized discourse about Tibet between these aspects of modernity and the effective occupation of Tibet by the PRC since 1950. Many scholars have written about various historical dimensions of the relations between Tibet and China in the buildup to this juncture of conquest and development, including me on the evolution of co-existence between Tibetans and (Chinese) Muslims (see Fischer 2005b; 2008b). This chapter focuses more specifically on the transition to PRC rule in Tibet and the subsequent inception of modern development, with the objective to establish a degree of clarity regarding various historical reference points on development for the rest of the book. Certain themes are also emphasized in this chapter. One is the legacy of the early incorporation of Tibet into the PRC for later patterns of disempowered and dependent peripheral development in Tibet, in particular the annihilation of the traditional wealth and institutions that could have served as the basis for an indigenously driven modernization in Tibet, and the subsequent financial void that was created in the wake of the radical restructuring of the rural society and economy. Another theme is the economic stagnation in the TAR that lasted from 1985 to 1995, with the suggestion that this might be one hitherto unexplored dimension of the urban protests in Lhasa in the late 1980s. If so, these earlier protests of the reform period would have been quite distinct from the wave of protests that started in 2008, and that have been occurring in a context of rapid economic growth. Lastly, the chapter offers historical background to understand the re-emergence of conflictive fault lines between Tibetans and Chinese Muslims in the reform period, particularly in Amdo, as well as patterns of inter-elite cooperation between Tibetans and Han Chinese through the vehicle of religion despite the memory of intense violence in the Maoist era and despite ongoing Tibetan contestation of the authoritarian and assimilationist hegemonic project of the Chinese state. These objectives and themes are presented in three parts. The first offers some brief reflections on the interethnic basis from which Tibet entered into the PRC and the increasing contestation of the interethnic power balance in Tibet—particularly in northeastern Tibet, where I concentrated much of my fieldwork—parallel to the rise of the modern post-imperial Chinese state in the early twentieth century. The second section deals with the historiography of development in Tibet, including some cautionary notes regarding standard PRC narratives and some hypothetical speculations that might serve as antidotes to these narratives. The third reviews the regional development policies instituted in both the Maoist and early reform periods of the PRC and their consequences

50

Chapter 2

for Tibetan areas. The conclusion ends with some reflections on the nationalist revivals that have occurred during the reform period together with the interethnic themes that emerge from this tumultuous modern history of development in Tibet.

The Pre-PRC Ethnic Co-Existence in Tibet The very inception of Tibetans, conceived today as a collectivity inhabiting the Tibetan plateau, speaking various Tibetan dialects based on a common written language, and, for the most part, practicing a form of Buddhism assimilated from India and adapted into what is now known as Tibetan Buddhism, arguably runs parallel to the specialized integration of Tibet into the world system since the rise of the Tibetan Empire in the seventh century.2 Despite its stereotyped isolation, Tibet has been intricately interwoven into international power dynamics since this time. The pattern of fragmented integration into larger spheres of power that dominated Tibetan history for most of the second millennium and which, to a large extent, pre-empted the revival of centralized rule following the breakdown of the Tibetan Empire, is itself is related to what Janet Abu-Lughod (1989) refers to as the first wave of globalization, predicated on the expansion of the Mongolian Empire in the thirteenth century. Throughout these centuries, inter-ethnic relations in Tibet were fluid, changing, and contested. By no means were they in a stable state, characterized as either harmony or animosity. Nonetheless, from the thirteenth to the nineteenth century these relations evolved along a fairly consistent conception and structuring of hierarchy within the Tibetan areas. With the collaboration, patronage, or even neglect of various empires emanating from China, Tibetans ruled their local turf and controlled land-based wealth, while Muslims generally played subordinate economic roles, mainly specializing in trade, commerce, and certain services.3 The presence of “Han” Chinese was very limited in Central Tibet before the 1950s, although less so in the eastern Tibetan borderlands of Kham and Amdo. Obviously, within these broad characterizations, it is important to note that significant differentiation and often violent divisions have existed among Tibetans (and among closely associated Tibeto-Buddhist groups, such as the Mongols or Monguors in Qinghai or the Qiang in Sichuan), as they have among other groups such as the Muslims and Chinese.4 The most manifest schisms among the TibetoBuddhists have been sectarian and regional.5 In this sense, although Tibetans are here identified as a group, this is not meant to imply homogeneity of purpose or loyalty throughout space, time, or even class. Rather, the ethnicity of conflict in Tibet often involves many complex



The Inception of Modern Development in Tibet under PRC Rule 51

layers of interpretation, most of which are not explored here due to the attention given to drawing out certain common patterns at the regional level, as explained in the introduction. With regard to Central Tibet, regional interaction with other cultural groups from Inner Asia generally seems to have continued as normal despite the restrictions on contact with Europeans that was established in the late eighteenth century. For instance, small groups of Muslim traders and workers came from Sichuan or northwest China over many centuries and were known in Tibetan as the Hopaling,6 while wealthier Muslim merchants came from Kashmir, Ladakh, and Nepal, and they all mostly settled in Lhasa, Shigatse, and several other main towns, numbering several thousand by the twentieth century. By 1959, when the Fourteenth Dalai Lama escaped into exile, their combined population in Lhasa, the largest population center, was estimated at around two thousand (Cabezon 1997, 20).7 These “Tibetan Muslims” intermarried with Tibetans and assimilated into the dominant culture, including the adoption of the Tibetan language.8 While these intimate cross-cultural exchanges were restricted to the main urban centers, Tibetans from all Tibetan regions typically made pilgrimages to these centers and thus would have at least heard of the Muslim presence in these centers. The brief British invasion of Lhasa in 1904 catalyzed a critical shift in the imperial (and post-imperial) representation of China in Tibet toward an increasing and then total representation by Han Chinese. As noted by Tuttle (2005, 43), prior to the British invasion nearly all the ambans in Central Tibet had been either Manchus or Mongols. The integration of Central Tibet into the Qing Empire was governed through a mutual conception of interstate diplomacy that was deemed beneficial to the elites of both states, albeit the Manchu more or less considered Tibet as one of their satellites. However, Han Chinese started to play an ever greater role in running the affairs of the Qing Empire in the late nineteenth century due to challenges to imperial rule all over China, and they took the upper hand in the control of Tibetan affairs for the first time following the British encroachment. Tuttle argues that the new involvement of Han Chinese fundamentally altered the nature of Qing-Tibetan relations because it corresponded to the rise of Chinese nationalism in China, whereby “a modern territorial policy took shape that has continued to the present day” (2005, 44). This policy remained unfulfilled for several subsequent decades given that the collapse of the Qing dynasty in 1911 allowed the Thirteenth Dalai Lama, who had been in exile during both British and Manchu occupations, to regain control of Central Tibet and to expel the few remaining Manchu or Chinese soldiers and residents. From 1913 until the invasion of the People’s Liberation Army (PLA) in 1950, the territory controlled by Lhasa was effectively independent, duly demonstrated by their protracted military campaigns in

52

Chapter 2

eastern Tibet in the 1920s and 1930s against a disorderly array of Nationalist and local Sichuanese and Muslim warlord forces (cf. Goldstein 1989b; Shakya 1999; and Peng 2002). In eastern Tibet, the demise of the Tibetan Empire in the ninth century left both Amdo and Kham largely under the control of a heterogeneous mix of petty kingdoms, tribes, powerful monasteries, and other localized organizations of power, with various degrees of Mongolian overlordship throughout the course of the second millennium (cf. Borjigidai 2002; Kapstein 2000, xix; and Nietupski 1999, 2002). Both had been nominally absorbed into the Qing Empire by the eighteenth century, although in many cases direct rule was applied with much difficulty if at all. Remote regions remained effectively self-ruling, such as in Golok, where some tribes would occasionally acknowledge outside authority by accepting official ranks from either Lhasa or Beijing, or both (Horlemann 2002b, 246). Local authority was seriously challenged only from the late nineteenth century onward, although there was arguably no regional administrative state structure or centralized authority in these regions that could have constituted the basis of a relatively autonomous state and hence a vessel for demands of sovereignty, as was the case with the Ganden Phodrang government based in Lhasa. Central Tibet did nonetheless wield considerable decentralized influence in these areas up to the mid-twentieth century through monastic and trulku networks (lineages of recognized reincarnations of lamas). Generally, religious networks throughout Tibet, and even farther afield in Mongolia and the Mongolic regions of Russia, were (and continue to be) affiliated with colleges in the large monasteries in Central Tibet, where monks and trulkus were (and are) sent for higher studies. The resulting networks thereby formed (and continue to form) very strong links of loyalty and politicoreligious authority throughout the Tibetan areas that centered on Central Tibet (and now partly center on the exile community in India). In other words, lack of central rule in the form of a regional state apparatus did not imply the absence of alternative regional systems of political authority.9 Prior to the Communist takeover, monastic and trulku estates also acted as important centers of political and economic influence in their local arenas, especially in eastern Tibet (in lieu of a centralized state apparatus). They were involved in trade, commerce, and finance; were the organizational hubs for militias and local armies; and held considerable influence over non-religious elites. Most of the natural Tibetan towns or cities today developed around important monasteries, which were usually located in strategic places.10 As a result, monasteries were usually in command of considerable local financial and military strength, either directly or through alliances, and often acted as the centers for organized armed resistance up to the mid-twentieth century, including those during the first decade of Communist rule.11



The Inception of Modern Development in Tibet under PRC Rule 53

Within these nexuses of power, Amdo presents a particularly strong example of the patterns of interethnic coexistence that characterize these eastern Tibetan regions.12 The main population centers of Amdo are located alongside the historic Gansu Province,13 which was a crossroad of the Silk Road between the Turkic west, the Mongolian north, the Tibetan south, and the Chinese east, and is one of the main centers of Chinese Muslims. Because of this location, Amdo has served as a focal point for Tibetan-Muslim relations for centuries. To illustrate this proximity, the Muslim center of Hezhou (today called Linxia) in modern Gansu Province and known euphemistically as “the Chinese Mecca” is very close to the extensive Labrang monastic complex (Ch. Xiahe), which up to the 1950s was one of the largest Tibetan monasteries in Tibet outside Lhasa.14 Hezhou/Linxia in particular has served as one of the main sources of Muslim trade in the Tibetan areas up to the present, even as far south as Lhasa or southern Kham.15 Nearby in Qinghai, the Salar Muslims are said to have settled and intermarried with local Tibetans as early as the thirteenth century in what is now the Xunhua Salar Autonomous County in the Haidong District of Qinghai.16 This county is surrounded by traditionally Tibetan areas and includes four Tibetan townships and the birthplace of the late Tenth Panchen Lama.17 It borders Rebgong, the county that seats Rongwo (Ch. Longwu) Monastery, another power and religious center of Amdo. Kumbum Monastery (Ch. Ta’ersi), the other major monastery of this core area of Amdo founded in the sixteenth century by the Third Dalai Lama, is very close to Xining City. Xining had been a Chinese garrison city since at least the twelfth century with large Muslim populations on its outskirts for almost as long (Lipman 1997, 12-14 and 159-65; Goodman 2004b, 384). Thus, ironically, although Amdo Tibetans clearly see Muslims as outsiders in Amdo, most of the Muslims within Amdo are indigenous to very proximate if not overlapping regions of northwest China, and many hail from Muslim families that have been residing in Tibetan towns for generations. The Tibetan-Muslim interface in Amdo has therefore been much denser and more contested than in Central Tibet, to a certain extent trumping the Tibetan-Chinese interface within Amdo itself. Nonetheless, despite contestation, cooperative synergies have been equally dense. For instance, even more present than in Central Tibet, intermarriage was relatively common between Tibetans and local Muslims,18 as well as with Mongolians and Monguors, who were mostly Tibetan-Buddhist, and with local Han Chinese, many of whom were also Buddhist.19 As a result, intermixing between these “nationalities” has been common, with identity and loyalty defined more or less along religious lines. As in to Central Tibet, Muslims in Amdo filled specialized economic roles that continue to the present. Traditionally, they were based in market towns and moved between pastoral and agricultural zones as traders, merchants,

54

Chapter 2

middlemen, brokers, and translators (Lipman 1997, 13). They controlled much of the trade in livestock products, particularly wool and hides, and also worked as butchers for Tibetans. In many cases Muslims were invited by Tibetan rulers to fill these specialized roles, and their settlements and trade routes influenced the development of many Amdo Tibetan towns.20 Unlike in the present, these trading and service roles of the Muslims would not necessarily have been perceived as a threat by Tibetans in the past, given that the Muslims were performing work that was generally stigmatized by Tibetans. In this regard, Tibetans, and in particular Tibetan nomads, with their abundant land in comparison to neighboring Muslim or Chinese areas, were relatively wealthy within a regional pre-industrial economic perspective.21 The Tibetan-Muslim symbiosis in Amdo nonetheless came to be underlain by competition and violent confrontations. Increasing tension was probably stimulated by the rise of Muslim militancy in the northwest during the demise of the Ming dynasty in the seventeenth century, which coincided with the spread of Sufism into Gansu (Lipman 1997, 53-57). In particular, Xining, Xunhua, and Hezhou became key organizational hubs for the Sufi movement, which included proselytizing among non-Muslims.22 Rising Muslim influence therefore came to contest local Tibetan/Buddhist dominance and probably stimulated activism among Tibetan/Buddhist elites, such as the founding of Labrang Monastery in the beginning of the eighteenth century.23 Until the Muslim uprisings in the second half of the nineteenth century, local Tibetan rulers nonetheless managed to hold the upper hand of Manchu favor under the imperial policy of using minorities to rule minorities in the borderlands. Tibetan armies or militias often assisted the Qing military in various pacification campaigns against Muslim rebellions in neighboring Muslim strongholds, such as Xunhua or Xining.24 However, the ethnicity of such alliances was also ambiguous given that both Chinese and Tibetans had joined earlier Muslim uprisings during the demise of the Ming dynasty and loyal Muslim forces were also often used in Qing pacification campaigns to put down rebellious Muslims, with loyal Muslims often fighting side by side with Tibetans, Mongols, and Chinese.25 In Chone and Taozhou, local Tibetans often offered protection or refuge to their Muslim neighbors against attacks by non-local Muslims (Lipman 1997, 152 and 188). Nonetheless, for most of the Qing dynasty, pacification gave special attention to Muslims, who were portrayed as naturally violent and prone to rebellion, while patronizing preference was given to Tibetan elites.26 Eroding Hierarchies of Ethnic Co-existence The hierarchy of ethnic co-existence started to erode toward the end of the nineteenth century, starting first in Kham and Amdo and in symbiosis with



The Inception of Modern Development in Tibet under PRC Rule 55

the breakdown of imperial China and the emergence of the modern Chinese nation-state. In Amdo, the patronage-instilled ethnic power balance slowly reversed toward the end of the Qing dynasty. Several of the loyalist Muslim warlord families in the northwest managed to gain considerable Qing imperial favor (and military co-dependence) due to their role in pacifying the Muslim rebellions and supporting central rule in Gansu in the late nineteenth century. Following the demise of the Qing dynasty in 1911, they successfully navigated both favor and local authority over to Nationalist allegiance (Lipman 1997, 171). The rise to power of the Ma Muslim warlord family based in Huangzhong in Qinghai (Ma Qi followed by his nephew Ma Bufang) in the first part of the twentieth century was particularly relevant for Amdo Tibetans. Ma Qi was appointed by the Nationalists as governor of the newly created province of Qinghai in 1928. Ma Bufang took over leadership of the province following the death of his uncle in 1931, and “established essentially a separate, and Islamic, state-within-astate in Qinghai under the Republic,” relying heavily on the Hui and Salar in the administration and military while excluding Tibetans, Monguors, and Mongols (Goodman 2004b, 386). The creation of Qinghai incorporated large parts of the vast Tibetan territories of Amdo and Kham (Yushu), which hitherto had been only symbolically incorporated into the historic Gansu Province. Ma Bufang attempted to enforce obedience in these territories through punitive military campaigns, which in turn were faced with much resistance. Counter-resistance tended to be violent, often involving the destruction of monasteries and the sacking of towns. Most of Qinghai eventually submitted, although about half of Golok was never subdued.27 Ma Bufang was also involved in military campaigns in northern Sichuan in the 1930s, fighting alongside Nationalist and Sichuan warlord forces against the army from Lhasa (Peng 2002, 66). The success of Ma brought a reconfiguration of territory, population, and even ethnicity in and around the northeast corner of Qinghai due to his policies of ethnic cleansing and forced conversion.28 Hualong, Xunhua, and Minhe Counties of the present-day Haidong District, where Ma sourced his Muslim power base amid sizable surrounding Tibetan populations, were particularly targeted.29 The aggressive policies combined with autarkic rule created an exodus of both Tibetans and Muslims from these core counties, which accounted for some of the new presence of rural Muslim households in places such as Chabcha to the west.30 In some locations, attacks from Ma strengthened local Tibetan-Muslim alliances, such as in the case of attacks on the Muslim compounds of Taozhou in the late 1920s, as discussed by Lipman (1997, 196-97). Given that the energies of the Nationalist government in China were absorbed by Japanese invasion and civil war from the 1930s, the Communist victory in 1949 was the first time that modern China had the capacity to

56

Chapter 2

enforce any direct central rule on these areas. The subsequent administrative reorganization of the rural areas by the Communists had the effect of fixing the de facto configurations of these ethnic power struggles as they had evolved up to the 1950s, given that counties were ethnically demarcated on the basis of their majority ethnic composition and individuals were assigned ethnic status according to their de facto religious affiliation in the 1950s.31 Collectivization and control over population movements subsequently prevented further encroachments on Tibetan rural land from taking place, other than in certain borderline or suburban cases, which will be discussed further in the next chapter. Ironically, when this fixing of territory and identity took place in the 1950s, Amdo Tibetans were in a subordinate position to the Muslim warlords, having lost significant amounts of territory and people in several key locations. Farther south, ethnic co-existence in Kham was more tied up with the increasing Manchu efforts to colonize and control this region from the eighteenth century onward. For instance, the drawn-out and very expensive Manchu military campaign in the neighboring region of Gyelrang in the eighteenth century, as discussed in detail in Greatrex (1994), contributed to a psyche among successive Manchu rulers that the Qing empire was vulnerable to destabilization via its Tibetan backyard and thus required forceful subjugation. This was particularly the case following the British invasion of Lhasa, which catalyzed a reevaluation of Manchu strategy in the Kham areas of Sichuan, resulting in intensified military campaigns in the last years of the Qing dynasty. These were later continued by the Nationalists with more or less the same psyche (Coleman 2002; Peng 2002; van Spengen 2002; and Tuttle 2005). It is also important to note that, up to the Communist period, Tibetan resistance in Kham (as in Amdo) had not always been directed toward central Chinese authority per se. In the case of three prominent Kham self-rule movements in the 1930s, each was closely allied to and supported by the Nationalist government in Nanjing. Their alliance was in mutual opposition to the local Chinese warlord Lui Wenhui as well as to the Lhasa government in central Tibet (Peng 2002; Tuttle 2005). Moreover, many prominent lamas also resided in Nanjing at this time. Most prominent was the Ninth Panchen Lama, who was in exile from Central Tibet as a result of conflicts with the Thirteenth Dalai Lama. During this exile, he was very active in propagating Nationalist policy and values in Amdo, Kham, and Inner Mongolia, and in turn, informing the Nationalists about the situation in these regions, particularly regarding the Mongolian princes who were defending against Japanese invasion (Jagou 2002). As documented by Tuttle (2005), he also gave numerous large-scale teachings, which included the attendance of Chinese leaders such as former presidents of the republic, and bolstered tremendous popular support for Tibetan Buddhist teachers and practices among Chinese Buddhists. Indeed, Tuttle argues that, following two failed strategies



The Inception of Modern Development in Tibet under PRC Rule 57

to entice Tibetan elites through secular nationalist rhetoric, the Nationalists started to leverage a religious rhetoric to advocate national unity in the 1930s. This strategy gained the support of prominent exiled Tibetans (such as the Panchen Lama), who apparently hoped that the Chinese government would restore them to their positions of power in Tibet. The strategy also “had a significant impact on the Chinese imagination of unity with Tibet for the next decade, suggesting that this now shared religion could bridge the gap between the estranged peoples” (Tuttle 2005, 12). Communist Conquest and Immediate Aftermaths The occupation of Central Tibet by the Chinese Communists in 1950 culminated in the Tibetan uprising in 1959, the escape of the Fourteenth Dalai Lama, and the uninhibited launch of Central Tibet into radicalizing Maoism. This was the first time that any form of central Chinese rule had ever become directly involved in the local management of Central Tibet. It did so through the force of conquest, given that the Communist Party of China (CPC) was lacking a pre-existing indigenous organizational base or supporting social movements, unlike elsewhere in China. It was also the first time that Han Chinese people had entered Central Tibet on a large scale. The history of this period in Central Tibet has been covered in detail by Goldstein (1989b; 2007) and Shakya (1999). Suffice to say that TibetanHan relations became synonymous with the imposition of Communist rule and an anti-religious ideology. Kham and Amdo led the way in resisting the Communists. In both cases resistance was not necessarily directed toward Chinese rule per se, which had come and gone many times over the centuries and was often received with various degrees of local support, as discussed above. For instance, Horlemann (2002b, 247-48) notes that the CPC did not immediately include Golok into their liberation plans in 1950. Official contacts were allegedly first established only in 1951 after some groups from Golok took the initiative and Golok apparently remained surprisingly calm until 1958. Rather, the main points of contention were the radical Socialist fast forwarding into collectivization and the dismantling of monastic influence. Both radical policy shifts started in earnest in the mid-1950s, with the CPC’s reneging on earlier promises even though these Tibetan areas outside the TAR had not yet undergone the intermediate stages of so-called Democratic Reforms, as had the Han areas (Goldstein 1998, 8). As a result, armed resistance became very stiff in Kham and Amdo from the mid-1950s onward. Resistance was often organized around some of the most prominent monasteries in each region, as exemplified by the uprisings organized around the two prominent monasteries of Litang and Batang in Kham (Sichuan), which resulted in their being bombed and shelled by the PLA

58

Chapter 2

(Goldstein 1998; also see Shakya 1999, 136-44). Refugee movements from Kham to Central Tibet also destabilized the situation in Lhasa. The early Communist leadership also apparently assumed that Amdo Tibetans would support them because they had defeated Ma Bufang in 1949, but soon thereafter they themselves were faced by resistance from both Tibetans and Muslims. As in Kham, Amdo Tibetan resistance started in the mid-1950s and in some cases joined forces with the resistance movement in Kham. The peak of resistance occurred in 1958, when several Tibetan regions in Qinghai broke out into full-scale guerrilla warfare, occurring side by side with an uprising organized by a Salar Muslim in Xunhua motivated by similar grievances. As in Kham, resistance fighters as well as refugees also fled from Amdo to Central Tibet.32 Together with the famine of the Great Leap Forward, the combined violence of these events far exceeded the loss of life experienced in China or by other insurgent minorities such as the Muslims. For instance, the official Tibetan population of Qinghai experienced a sharp drop by about one-fifth between 1957 and 1963, despite population growth before and after these years.33 This includes the effects of the famine of the Great Leap Forward, during which mortality rates in Qinghai were among the highest in the country.34 Considering population replacements over these years, a more realistic estimate of effective population loss would exceed one-third of the 1957 Tibetan population.35 Qinghai Muslims experienced a proportionately smaller, albeit significant, loss of population, although only between 1959 and 1963.36 In comparison, it is generally estimated that the loss of life in China due to the famine of the Great Leap Forward amounted to about 5 percent of the population.37 These insights imply that the loss of life experienced by Tibetans in this province—whether through insurgency or famine—was exceptionally severe. The memory definitely remains vivid in oral histories and in contemporary conceptions of grievance. Informal accounts and some statistical records suggest that Tibetan experiences were similar in Sichuan, Gansu, and Yunnan. The TAR was less affected by the famine and violence started later, although informal accounts suggest a similar severity of repression.

Reflections on the Historiography of Development in Tibet From this perspective of the bath of fire and blood that Tibetan areas passed through in their incorporation into the PRC, some cautious and speculative observations on the inception of modern development in Tibet during this period are worth highlighting. First, caution is required with regard to the use of summary statistics describing various aspects of social or economic



The Inception of Modern Development in Tibet under PRC Rule 59

change during this period from the 1950s onward. An uncritical acceptance of the usual indicators incanted by the government and unfortunately relayed by many scholars and journalists acquiesces to the politics of statistical representation played by the government. There are three main reasons for caution: earlier data are highly inaccurate, mostly involving estimated computations often on the basis of dubious assumptions;38 the lack of price indices for the TAR prior to 1990 renders a proper evaluation of real (i.e., inflation adjusted) economic trends very difficult; and, most importantly, presentations of smooth cumulative trends in certain long-run variables (such as economic output) mask the massive and enforced upheavals and transformations of the society we are purporting to measure. As a result, we effectively have little idea whether the statistics (especially from the 1950s and 1960s) refer to actual additions to wealth or simply expropriations of indigenous wealth and displacements from indigenous to PRC-centered forms of accumulation. Keeping these cautions in mind, while it is true that Tibet more or less embarked on modern development following the Chinese takeover in the 1950s, this is not to say that it would have remained static in the absence of Chinese rule. More likely, it would have embarked on its own process of modernization as occurred in all countries of Asia, the path of which can only be speculated. The only certainty is that the Tibetan economy of the late 1940s—and its elites in particular—would have served as a starting point for an autonomous economic transition, probably aided by China or other countries, much as Bhutan has been heavily aided by India in its own development. On the other hand, official PRC portrayals of the old society of Tibet as a feudal society with most of its population living as immiserized serfs and chattel slaves (e.g., see PRC 2001a) is heavily influenced by ideology and offers little insight into the realities of pre-modern Tibet. Specifically, it draws from Marxist theories of ethnicity that were adopted and then further elaborated by Stalin in the 1930s and later adapted to China by the CPC. Minority nationalities were categorized according to their stage of economic development, ranging from slave societies, as in the case of several of the Yi and Yao tribes in southwest China, right up to feudal, advanced feudal, and even nascent capitalist stages of development in the case of various Han regions (see Lipman 1997, xx; Harrell 2001, 7). Tibet was categorized as a theocratic feudal system (even though Tibetan Buddhists do not believe in a “theos”). Ethnic and religious identities were accordingly considered to be manifestations of these lower stages of historical material development. They would therefore eventually recede under material and scientific progress, best achieved in Tibet by opening it up to the more advanced regions of China and allowing for the dissemination of rationality, technological capabilities, and the like.

60

Chapter 2

Such simplistic theories of modernization, whether derived from a Marxist or a mainstream perspective, are problematic in that they impose categories and classifications from the outside, often with little relation to local realities or indigenous systems of reference. The Marxist lineage also leads to a conflation of ethnicity and religiosity with material conditions, despite the fact that even some of the most advanced economies of the world today, such as the United States, are still very ethnically stratified and very religious. Unfortunately, most studies of the historical Tibetan economy produced in China are heavily constrained by this official ideology. It is in fact very difficult to make any rigorous comparisons between poverty in the old economy and poverty at the beginning of the intensive development drives in the mid to late 1990s. Data is lacking from the past whereas present data is of poor quality. The old and new economic systems are also completely different, meaning that processes and conditions of impoverishment have changed substantially over time. If the notion of “absolute” poverty is considered, that is, the ability of a household to produce enough output or earn enough income in order to support its essential food needs (2,100 calories per day per person) and non-food needs, absolute rural poverty in the TAR was as high as 25 percent in 1999, according to official income distribution data and national poverty lines (Fischer 2005a, 96-110). These rates of absolute rural poverty were among the highest in China at the time and they corroborate with the high rates of stunting among Tibetan children observed and measured by health workers in the TAR in the 1990s (e.g., see Harris et al. 2001). It is conceivable that absolute rural poverty could have been as high in the late 1990s as it was in the pre-PRC past, when the population might have been quite effective in providing for its own basic food needs, albeit with seasonal fluctuations and periodic shortages as is typical in pre-modern agricultural systems, even if it did not have paved roads, airports, telecommunications, modern cities, Internet cafes, and karaoke bars. Indeed, the economic system up to the 1950s was largely subsistence based, with wool as one of the main internationally tradable commodities. In that context, the main poverty consideration would have been the ability of households to feed themselves based on their own production or labor, or on barter and trade. However “backward” the old economy of Tibet was, there is no reason to assume a priori that one-quarter of its population was unable to provide itself with basic subsistence food needs (i.e., the absolute rate of poverty measured in 1999). The relatively high rates of poverty today (if we assume these measures to be accurate—see chapter 7) might be due to a variety of factors, such as population growth on limited land resources and falling agricultural prices, many of which are not necessarily the direct intention or result of government policies specific to the TAR since the 1950s. Nonetheless, the validity of any poverty comparisons between the old and



The Inception of Modern Development in Tibet under PRC Rule 61

new economies requires serious questioning, particularly in contexts where the rural population remains embedded in its agrarian subsistence base, as it was in Tibet until at least the late 1990s. Indeed, it is plausible that the more collective approach of managing land in the old economy (and in the Maoist collective economy) provided more potential for labor specialization than in the fragmented individual household responsibility system established in the reform period, especially in the pastoral areas, thereby allowing for greater labor productivity and surplus, albeit with a more extensive use of land than in the present. Furthermore, certain aspects of the old economic system functioned with an efficiency that the modern state still has a difficult time matching in the dispersed and low population density rural areas of Tibet. For instance, the sheer pervasion of the monastic system, while to a certain extent aborting the development of a strong centralized state in Tibet in the first half of the twentieth century, nonetheless played a key circulative role in the dispersed rural economies throughout the Tibetan areas, which functioned parallel to the state and far exceeded the reach of the Ganden Phodrang government based in Lhasa. In particular, monasteries acted as the dominant financial institutions, providing for an impressive accumulation of financial assets via non-state channels, channels that were to become the main object of attack and expropriation in the Communist occupation of Tibetan areas. For instance, the PRC government estimated that just prior to the dissolution of Drepung Monastery in Lhasa in 1959 the monastery had outstanding loans worth ten million yuan in cash (worth five million 1959 US dollars, or equivalent to about 250 million yuan in 2000)39 and 130,000 tons in grain (Goldstein 1998, 23). Other monastic institutions, such as Sera and Ganden, would have had similar financial magnitudes. Considering that loans in the banking system of the TAR amounted to just over eight billion yuan in 2000, of which only 87 million yuan was designated for agriculture (TSY 2010, table 14-2), it is clear how a few monastic institutions in Lhasa alone, in the pre-modern and unsubsidized conditions of the 1950s, would have accounted for a significant proportion of even the entire financial system in the TAR today. Moreover, as a reflection on the above comment that the official data do not represent the displacement of indigenous forms of wealth, these data in TSY (2010) record loans of only 13.79 million yuan in the TAR banking system in 1959, which evidently does not include the cash loans of the monastic institutions mentioned above, that in total would have far outweighed this amount. When the thousands of large and small monasteries spread out across the Tibetan areas are considered, the extensive breadth and depth once sustained by the indigenous financial system can be appreciated. Given the tendency for every family to have had a relative in a monastery or nunnery and for every populated valley to have had a monastery or nunnery of some

62

Chapter 2

sort, the outreach, accessibility, and efficiency with which the indigenous financial system would have been able to operate is indeed striking. It would have been able to perform functions from providing micro credit for small-scale production or welfare in the context of local shortages, up to financing large trading companies in Lhasa, the government, and a wide range of social functions. The system would have been essential for the financing of disperse mountainous agriculture and herding, and of the extensive trade networks both within Tibet and abroad with Nepal, India, and even England, such as the wool trade. Indeed, monasteries or reincarnate lamas often played central roles in trade as well as finance, particularly in Kham and Amdo. The Communists typically portray this role of monasteries in the old economy as predatory and exploitative. Exploitation was, no doubt, involved, as in all financial systems, including those of China or the United States today. Nonetheless, the veracity of these assertions is evidently imbued with ideological motives. Indeed, there is no reason to assume that the current state-run financial system in the Tibetan areas is any less exploitative. For instance, it has been generally observed in China that the financial system drains resources away from rural areas by lending or investing a large part of rural savings outside of the rural areas—or at least this was the case in the first two decades of the reform period.40 This is reinforced by the fact that rural dwellers, more than half of whom were illiterate in the Tibetan areas in the 1990s, would have had great difficulty accessing the formal state-run banks. As a result, rural lending in the Tibetan areas has tended to be monopolized by officials or other wealthy and well-connected people. Whether corrupt or not, the formal system tends to drain savings away from the rural economy, while the monasteries tended (and still tend) to retain them and even bring in savings from elsewhere (such as in the case of attracting donations from wealthy Chinese sponsors in Shanghai, Hong Kong, or Taiwan today, as discussed in chapter 8). Hence, from the perspective of the circulation of local aggregate demand, it is quite plausible that the indigenous financial system was more pro-poor and expansionary than the financial system that came to supplant it, at least up until the early 2000s. The attention to rural development via subsidized bank lending under the Eleventh Five-Year Plan from 2006 onward, as discussed by Goldstein et al. (2010), might have changed this tendency in the formal banking system in the TAR. These questions also need to be considered more broadly in terms of the developmentalist objectives of financing industrial policy and structural transformations of the economy. State-directed finance in the PRC has definitely played a huge role in this regard whereas monastery-centered finance did not (and does not today, although some monasteries have tried to adopt various social welfare functions, with much resistance by local governments). However, this consider-



The Inception of Modern Development in Tibet under PRC Rule 63

ation is different from the concern with exploitation and the comparison is also tenuous, given the pre-capitalist and pre-developmentalist context of the old indigenous financial system in Tibet. In any case, events related to the radicalization of Maoism in the 1950s brought about the destruction of this monastic system. The main landholding groups (the monasteries, the aristocracy, and the Ganden Phodrang government in Central Tibet) were eliminated by the 1960s both inside and outside the TAR, and their assets were expropriated. This started earlier outside the TAR, that is, in 1955-1956 in the Kham areas of Sichuan, while it was postponed until 1959 in the TAR. The closure, expropriation, and eventual destruction of the vast majority of monastic institutions in the period leading up to the advent of the Cultural Revolution in 1966 therefore had the effect of wiping out the entire indigenous financial basis of the Tibetan economy. Loans were canceled, estates and granaries were confiscated, the valuables and collateral of the monasteries, such as gold and precious metals, were largely sent to China, and the physical infrastructure of the monasteries was either disassembled, destroyed, or vandalized. The experience was similar across all the Tibetan areas. The ensuing collapse of the financial sector would have wreaked havoc in the other sectors of the economy dependent on such finance, such as agriculture or trade, albeit these were simultaneously or soon thereafter collectivized or taken over by the state. Trade in Central Tibet was also heavily impacted by the closure of the main trade routes with India, particularly during and after the war between India and China in 1962, and trade was more or less entirely rerouted through Sichuan or Qinghai. Furthermore, a large proportion of those Tibetans who were skilled in business, such as monastery officials or aristocrats, along with the “Tibetan Muslims” of Ladakhi origin, escaped into exile or were ruined, incarcerated, or killed. Thus in a matter of years, the elites and their wealth that could have served as the basis for an indigenously driven modernization in Tibet were effectively wiped out. It therefore comes as no surprise that the Tibetan economy was heavily dependent on funding from the central government by the advent of the reform period given that, among other factors, a financial void would have been created in the wake of occupation. The state took over the direct management of this distant and distinct region through administrative means on all levels, from the micromanagement of the rural economy all the way up to provincial government finance. This involved a huge security and military dimension given that the 1962 war with India was fought from two locations within the TAR itself. Wang and Bai (1991 [1986], 73) note that the TAR government revenue fell sharply into deficit and the level of subsidy started to increase sharply precisely in 1968, in the midst of the Cultural Revolution and just before the onset of wide-scale rural collectivization in the TAR. Yan (2000, 161) also cites one of the Chinese books

64

Chapter 2

she reviews that the TAR government expenditure was already heavily subsidized in 1952. However, the data she cites must refer to the parallel administration set up by the PRC following the 17-Point Agreement in 1951, which by agreement was to avoid placing undue burden or strain on the local economy at the time and thus was funded mostly by the central government from its inception. Hence, these data would not refer to the fiscal situation of the Ganden Phodrang government, which co-existed with the parallel PRC administration in Lhasa up until 1959. Indeed, the TAR was officially only created in 1965. Such misrepresentation, whether intentional or not, is an example of how data from this period can create a false impression of incremental progressive change rather than displacement. Instead, the pattern of intensifying subsidization in this region from the 1960s onward paralleled the militarization of the region and the radical restructuring of the rural society and economy, which would have accounted for well over 90 percent of the Tibetan population at the time. This legacy is vital to understanding development in Tibet ever since. This legacy was the result of the radicalization of Maoism under occupation rather than occupation per se. During the first decade of Communist rule, state ideology became progressively more hostile to the indigenous socio-economic system, which in the end was simply not allowed to co-exist with the new rule and was obliterated throughout the course of the Maoist period. Had Maoism not radicalized in the mid-1950s, it is conceivable that Tibetan elites and their wealth could have played a significant role in developing a modern Tibetan economy, even within the context of Chinese occupation. In particular, had moderate land reform been undertaken, as in the early 1950s elsewhere in China, without engaging in a wholesale attack on the monastic institutions, it is conceivable that Tibetan elites, having lost their traditional sources of wealth (landed estates), would have transferred their wealth and business acumen into a variety of new activities in the emerging economy, as was the case with the land reforms in South Korea and Taiwan in the early 1950s. Indeed, this is precisely what many of these elites did in exile, using their wealth that they escaped with (or accumulated in places such as Darjeeling) along with their entrepreneurial capabilities to successfully integrate into the modernizing economies of their host societies. In contrast, collectivization and animosity toward the monasteries and other aspects of the indigenous socio-economic system resulted in the major Tibetan uprisings in Kham and Amdo from 1955 onward. The economic dimensions of these conflicts should not be underestimated. Of course, Maoism radicalized in both Tibet and China, and the social structure was turned upside down in each. However, revolutionary zeal emerged as an indigenous process in China, whereas it did not in Tibet. There was some Tibetan participation in the mayhem of the Cultural



The Inception of Modern Development in Tibet under PRC Rule 65

Revolution, such as in the destruction of whatever few monasteries remained by that time. For instance, Wang (1998, 314-23) emphasizes that most of the destruction of monasteries during the Cultural Revolution in Tibet was organized by thousands of returning Tibetan students from Han inland cities who were, he argues, the sons and daughters of poor Tibetans and were the most enthusiastic participants of the destruction, whereas Lhasa and other areas that were controlled by the conservative-leaning PLA and where most the Han were located saw less damage. Nonetheless, this point is relatively trivial given that, as pointed out by Goldstein (1998, 9), the destruction of the monasteries as functional institutions already took place between 1959 and 1965, and was more or less complete before the advent of the Cultural Revolution. The Cultural Revolution in Tibet merely eliminated “all remaining vestiges of religion. ” Rather, the radical restructuring of Tibetan society was imposed from the outside, in accordance with political trends evolving in the Han areas of China with little relation to Tibet. The reform period similarly originated in China and was imposed on the Tibetan areas from the outside. Moreover, despite the demise of Maoism, the legacy of Chinese nationalism has been maintained as a central part of state ideology, whereas Tibetan nationalism has been treated with suspicion and as counter-revolutionary. As a result, once the reform period began in the late 1970s, many elements of the Chinese elite who survived the Cultural Revolution quickly re-emerged as leaders in the new market economy, whereas the suspicion of these classes in Tibet remains as a predominant legacy even to the present and is compounded by the fact that many of these Tibetan elites are now in exile. As a result, the reversion of the social order has been much slower and much more entangled in paranoiac politics.

Maoist and Reformist Regional Development Policies Fate was such that Central and Western Tibet became locations of key strategic military significance for the PRC, particularly given the 1962 war with India. Similarly, most of Kham and Amdo were incorporated into provinces that comprised the privileged core of Maoist industrialization strategy, namely Gansu, Qinghai, and Sichuan. From the mid-1960s until the early 1970s, this strategy pushed industrialization into the interior of China, particularly heavy industry, in what later became known as the “Third Front Strategy.” The orientation was predominantly militaristic and although the ethos emphasized egalitarianism and self-reliance, the Third Front industries were nonetheless heavily subsidized by the central government,

66

Chapter 2

particularly those connected to the military. The beginning of the end of this strategy was détente with the United States starting in 1971 rather than reform in 1978, which allowed for a reversion of the national industrial strategy back toward the more efficient coastal provinces (Wei 2000, 28; Cannon and Jenkins 1990, 36–38; Yang 1997, 25–26). The Maoist strategy of interior industrialization was supported by what Yang (1997) calls the “circular payments system.” A regional system was established whereby state-fixed prices for energy and raw materials (including agricultural products) were kept low in order to subsidize processing industries, which were predominantly concentrated in the coastal areas despite interior industrialization. These industries in turn were heavily taxed and their surpluses were returned in the form of transfer payments to the raw material–producing areas, predominantly located in the interior and western regions (Yang 1997, 62). Thus, Shanghai turned over to the central government a surplus equivalent to more than 50 percent of its GDP in 1978–1980, while provinces like Inner Mongolia, Ningxia, Xinjiang, and Qinghai received subsidies from the central government equivalent to over 20 percent of their GDP, and the TAR over 60 percent (UNDP 1999, 65). Much of these subsidies were oriented toward the interior industrialization strategy or the parallel and autonomous military structure in the western provinces, mostly bypassing the rural areas. The regional development strategy was further complemented by what has been called the “price-scissor” strategy of rural-urban development within regions, whereby agricultural prices were undervalued in order to subsidize urban output and wages (cf. Yang 1997; Knight and Song 1999). Notably, this strategy was in accordance with fairly classic late industrialization policies used in Japan in the late nineteenth century, in the USSR in the 1930s, and in South Korea and Taiwan in the 1950s and 1960s.41 However, unlike the regional transfer system, this rural-urban transfer did not necessarily include a backflow of subsidies from urban to rural areas given that rural areas were expected to be self-sufficient despite the fact that their output was undervalued. This system was particularly relevant for the TAR given that it institutionalized the intensive subsidization of this autonomous region by the central government in the 1960s. Subsidies in turn rose to an unprecedented level relative to GDP from the late 1960s onward, as noted above. However, industrialization under the Third Front Strategy was generally concentrated in strategic centers that were isolated from military fronts and sensitive border areas, which by definition precluded most of the TAR. The Tibetan areas outside the TAR were more influenced by the industrial components of the strategy. In particular, Qinghai served as one of the main national centers for prison labor camps, earning it the notorious reputation as the “gulag” of China, as well as a major nuclear weapons research and development center



The Inception of Modern Development in Tibet under PRC Rule 67

and test site (located in a Tibetan area), which has since been turned into a museum. The Third Front requirements for regional self-sufficiency also generated considerable demand for forestry and mining resources in the Tibetan areas located close to Chengdu, Lanzhou, or Xining, and it marked the intensification of mineral exploitation in the Tibetan areas of western and northern Qinghai. Even in these Tibetan areas outside the TAR, most Tibetans were nonetheless sidelined by the enclave nature of the Third Front Strategy. Yang (1997, 24) contends that the strategies were quite inefficient, particularly in the western economies where scarce resources were squandered in inefficiently planned and located heavy industries, and “projects were built under the direction of a small group of central leaders and with little regard to local interests.” Because of “the dominance of the ‘unit’ mentality, which was strengthened owing to the military nature of most of the projects, most of these plants had few linkages with the local economy and contributed little to the development of these localities, including most areas inhabited by ethnic minorities” (Yang 1997, 23). These considerations would have been especially significant in the Tibetan areas given their political and security sensitivity, and mistrust for the patriotism of Tibetans more generally, particularly considering that some Tibetan areas in Qinghai and Sichuan remained restive into the 1960s. With the advent of the reform period from the late 1970s onward, major shifts in regional development policy led to immediate macroeconomic stagnation in the northwest provinces and the TAR in the 1980s despite the simultaneous resurgence of the rural economies in these areas due to de-collectivization in the early 1980s. First, industrial investment was further redirected away from the interior and toward the coastal areas. Second, fiscal decentralization and enterprise reforms quickly eroded the system of regional redistribution by allowing for greater local retention of tax revenues and profits, while imposing greater fiscal self-reliance on local governments. This resulted in a marked decline in the level of transfer payments that was directed by the center to the interior and western provinces. However, the price-scissor strategy of under-pricing agricultural and raw material output was maintained up until the early 1990s as a means of subsidizing coastal industries, despite the fact that the disappearing transfer payments were originally intended to offset this under-pricing (cf. Wong 1997; Yang 1997; UNDP 1999, 65; Liu and Lin 2001, 87; Knight and Song 1999). These combined factors had a disproportionately contractionary impact on the interior and western provinces, given that their economies were more agrarian and raw material based than the coastal areas, and their public revenues were heavily dependent on subsidies (whereas coastal provinces were largely in surplus, financing central revenues rather than being financed by them). The Tibetan areas

68

Chapter 2

were especially impacted given that they were the most agrarian in China and were among the most heavily subsidized throughout the 1970s. The strategy was successful in generating rapid, albeit unbalanced, national growth. Coastal poles led the growth, where an increasing share of national investment was redirected and most foreign direct investment was concentrated. Fiscal decentralization further reinforced the rich areas but had a debilitating effect on the poorer “deficit” areas that were burdened with lower health and education attainments, poorer infrastructure, and an inefficient industrial legacy. The combination of these factors naturally strengthened the coastal growth poles, which in turn acted as sponges, attracting human, material, and financial resources throughout the country. As a result, even though the pre-reform pricing policies were gradually dismantled and liberalized in the late 1980s and early 1990s, the momentum of regional disparities continued more or less without respite into the mid-1990s. In addition to these policy issues, the slow growth of agricultural incomes after the mid-1980s had a further disproportionate impact on the interior and western regions. The early 1980s had seen a quick boost in rural incomes throughout the country as a result of the one-off effects of de-collectivization and a large and sudden (administered) increase in the terms of trade for agricultural goods. Nonetheless, agricultural output growth slowed considerably from 1985 up until about 1994 (Brandt et al. 2002, 68). As a result, rural per capita incomes in the predominantly agricultural regions stagnated in real (inflation-adjusted) terms for much of this period.42 Growth in national average rural incomes was driven by the expansion in non-farm rural activities such as township and village enterprises (TVEs), yet TVEs were also concentrated in the coastal areas, as were opportunities for non-farm wage employment (Woo 2002, 245). In contrast, the fate of the western rural areas remained much more closely tied to the fortunes of agriculture, in particular to grains and animal husbandry. Grains in turn were particularly stagnant between 1985 and 1994. Given that the Tibetan areas were the most rural and agrarian of the country, rural Tibetans—especially Tibetan farmers—were especially susceptible to these agrarian dynamics. Such dynamics are aptly described by the story of wool. Wool became one of the first commodities to be fully liberalized in the 1990s, in terms of domestic pricing and supply as well as international trade (Yang 1997, 7172). The role of administrative controls to keep the price of wool low was thereby replaced by the impact of international competition.43 As a result, the terms of trade for wool relative to rural consumer prices in Qinghai fell by around three-quarters between the late 1980s and 2004. This calculation is based on farm-gate (or “ranch-gate”) prices at the time of field research in 2004 and equivalent prices in the late 1980s that were recalled by various



The Inception of Modern Development in Tibet under PRC Rule 69

informants, including pastoralists, wool traders, manufacturers, and scholars in Qinghai, Gansu, and the TAR. For instance, according to informants in Qinghai, average quality sheep wool from lower-altitude locations such as Malho and Tsolho (and Labrang in Gansu) was selling at about twenty yuan per kilogram in the late 1980s, whereas in 2004 it was selling for about six to ten yuan per kilogram, depending on quality.44 Over the same period of time, the rural consumer price index of the TAR and Qinghai more than doubled. Hence, the purchasing power of one kilogram of wool effectively fell by at least three-quarters. My informants generally concurred that most of the fall in nominal prices occurred in the early to mid-1990s. The drastic decline in these terms of trade acted as a powerful downward pull on the real value of Tibetan rural incomes given the importance of herding in most Tibetan areas. More recently, since the late 1990s, this has been compensated for by rising prices for meat and skins, as well as by the very sharp price increases for caterpillar fungus and the subsequent bonanza in many pastoral areas where caterpillar fungus is found.45 Similar price trends were also observed in the main Tibetan farming commodities, such as barley, wheat, and rapeseed. For instance, Goldstein et al. (2008, 519) report that barley prices in rural Shigatse Prefecture in the TAR rose from 1.4 yuan per kg in 1982 to 1.74 yuan in 2006. Given that rural consumer prices in the TAR increased by about five times (or 400 percent) between 1982 and 2006,46 this implies that the real purchasing power of a kilo of barley also fell by about three-quarters over these years. As with wool, it is likely that much of this decline happened in the 1990s. Thus, despite profuse PRC propaganda to the contrary, the 1990s were a particularly stagnant time in the rural economies of Tibet. Indeed, if the official income data are to be believed, rural incomes in the TAR were stagnant in real value from 1992 to 2003 (see further discussion of these income data in chapter 5). Despite such apparent stagnation, there was not yet much indication in the 1990s of any significant movements of local Tibetan labor out of agriculture, at least not in terms of the overall composition of employment, as also reported in the official data (see chapter 5). As discussed in the following chapter, this might have been due partly to relatively rapid population growth in the working age groups of these Tibetan rural areas in the 1990s, which would have maintained the share of the labor force working in agriculture despite an increasing number of people moving out of agriculture and out of the rural areas altogether. As a result of all these combined dynamics, the TAR and Qinghai were among the worst cases of economic lagging in China from the beginning of the reform period in the late 1970s until the mid-1990s, together with Ningxia and Gansu, two other northwestern provinces. As discussed below, the TAR was in deep recession for about ten years. Qinghai was also stagnant or slightly recessionary in real terms from 1993 to 1996, and Gansu

70

Chapter 2

from 1993 to 1995. The fall from pre-reform grace in the TAR and these four northwest provinces was as dramatic as the meteoric rise of southeast China over the 1980s and 1990s (cf. Naughton 2002; Hussain 2002b; Fischer 2005a, 16-31). For instance, at the beginning of this period, the per capita GDP of Qinghai was substantially higher than the national average, in contrast to provinces in the southeast such as Fujian, a coastal province that had been neglected during the Maoist years due to its proximity to Taiwan and which subsequently became one of the coastal trading powerhouses. By 1991, the per capita GDP of Qinghai was still 1.03 times the national average (measured in constant values according to province-specific CPIs), while that of Fujian was still only 0.9. In an almost mirror reversal, by 1997, the ratio of Qinghai had fallen to 0.75 whereas that of Fujian had risen to 1.42. The ratio of Gansu fell from 0.67 to 0.48 over the same years. The ratio of the TAR to the national average similarly fell from 0.8 in 1991 to 0.48 by 1995, which was the sharpest and fastest drop among the provinces in China. In other words, the two provinces that account for about three-quarters of the Tibetan population and territory, as well as an important concentration of Chinese Muslims, emerged as the leading laggards amid rapid national economic growth. Due to this dismal performance, the TAR had become one of the poorest provinces in China by the mid-1990s, falling to the second lowest per capita GDP in China in 1996, as shown in figure 2.1.47 While Guizhou remained by far the poorest province in China throughout all these years (and to this day), the per capita GDP of Gansu (typically the secondpoorest province) had surpassed that of the TAR for the first (and last) time from 1994 to 1997. Similarly, macroeconomic conditions in the Tibetan Figure 2.1: Real per capita GDP, selected provinces, 1991-2000 (constant 2010 yuan)

Sources: calculated from CSY (2011, tables 2-1, 2-15 and 9-5) and equivalent in previous CSYs.



The Inception of Modern Development in Tibet under PRC Rule 71

areas outside the TAR were probably similar to if not worse than those in the TAR given that they were much less fiscally privileged than the TAR and they were additionally located in the less-privileged hinterlands of other western provinces, hence they would have been more severely affected by these dynamics of regional development. Indeed, there were indications that absolute (or basic needs) poverty rates were increasing in some of these western provinces between the late 1980s and mid-1990s, such as in the survey results presented by Khan and Riskin (2001, 65-67), although their two surveys did not include the TAR, nor Qinghai in the second. The likelihood of worsening living standards in some poorer (and minority) parts of western China amid rising national prosperity was a consideration that the authorities took seriously, especially given the political sensitivity of these regions.48 The Lost Decade of the Tibet Autonomous Region While many of the western provinces nonetheless managed to grow throughout these years, the TAR was languishing in a severe macroeconomic recession for about ten years from 1985 to 1995. This becomes evident only once price inflation is accounted for—it was quite high in the TAR over these years. In other words, while nominal aggregate GDP was rising, as is commonly cited in official government and related scholarly publications, price inflation was rising significantly faster. Hence purchasing power per person was falling, especially once relatively rapid population growth is accounted for. This is can be seen in figure 2.2, which charts the per capita GDP of the TAR—as declared in official statistics—indexed for inflation and normalized by setting the value for 1978 equal to one hundred. Given that the consumer price index (CPI) for the TAR is only available from 1991 onward, the national CPI is used for indexing from 1985 (when the national CPI starts) to 1990, and the national retail price index (RPI) is similarly used from 1978 to 1985. This assumes that price inflation, as recorded by these two indices,49 was roughly the same in the TAR as it was nationally on average, which is plausible because most other western provinces recorded CPI inflation rates in a very similar range as the national rate between 1985 and 1990. Price controls would have also still been largely in force in the early 1980s, especially in western China.50 While the results in figure 2.2 up to 1990 must therefore be seen as approximate, they nonetheless reveal very interesting trends that provide valuable insight into the economic conditions surrounding political unrest in the TAR in the late 1980s, which to the best of my knowledge have not been analyzed elsewhere. Figure 2.2 reveals three distinct phases: rapid (real) economic growth in the early years of the reform period, ending in 1985; a period of (real) economic recession from 1985 to about 1994, punctuated by a one-year

72

Chapter 2

Figure 2.2: Real per capita GDP of the TAR, 1978-1998 (1978=100)

Sources: TSY (2010: table 2-6); CSY (2000, table 9-2; 1999: table 9-5; and equivalent in earlier CSYs.)

improvement in 1990; and (real) recovery starting from 1995 onward. Real per capita GDP surpassed its 1985 peak only by 1998. In the first phase, strong growth from 1978 to 1981 likely represents a recovery from nearfamine conditions in the late 1970s, as described by Dreyer (2003, 415), due in large part to the late introduction of communes from the late 1960s onward together with the forced application of inappropriate “scientific” farming techniques ill suited to the high-altitude ecology of Tibet. Hence, the choice of 1978 as a starting point probably gives a biased impression of subsequent upward trends in the economy given that this year was probably a dismal low point in the TAR, from which any return to normality would appear as a strong improvement in comparison if taken out of context. The strong real growth from 1983 to 1985 most likely represents the initial one-off effects of de-collectivization and the implementation of the individual household responsibility system in the rural areas, which largely took place in 1983 in the Tibetan areas. In the aggregate GDP statistics (not shown), rapid growth in these two years was due to an almost doubling of the nominal (i.e., current yuan) value of the primary sector (mostly farming and herding), and a more than doubling of the value of the tertiary sector. The latter could be explained by an increase in trade and related service-sector activities also due to the reforms, including some initial tourism development, and an increase in government subsidies, appearing as increased administrative spending and cadre salaries following the Second Tibet Work Forum convened in Beijing in 1984. Both of these sectors continued to grow in real terms up to 1990, roughly doubling again in nominal value from 1985 to 1990, versus a national rate of (CPI) price inflation of about 65 percent over these years.



The Inception of Modern Development in Tibet under PRC Rule 73

The economic contraction after 1985 basically stemmed from the secondary sector, which was stagnant in real terms from 1983 to 1985 and then collapsed in both nominal and real value after 1985. Despite the talk of reconstructing the economy in the Second Tibet Work Forum in 1984, the nominal value of construction officially fell by over a third in 1986 and only recovered to its 1985 nominal value by 1991 (and its real value by 1993). The real value significance of this contraction in the secondary sector was much greater given rising price inflation over this period—the national rate of CPI price inflation was 19 percent in 1988 and 18 percent in 1989, while the TAR-specific CPI rate of inflation reached 28 percent in 1994, or a cumulative rate of 117 percent from 1990 to 1995. The price increases from 1990 to 1995 were greater than the nominal yuan increases in the primary and tertiary sectors in these years. The combination of price inflation with relatively rapid population growth explains the sharp contraction in real per capita GDP from 1985 to 1989, and then again from 1990 to 1994. Recovery only started in 1995 through a strong increase in construction, which tripled in nominal value from 1993 to 1995 due to the new wave of subsidies launched by the Third Tibet Work Forum in 1994. Overall, these vagaries in the macro-economy of the TAR were due in large part to the erosion in the value and magnitude of subsidies in the local economy (again, in real terms, once accounting for price inflation). This erosion might not have impacted the Tibetan rural areas significantly, given that these were still largely based on subsistence modes of production and were quite marginalized from the subsidization policies, as noted above, at least until the government started to direct its subsidies toward improving rural household incomes from about 2003 onward. However, the erosion of subsidies and its economic effects would have had strong impacts in the urban areas where most subsidies were concentrated, particularly among salaried urban workers. These economic dynamics of the late 1980s appear to have been largely overlooked in most analyses of the protests that broke out in and around Lhasa at the time. They point to a possible economic dimension of grievance that might have fueled these protests in combination with other considerations. This might help to explain why, as pointed out by Barnett (2009), the protests in the late 1980s were largely urban in focus, given that the economic grievances up to that point would have had mostly urban implications. Rural areas would have still been riding on the many perceived benefits of de-collectivization and the relative tolerance toward restoring the cultural and religious institutions of rural society, whereas the sharply declining purchasing power of Tibetan farming and herding outputs would have been felt mostly in the 1990s. Moreover, such economic grievances need not have been clearly cognized by the urban population. Rather, within a context of rising wages and incomes as well as rising prices, people

74

Chapter 2

would have had a hard time objectively evaluating the balance, although they would have nonetheless experienced significant austerity, reflecting the contraction of real aggregate demand in the urban economy. The resulting cognitive dissonance goes to the heart of many of the contradictions of development experienced by Tibetans—and even by government officials struggling to keep up with events both local and national—during these early years of the reform period.

Conclusion: Nationalist Revivals in the Reform Period amid Rapid Growth Because Chinese occupation and rule during the Maoist period quickly evolved into an attack on the traditional religious social structure, religious freedom quickly came to be associated with liberation struggle and the right to self-determination. This allowed for a strong fusion of national and religious purpose, with the Chinese portrayed as the arch adversaries of both the Tibetan nation and the Buddhist religion. Religious expressions of Tibetan nationalism were therefore naturally reinforced, in Tibet as well as in exile. As a result, the revival of religious space that slowly emerged in the early 1980s inevitably became politicized. For instance, monks and nuns have been at the forefront of dissident activity in the reform period, particularly during the large demonstrations in Lhasa from 1987 to 1989 and afterward. Goldstein (1998, 42) points out that for many ordained, particularly the younger monks and nuns, political dissidence became a religious as well as a nationalistic and political duty given that they had no spouse, children, or other obligations, and thus it was seen as appropriate that they sacrifice themselves for the good of the Dalai Lama, their religion, and Tibet. Schwartz (1994) makes similar arguments. By the late 1990s, available statistics indicated that monks and nuns made up a majority of people detained for political reasons in Tibetan areas (TIN 1999a; 1999b). In the face of such dissent, the Chinese authorities have naturally maintained tight reins of control over monasteries up to the present, particularly in the TAR. In this light and echoing the seminal work by Yeh (2003a; 2013) on applying the idea of hegemony to Tibet, many of the actions of the Chinese state throughout the reform period can be understood as efforts to transform their overwhelming dominance into hegemony. Per a Gramscian notion of hegemony, dominance rests primarily on coercion, whereas hegemony can be understood as the additional power that accrues to a dominant group by virtue of its ability to lead a system of groups in a desired direction in a manner that is perceived as pursuing a general interest.51 In this sense, Chinese rule of the Tibetan areas during the Maoist period was coercive rather than hegemonic. The struggle of the PRC ever since has been



The Inception of Modern Development in Tibet under PRC Rule 75

to forge hegemony within their rule of the Tibetan areas (and in the Uyghur areas of East Turkestan/Xinjiang), such that their dominance is perceived by Tibetans (and Uyghurs) as pursuing a common interest. Most of the recent manifestations of conflict between Tibetans and Chinese, beyond incidences of general hooliganism, appear to revolve around these issues of contested hegemony, by way of either Tibetan protest against PRC rule (even before the 2008 protests) or else state repression of activities deemed counter to its hegemonic project. Tibetan protests were clearly ongoing prior to the 2008 protests, particularly in both Kham and Amdo, which have been known for vibrant and often vocal nationalism.52 They then exploded in 2008 and have since been simmering, with varying degrees of localized outbursts depending on circumstances, most notably exemplified by the recent wave of self-immolations that started in northern Kham and southern Amdo in 2011 and 2012, which have probably been driven by the severity of patriotic education campaigns in many monasteries of this region combined with a variety of other local factors.53 State repression notably targets any religious leader or organization, or any person more generally, that oversteps the informally deemed limits of politically appropriate, state-sanctioned religious activity, which have narrowed considerably since 2008. This includes the well-documented patriotic education campaigns in monasteries and in various work units that started in earnest in the mid-1990s,54 were continued through the 2000s, and were reinvigorated following the 2008 protests. It also included the demolitions at the Serthar Buddhist Institute in 2001 in the Ganzi TAP of Sichuan.55 Such obsessions with hegemony help to explain the ban on photos or worship of the Dalai Lama in the TAR or in some sensitive counties outside the TAR, or the absurd attempts of an atheist state to control the selection of the Eleventh Panchen Lama and to vigorously promote their choice of boy among Tibetans despite almost universal disbelief as well as the more recently announced law to control all reincarnations more generally.56 However, as argued by Robin (2011; 2012; forthcoming) in her work on Tibetan artists, the boundaries of both expression and censor are often an unknown and ambiguous territory that both sides have been exploring and testing, particularly with respect to the Internet. The ongoing contestation of this hegemonic project helps to shed light on the tensions between assimilation and “autonomy” that have characterized Chinese development policy toward these areas since the beginning of the reform period. Much of the development rhetoric of the CPC plays an important discursive role in these endeavors, given the emphasis on the vanguard role of the Chinese state in bringing Tibetans out of backwardness. Religious revival is linearly portrayed within this discourse as a residue of superstitious unscientific beliefs that fetter or impede development, to the extent of causing poverty in Tibetan rural communities due to ongoing transfers of household wealth to monasteries in support of economically

76

Chapter 2

wasteful ventures such as building religious monuments.57 The degree to which educated and secularized Tibetans adopt similar attitudes toward the religious economy or else internalize feelings of inferiority effectively contributes to these hegemonic endeavors, insofar as Chinese rule comes to be seen as a superior alternative to a continued state of perceived cultural and social “backwardness.”58 Secularism therefore inadvertently feeds into the hegemonic project of the PRC and sits uncomfortably with notions of minority nationalism or ethnic pride, whereas religious nationalism captures an impulse of resistance and defiance with monasteries serving as some of the most effective loci for channeling this impulse. The adherence to religious symbols of nationalism is also strong among younger and well-educated generations of Tibetans in the exile community, many of whom argue, for instance, that a natural accord exists between Buddhism and the modern concepts of human rights and democracy.59 Tibetan relationships with the Chinese are rendered quite complex in this regard given the increasingly close private social collaboration between Tibetan and Han Chinese people through the vehicle of Tibetan Buddhism, which will be discussed in chapter 8. Parallel to such contestation and collaboration with the Chinese, tensions between Tibetans and Muslims have risen considerably since the 1980s, particularly in Amdo. Conflicts with Muslims, including occasional flares of violence, are notably social, dealing with day-to-day interactions, often over issues of trade and commerce. In 2003, these local conflicts coalesced with sufficient momentum in Amdo to produce a regional boycott movement by Tibetans against Muslim businesses in Tibetan areas, reaching as far as Lhasa and some parts of Kham. This boycott and its related prejudices are discussed in detail in chapter 8, although suffice to say here that the increasingly conflictive fault lines between Tibetans and Muslims in the context of the modernizing Chinese nation-state from the late nineteenth century onward re-emerged in the reform period as one of the principal foci for Tibetan disaffection with their experiences of development. Both the conflictive and cooperative aspects of these interethnic relations would appear to suggest an aspect of primordialism, meaning the persistence of long and entrenched patterns of co-existence that long pre-dated and survived both the Maoist period of radical social restructuring and the reform period of accelerated growth. However, primordialism is merely apparent due to the fact that for as long as Tibetans and Muslims have been in contact, their relations have been mediated through the ebb and flow of Chinese dominance in this buffer zone between China and Central Asia. Furthermore, the quantum leap in Chinese dominance in Tibet since the formation of the PRC did not guarantee that their rule would become hegemonic. Engaged as it was in radically altering local constellations of power



The Inception of Modern Development in Tibet under PRC Rule 77

and social structure, it is normal to expect that the rule of the CPC would be politically contentious, even while the liberalizing reform period offered opportunities to create cooperative networks in Chinese society that leverage cultural or religious affinity. It is also normal to expect that Muslims would remain the main rivals for Tibetans within such patterns of integration given their equally subordinate position within Chinese dominance. Nonetheless, these patterns also represent a fundamental break with the past. The break is twofold. First, the quantum leap of Chinese dominance in Tibet under the PRC was utterly new, in that it was the first time in history that a state emanating from what we now call China became directly involved with the direct rule of the Tibetan areas. Direct Chinese rule was facilitated by modern technological and organizational state capacity, which was itself achieved through revolutionary consolidation under the CPC. However, unlike elsewhere in China, Communist rule in Tibet was mostly lacking popular roots. Instead, it started as a military endeavor, particularly in Central Tibet. Subsequent rule was dictated by a centralized command structure far removed from the region, in a manner that many refer to as akin to a form of colonial rule. Second, this sudden leap in Chinese dominance took place alongside dramatic social and economic transformation. At first this was the result of radical Socialist planning, such as collectivization, the destruction of the monastic institutions and the landed aristocracy, the introduction of a control economy and a modern state bureaucracy, rapid infrastructure expansion, and early attempts at industrialization, albeit the last three were in large part oriented toward security and military concerns. The expansion of the state also included the first large-scale introduction of modern social service provisioning. Secular public education was expanded, even though in an assimilationist and incomplete manner, with illiteracy rates still ranging well above 50 percent in the mid-1990s.60 Basic health care and pharmaceuticals were also expanded, although the rural health care system subsequently collapsed at the beginning of the reform period, as elsewhere in China.61 These revolutionary changes, even though violently imposed from without and lacking a local support base, subsequently launched Tibet into modern development, with a sociological trajectory very similar to many of the decolonizing countries of Africa and Asia in the 1950s and 1960s, except within an opposite political trajectory of forced incorporation rather than independence.

Notes  1. For instance, see the formative debate between Goldstein (1986; 1987; 1989a) and Miller (1987; 1988). Also see the earlier work by Goldstein (1971).

78

Chapter 2

  2.  For authoritative histories of this early period of Tibetan origins, see Beckwith (1987), Samuel (1993; 2002), Kapstein (2000), Horlemann (2002a) and Meinert (2002).   3.  On Muslims in Tibet, see Gladney (1991), Cabezon (1997), Lipman (1997), and Dillon (1999).   4.  For an interesting deconstruction of stereotypes of Chinese Muslim unity, see Lipman (1997).   5.  Classic cases of regional feuds were in Central Tibet between Lhasa and Shigatse, or land wars in southern Golok between Darlak and Serthar Counties. Subregional feuds include rangeland feuding between pastoralists. See Yeh (2003b) and Pirie (2005) for details on the latter.   6.  According to Yeh (2009b), the term “Hopaling” referred to the neighborhood where these Muslims lived in Lhasa.   7.  Gladney (1991, 33) cites an estimate that there were about 6,000 “Tibetan Hui” in Lhasa in the mid-1980s, while apparently almost all the “Kashmiri Muslims” fled to India with the Dalai Lama.   8.  On these Muslims, see Gladney (1991, 33-34) and Cabezon (1997, 16, 20, and 25). Gladney (1991, 34) remarked that during field work in Lhasa in the mid1980s these local “Tibetan Hui” did not interact with the Chinese Muslims who started arriving in Tibet from elsewhere in China in the 1980s. I have also noticed that Tibetans in Lhasa clearly differentiate between “old” and “new” Muslims.   9.  On religious networks of authority in Amdo, see Makley (2007). On political authority in Tibet despite the lack of a centralized state, see Samuel (1982). 10.  For instance, see Neitupski (1999; 2002) or Makley (2007) on Labrang, Yi (2005b) on Rebgong, and Goldstein (1989b, 25) on Lhasa. 11.  For examples, see Goldstein (1998), Shakya (1999), Coleman (2002) and van Spengen (2002). 12.  I focus on Amdo because I did most of my fieldwork in this region. Similarly complex histories and interactions can be traced in the southeastern parts of Tibet (Kham), such as between Tibetans, Qiang, Yi, and Han Chinese. 13.  The historic Gansu Province included present-day Ningxia Hui Autonomous Region, Gansu Province, and the northeast corner of Qinghai, known as Huangzhong, located around Xining City and along the Xining River valley between Xining and Lanzhou. See Lipman (1997) for further discussion. 14.  On Labrang, see Nietupski (1999; 2002) and Makley (2007); on Hezhou/ Linxia, see Lipman (1997). 15.  This observation is from my fieldwork, particularly with respect to the textile trade. 16.  Interview with Dr. Ma Cheng Jun (a Salar professor), Xining, June 2004. 17.  See the glossary regarding the Panchen trulku lineage. 18. Lipman (1997, 19), notes that intermarriage between Amdo Tibetans and Muslims often acted as a means of solidifying business relationships, among other considerations. Also see Dillon (1999, 48). 19.  According to some estimates, about half of the Chinese in Qinghai are actually indigenous to the province (Goodman 2004b, 381 and 384). 20.  For the example of Rongwo in Rebgong, see Yi (2005b). 21.  Tibetans and Muslims repeatedly confirmed to me during fieldwork.



The Inception of Modern Development in Tibet under PRC Rule 79

22.  Corroborating with Lipman (1997, 63-65), several Tibetan lamas, officials, scholars, and teachers in Amdo repeatedly told me about the loss of Tibetans to the Muslims over the centuries due to conversion. 23.  On Labrang, see Nietupski (2002). 24.  For instance, armed militias of Amdo Tibetans and Alashan Mongols participated in the defeat of a Salar uprising in Xunhua in 1781 (Lipman 1997, 110), another in 1872 (interview with Dr Ma Cheng Jun, Xining, June 2004), and a Muslim uprising in Xining in 1895 (Lipman 1997, 161). 25.  See Lipman (1997) for many examples of cross-ethnic alliances and interMuslim conflicts. 26.  For examples, see Lipman (1997, 55). 27.  These insights are from interviews during fieldwork in Qinghai from 2003 to 2005. 28. These events are mentioned by Goodman (2004b, 386). My own understanding was built from interviews with a variety of Tibetan, Muslim, and Chinese scholars and elderly locals. 29.  Even in the 2000 census, Tibetans accounted for about one-quarter of the population of Hualong, and the county in fact contains the largest Tibetan population of all the counties in Qinghai (see next chapter). 30.  Interviews with several Tibetan scholars in Chabcha, July 2004. 31. For instance, the Tibetans in Hualong who were converted to Islam during the time of Ma Bufang were labeled Hui rather than Tibetan, even if they had remained culturally and linguistically Tibetan in most other respects. This is ironic given that the official Communist approach to nationality stipulates that nationality should not be determined by religious belief. See Lipman (1997, xx) on the official PRC theory of nationality. 32.  Accounts of these events were collected from interviews with local Tibetan scholars during fieldwork in 2004. The references to CPC misinterpretations and the Salar uprising are from Goodman (2004b, 387), which I confirmed with Dr. Ma Cheng Jun, in Xining, June 2004. 33. Officially, Tibetans in Qinghai numbered 513,000 in 1957, 478,000 in 1959, and then 408,000 in 1963 (QSY 1991, 163). Refugee exoduses might have accounted for some of this shortfall although it is generally considered that only a small share of the 85,000 refugees who left Tibet for India were from outside Central Tibet. Obviously, caution must be used with the statistics from this period given their doubtful accuracy (see the next chapter). 34.  In the midst of the famine in 1960, death rates in Qinghai were surpassed only by death rates in Guizhou and Anhui (Yang 1996, 33). The severity of the famine in Qinghai was related partly to failures of agrarian colonization in highaltitude pastures (Goodman 2004b, 387). We can therefore speculate that much of the excess mortality occurred among the Han migrant colonizers of these pastures, as discussed in the next chapter. 35. If the moderate pre-1957 population growth rate of Tibetans had continued unhindered throughout this period, their population would have approached 600,000 by 1963, indicating a shortfall of 150,000-200,000 Tibetans over this period, or around one-third to two-fifths of the 1957 population. This is a more realistic measure given that the 1963 population includes children born in the midst of the mayhem.

80

Chapter 2

36.  Similarly calculated from QSY (1991, 163). 37.  This is considering a rough estimate of 20 to 30 million deaths in a population of about 600 million, which is a rough estimate cited in the literature. For instance, see Peng (1987). 38.  See an excellent discussion of this with regard to demographic and health data in Childs (2008). 39.  Given that price indices going back to the 1950s are not available for China, the yuan value here is converted into US dollars according to the 1959 exchange rate, indexed according to the consumer price index of the US dollar up to 2000, and then converted back to yuan at the 2000 exchange rate. Ideally, a price index for China would be used, which would also possibly result in a higher estimate. 40.  See work on this by Knight and Song (1999) or Huang et al (2006). 41.  For instance, see a discussion of Japan and the USSR in Lewis (1954, 174-75). 42.  See chapter 5 for rural income data from 1990 onward. 43.  See chapter eight for a discussion of the “wool wars” in western China in the 1980s that were related to these pricing and distribution dynamics. 44. This being said, prices vary enormously depending on the quality of the wool. For instance, prices in 2004 ranged from 18 yuan per kilogram for yak wool from locations above 3,800 meters, 22 yuan per kilogram for sheep wool from desert locations in Western Qinghai, or 200 yuan per kilogram of cashmere. 45. See Winkler (2008a; 2008b) for some excellent work on this booming “fungal” economy. 46. This is assuming that rural price inflation rates in the TAR were approximately the same as the national rates from 1982 to 1990, given that the CPI started to be reported for the TAR only from 1991 onward. This is a plausible assumption given that very similar rates of rural price inflation—in the magnitude of around 350 to 400 percent between 1985 and 2006—were reported in most other western provinces and that the TAR had among the highest rates of inflation in China in the 1990s. 47.  Note that the data used in figure 2.1, which are compiled from the national statistical yearbooks, show a slight discrepancy with the data used in figure 2.2, which are compiled from the TAR statistical yearbook. The TAR was still slightly recessionary in 1995 according the national data, whereas there was a slight recovery in 1995 according to the TAR data. 48.  For instance, see Yang (1997) for a detailed discussion of the debates that were raging across China over these issues in the early to mid-1990s. 49.  Ideally we would use a GDP deflator for this exercise, although this index is not publicly available in the statistical sources. 50. It is possible that price inflation in the TAR might have been slightly lower—and thus the depth of real recession as recorded in figure 4.1 slightly overestimated—given that price controls might have been maintained longer and with more vigor in the TAR than in other western provinces, although this is not possible to confirm with the available data. Notably, in the 1980s, prices were liberalized gradually in the east while controlled in the west, and price control was probably strongest in the TAR. Following price liberalization in the early 1990s, inflation was higher in the west than in the east as western prices adjusted toward a national level. With liberalized prices, adjustments in the TAR were accentuated by several other



The Inception of Modern Development in Tibet under PRC Rule 81

inflationary factors, such as the monetary impact of fiscal transfers, the transportation costs of importing goods to the TAR, and the ending of certain subsidies on various goods, hence explaining the particularly high inflation rates in the TAR in 1993 and 1994 (Fischer 2005a, 94). Price inflation has generally continued to be higher in the western provinces in the 2000s, albeit still from a lower cost of living, except in the case of the TAR, which generally has a much higher cost of living (in urban areas) than other western provinces (as confirmed by me during fieldwork). 51.  This reading of Gramsci’s notion of hegemony is from Arrighi (1994, 27-29). I do not delve further into the theme of hegemony in this book. 52.  I collected numerous accounts of ongoing nationalist activism (secessionist and non-secessionist) during fieldwork in Kham and Amdo from 2003 to 2007. 53.  See Fischer (2012a), in addition to the other essays edited by McGranahan and Litzinger (2012). 54.  See TIN (1999a; 1999b) regarding earlier patriotic education campaigns. 55.  See TIN (2001). This demolition involved the expulsion of several thousand monks and nuns, including an estimated one thousand Han Chinese monks, and the destruction of about one thousand dwellings. 56.  The selection of the Eleventh Panchen Lama in 1995 has been the source of intense dispute between the Chinese government and the Tibetan Government in Exile. See Hilton (1999). 57.  These arguments were repeated to me on numerous occasions during fieldwork by a wide variety of Chinese and by some educated Tibetans. They are also often recounted in the official press or else in many academic papers by Chinese scholars, echoing the important work by Wang and Bai (1991 [1986]), which will be treated in chapters 4 and 7. Also see the work by Yeh (2007; 2009b) on the cultural politics of development in Lhasa, and Zenz (2013) for a critique of Yeh (2007). 58.  Conversely, see the excellent discussion by Zenz (2013) on how traditionalist Tibetan discourses have turned this developmentalist logic on its head by portraying backwardness as the failure to study Tibetan language and culture, especially Buddhist culture. Assimilated Tibetans would possess poor levels of “development” thus conceived. 59.  Dreyfus (2002, 48) comments cautiously on this, although he avoids mentioning the sectarian violence that he discusses elsewhere (1998) as well as the undemocratic propensities of the trulku institution. 60. On education in Tibet, see Postiglione (1992; 2008; 2009), Bass (1998; 2005), Iredale et al. (2001), Postiglione et al. (2004; 2006a; 2006b), Bangsbo (2008), Iselin (2011), and Zenz (2010; 2013). 61.  On the rural health system in China, see Liu and Bloom (2002), Bloom and Fang (2003), and Shaoguang (2004).

3 Population Foundations of Marginalization in Tibet

Following from the general historical discussion of the previous chapter, this chapter delves more deeply into the population foundations of modern development in Tibet. This highly politicized and sensitive issue needs to be placed within a broader understanding of modern population transitions as a means to decipher those aspects of population change in Tibet that are, in fact, quite typical to other parts of the developing world, versus those aspects that might be considered quite exceptional to the case of Tibet. As elsewhere in the world, population transitions have been fundamental to the appearance of modernity in Tibet. In essence, these population transitions include demographic transitions, urbanization, and increased non-Tibetan in-migration from other parts of China (largely Han Chinese and, to a lesser extent, Chinese Muslims). However, the fact that these population transitions started around the same time as Communist rule in 1950 has led to their conflation with the related but distinct issue of occupation. As a result, some of the most vigorous condemnations as well as defenses of PRC rule in Tibet have come to be framed in terms of population. The reigning view in the Tibetan exile community—as well as among most Western observers and “Tibet support groups” (TSGs) and most Tibetans and even some Chinese whom I interviewed in Tibet—alleges that Han Chinese and Chinese Muslim migrants are quite simply overrunning the existing Tibetan population in Tibet. This is typically referred to as “population invasion,” “population swamping,” “population transfer,” “demographic invasion,” or “demographic aggression.” For instance, following the outbreak of widespread Tibetan protests in March 2008, the 83

84

Chapter 3

Dalai Lama reasserted his view that China’s policy of resettling Chinese migrants in Tibet amounts to “demographic aggression” and even “cultural genocide,” referring to the disappearance of Tibetans as an identifiable cultural group due to sinicization (see VOA 2008).1 Similar views were echoed by Tibet advocacy organizations in exile and the West, in particular in the 2012 report on “cultural genocide” by the International Campaign for Tibet (ICT 2012b). Indeed, 2007 was the first year ever when tourists (mostly domestic Han Chinese) outnumbered the local population in the Tibet Autonomous Region (TAR). While this must be qualified by the fact that most of such tourists visit only a few locations in the TAR, such as Lhasa, on package tours lasting only a few days, the numbers are nonetheless indicative of the sheer volume of Han Chinese movement in and out of this region, including Han migrants to service the tourist traffic and the more general government-subsidized economic boom. Parallel to this, many Tibetans fear that Beijing has been attempting to limit the growth of the Tibetan population through stringent family-planning measures, including forced abortions and sterilization programs. The concern is that Tibet will soon come to resemble Inner Mongolia, where ethnic Mongolians accounted for 17 percent of the population in the 2000 census, or Xinjiang, where Han Chinese accounted for 40 percent of the population, up from around 6 percent in the early 1950s. Many exiled Tibetan leaders claim that Tibetans might already be a minority in many Tibetan areas.2 At the extreme, these concerns led to accusations of genocide in the past, which have since been tempered to the ongoing accusations of cultural genocide. Conversely, Beijing denies that non-Tibetan migrants have been overrunning the local population. It regularly points out that ethnic Tibetans accounted for about 93 percent of the resident population of the TAR in the 2000 census.3 It continues to assert this position, despite ten years of evidently high levels of in-migration and booming tourism since 2000. Tidbits of evidence gleaned from the 2010 national census were reported by Xinhua (2011a), indicating that a little more than 90 percent of the “permanent resident”4 population in the TAR (numbered at just over three million) were “native Tibetans” (or a total number of 2.716 million people) and that just over 8 percent were Han.5 This 2 percent increase in the Han share compared to the 2000 census is hardly evidence of “swamping.” Instead, Beijing argues that migration to Tibet has mostly helped Tibetans. After all, migrants have been the ushers of government policies that have improved the quantity and quality of life of Tibetans since the 1950s, as demonstrated by rapid population growth (versus stagnation for centuries prior to 1950), a doubling of life expectancy, and impressive improvements in a variety of other human development indicators since the 1950s.6 From this perspective, it argues that the disparities of regional development be-



Population Foundations of Marginalization in Tibet 85

tween Tibet and other regions of China are due to the backwardness and remoteness of Tibet, which Beijing has valiantly invested enormous efforts to overcome with the help of migrants and some rich coastal provinces.7 The government thus credits itself for benevolently uplifting the backward and “feudal” social conditions of Tibet.8 As discussed in depth by Childs (2008, 214-31), while the emphasis on population growth within this discourse is ironic in light of government objectives to limit population growth elsewhere in China, it must be understood in light of exile accusations of genocide or population control. Both views are partially valid. In the case of population invasion, it is only in the main strategic cities and towns of Tibet where it can be plausibly argued that Han are outnumbering Tibetans. Ironically, the perception of population swamping is essentially an urban-centric assessment of the changing ethnic composition of Tibet, even though the Tibetan areas have been, up until recently, some of the most rural in China. While the perception of population swamping is arguably misconceived—and, hence, easy for Beijing and scholars such as Sautman (2008) to refute—it nonetheless captures an important sense of marginalization experienced by both common and elite Tibetans, which in turn reinforces a sense of Tibetan nationalism by providing a basis for cross-class perceptions and discourses to emerge. In terms of disparities, it is also true that ethnic inequality in Tibet has been strongly associated with urban-rural inequality given that, up until recently, most Tibetans have been rural (although this has been changing rapidly since 2000), whereas most Han in Tibet are urban. As discussed in the introduction, this observation leads some scholars, such as Sautman and Eng (2001) or Hu and Salazar (2008), to claim that the apparent ethnic bias of development in Tibet is mostly a reflection of an urban bias and, as such, it is no different from similar development dilemmas elsewhere in China, except that rural Tibet happens to be populated mostly by Tibetans. There are a variety of problems with this argument; most importantly, it distracts attention from ethnic exclusions occurring within the urban areas and intra-urban inequalities that are among the highest in China. In an attempt to clarify these politicized debates on population and development in Tibet, this chapter seeks to refocus attention on the issue of urban exclusionary dynamics. There is no doubt that Tibetans have been in demographic transition since the 1950s or 1960s, like other nationalities in China or, indeed, in Asia, and this demographic transition has been conjunctural with Chinese rule. Urbanization in turn sets the stage to understand the crux of modern ethnic tensions in the Tibetan areas. In the process of urbanization, there might be a visual impression that the more urbanized or economically successful groups (Han or Muslims) are becoming more dominant even while they are only

86

Chapter 3

maintaining or even losing their share in the overall population. In this way, perceptions may be deceptive even while they fuel resentment and signal other underlying dilemmas of considerable gravity, such as urban employment exclusions occurring in key moments of demographic and economic transition for the local population. However, this understanding of urbanization does not necessarily accord with the argument that perceived ethnic biases are merely reflections of spatial biases. There is a profound spatiality involved, such as along the lines discussed by Yeh (2013), but not in simplistic urban-rural binary terms. Rather, Han Chinese dominance in the urban areas of Tibet is crucial because urban areas are the portals of political and economic power and thereby serve as the basis from which new stratifying and subordinating tendencies become articulated and are emanated in the rapidly changing regional economy. In particular, educational, political, and socio-cultural biases operating in urban employment, in a context where illiteracy is still very high among even permanent urban Tibetan residents, are arguably the dominant factors producing marginalization in the contemporary development of Tibet. In this light, it is not correct to say that the urban bias is the cause of discrimination and marginalization in Tibet, although it definitely structures the way that discrimination and marginalization manifest, in contrast to other minority nationality regions in China, such as Xinjiang or Inner Mongolia, where rural populations are more ethnically mixed and minority populations possess higher levels of schooling. In laying out this argument, the chapter does not offer a comprehensive demographic and population survey of Tibet, which already exists in the literature from both sides of the debate.9 Rather, it analyzes macro-level demographic, spatial and migration trends in conjunction with development trends, drawing from both official population statistics (mainly the 2000 census) and fieldwork, as described in the introduction. The first section discusses various issues raised by the debate: the demographic transition in Tibet; the precise spatial delineation of Tibet; and the spatial distributions of ethnic populations within this delineated territory. The second section examines the balance between migration and local demographics, followed by a reflection on urban employment exclusion. The conclusion contends with the proposition that ethnic inequality is a reflection of an urban bias transposed on the rurality of Tibet, arguing that this logic distracts attention from power relations rooted in urban areas.

Clarifying the Population Debate Given that population has come to stand as a key symbol in the politicized debate over Tibet, many of the assertions on population made by



Population Foundations of Marginalization in Tibet 87

both sides tend to become blurred with political posturing and ideology, requiring that we put all of the arguments into context. However, one terminological clarification is necessary before starting. The TAR usually receives most of the attention in these debates, particularly from the Chinese side given that Beijing insists that the “Tibet Question” is only about the TAR and not the other Tibetan areas in China, despite the latter’s accounting for almost half of the Tibetan areas and more than half of the Tibetan population in China. The dominant focus of this and several of the following chapters is also on the TAR although this is not meant as acquiescence to Beijing’s view. Rather, it is for practical reasons given that official population surveys of the TAR represent a mostly Tibetan experience, unlike Qinghai, the province with the next-highest proportion of Tibetans in its population, where Han Chinese nonetheless dominate the population and thus the survey results. As argued in the introduction, the TAR can serve as a good rough proxy to represent all Tibetan areas in terms of broad structural variables in comparison to the rest of China, especially with regard to population data (less so with regard to economic data, as discussed in the following chapters). Population Growth and Demographic Transition Life expectancy in the TAR officially doubled to sixty-four years between 1951 and 2000. The Tibetan population of the TAR nearly doubled, reaching 2.43 million in 2000.10 The Tibetan areas outside the TAR experienced similar developments. Overall, the total Tibetan population in China reached 5.42 million according to the 2000 census, including 1.27 million Tibetans in Sichuan, 1.09 million in Qinghai, 0.44 million in Gansu, and 0.13 million in Yunnan (Tabulation 2002, table 2-5). It should be noted that the changes since 1951 are at best estimates or even guesswork given that most of the TAR and many remote Tibetan areas outside the TAR were not directly enumerated in either the 1953 or the 1964 censuses. As a result, the 1982 census is the first broadly accurate record that we have available on the Tibetan population.11 While the increase in life expectancy is commendable, it was about the same as the average for all developing countries in 1998, at sixty-five years according to the World Bank (2002a). Similar advances took place in neighboring Himalayan countries. For instance, life expectancy in Bhutan was estimated by the United Nations at sixty-two years for males and 64.5 years for females in the period of 2000 to 2005. Indeed, the comparison with Bhutan is interesting because, with conditions similar to Tibet, comparable health indicators (official in both cases) have been achieved under an entirely different political framework, albeit Bhutan is heavily subsidized by India, just as the TAR is by Beijing. In India the respective rates were

88

Chapter 3

63.2 and 64.6 years. Even in conflict-ridden Nepal, one of the most impoverished countries in Asia, the rates were 60.1 and 59.6 years (UN 2002). Similarly, rapid population growth itself is a consequence of the late onset of demographic transition, as observed in all developing countries over the same period, regardless of the quality of development.12 Tibetans have clearly entered the demographic transition, contrary to simplistic explanations sometimes given for the poor development of Tibet.13 This has been well established in the literature on Tibetan demography. Death rates probably started to fall in the 1950s due to the expansion of modern public health care in Tibetan areas following their absorption into the PRC, although the exact magnitude and timing is difficult to ascertain given poor data for this period.14 Improvements were also probably interrupted by the periods of uprising, counterinsurgency, famine, and social upheaval in the late 1950s and early 1960s.15 Nonetheless, by the late 1960s, sustained improvements in mortality and life expectancy were evident across Tibet, albeit with a significant lag behind China.16 Total fertility rates (TFRs; hereafter simply “fertility rates”) in the TAR officially peaked at around 5.5 births per woman in 1968, fell below 5 in the 1970s, jumped to a lower peak of about 5 in the early 1980s, and fell gradually after 1982 and then sharply from the late 1980s onward (Zhang and Zhang 1994, 54; Childs 2008, 202).17 The fertility rate in the TAR was 4.2 in the 1990 census and 1.9 in the 2000 census (Childs et al. 2005, 343). In Qinghai, fertility rates in the more central Tibetan autonomous prefectures were below replacement in the 2000 census, for example, 1.7 for Malho (Ch. Huangnan) and Tsolho (Ch. Hainan), and 1.8 for Tsojang (Ch. Haibei), while they were above replacement in only the two most remote prefectures, namely, 2.3 in Golok and 2.7 in Yushu (Qinghai Tabulation 2002, table L.6-1). These findings are further supported by data from the 2000 census on the age structure of all Tibetans in China (figure 3.1). These are almost identical to the equivalent data for the TAR analyzed in Childs et al. (2005, 341), except that the number of Tibetans is more than doubled, indicating that the two populations inside and outside the TAR are almost identical in this respect. In both sets of data, the largest age cohort was between ten and fourteen years of age and the size of the younger cohorts tapered off, indicating that demographic momentum reached its peak among Tibetans in the middle to late 1980s, both inside and outside the TAR, and has since been slowing. In order to assuage concerns that such official data have been concocted or manipulated, the fertility transition is also evident in the field. For instance, declining fertility rates were reflected in a demographic survey of 13 farming villages in the TAR by Goldstein et al. (2002), where fertility rates dropped from 6.4 in 1986 to under 2 in 1998 (calculated in Childs et al. 2005, 343). My own qualitative field observations in the Tibetan areas of Qinghai suggest similar trends. In other words, all indications suggest that



Population Foundations of Marginalization in Tibet 89

Figure 3.1: Tibetan population of China, all provinces, age and sex, 2000 census

Source: Constructed from CPSY 2003. From Fischer (2008d), reused with permission from Population and Development Review.

Tibetans inside and outside the TAR are currently well advanced into the demographic transition, to an extent that total fertility rates are now below replacement levels in all but the most remote areas. Natural population increase rates (NPIR; see figure 3.2) clarify these processes of demographic transition in comparison to the rest of China (Note that “natural population increase” is a demographic term that designates population changes due only to births minus deaths; it does not include net migration, the other factor causing population growth or shrinking. The rate is typically measured per one thousand people, that is, crude births minus crude deaths per one thousand people in a population.). The TAR (mostly Tibetan) had an NPIR that was considerably slower than China as a whole (mostly Han) up to the early 1970s, at about the same rate in the late 1970s and 1980s, and considerably faster since 1989. The similarity of the TAR and China in the late 1970s and 1980s represents the fact that lower fertility rates among the Han during these years (compared to the rates cited above for Tibetans) were counterbalanced by greater population momentum from their baby boom in the 1960s that followed the famine, when the rate of natural population increase in China hit its all-time peak of around 2.5 to 3 percent a year. The TAR apparently missed out on this post-famine baby boom. By the 1970s, fertility, birthrates, and rates of natural popula-

90

Chapter 3

tion increase were all decelerating rapidly in China as a result of the family planning policies that were implemented with vigor from 1970 onward. Rates of natural population increase for China then increased in the early 1980s (simultaneous to the implementation of the one-child policy from 1979 onward due to population momentum), reaching a lower peak in 1987 of about 1.7 percent a year, and gradually declined thereafter, to just under 0.5 percent a year by 2010. In the TAR, higher fertility rates into the 1970s and 1980s appear to have been counterbalanced by less population momentum, perhaps due to the fact that the disruption and upheaval from the late 1950s and early 1960s resulted in a much smaller reproductive cohort in that generation relative to the rest of China. As noted in the previous chapter, the loss of life in the years from 1957 to 1963 was apparently much greater in Tibetan areas— especially outside the TAR—than elsewhere in China. The refugee exodus from the TAR to Nepal and India in 1959-1960 would have also amounted to about 5 to 10 percent of the TAR population depending on which estimates are used for both refugees and population. Moreover, given the late entry of the TAR into the mainstream of Maoist upheaval, the consequent social disruption probably lagged behind the rest of China as well, such as collectivization, which was implemented on a large scale only by the end of the 1960s and resulted in “ecological disaster, and widespread hunger” in the 1970s (Dreyer 2003, 415). However, according to official data from 1965, the birthrates, death rates, and natural population increase rates in the TAR were all apparently quite low in the 1960s, strangely much lower than those in the rest of China. It is not clear why these rates for the TAR were so low. These data are not shown in figure 3.2 because they are nonsensical and of dubious accuracy (and are possibly related to propaganda purposes related to the establishment of the TAR in 1965).18 Discounting these 1965 data, (official) birthrates were otherwise substantially lower in the TAR than in China up to the early 1970s—and death rates higher—resulting in a slower rate of natural population increase. Given that fertility rates were higher in the TAR than in China during the 1970s (Fertility rates in China fell rapidly in the 1970s), the lower birthrates in the TAR must be reflective of a smaller cohort of women in childbearing years, suggesting a relatively much greater loss of life in the TAR than in China in the late 1950s and 1960s, as mentioned above. The sharp drop in population of the cohort aged 40–44 (born from 1956 to 1960) compared to the cohort aged 35–39 (born from 1961 to 1965), as shown in figure 3.1 with respect to the 2000 census, is definitely suggestive of this possibility. In any event, Chinese birthrates fell below the TAR rates by the mid-1970s, whereas the TAR rates more or less stayed at the same level up until the late 1990s (according to official data), resulting in an annual rate of natural population increase for the TAR at around 1.6 percent throughout this period, with peaks in

Sources: CSY (2011, table 3-3 and 3-4) and equivalent in previous yearbooks; CSY (1999, Table 4-2); TSY (2004, Table 3-2).

Figure 3.2: Birth, death, and natural increase rates, China and TAR, 1970-2010

92

Chapter 3

1983 and then again in 1990. Birthrates and, hence, natural population increase rate, started to fall sharply from 2000 onward, thereby catching up with China following a lag of about ten years. Whether the sudden decline from 2000 onward represents a policy change is not clear; it is more probable that the abruptness of the change represents the input of much more accurate data from the 2000 census and subsequent population surveys. In that case, the trend of gradually falling birthrates and natural population increase rates might have happened slightly sooner, in tandem with fertility rates as discussed above. As discussed in much of the contemporary demographic literature, these demographic differentials between Tibetans and Han are explained by a variety of structural, sociological, and policy factors, rather than any intrinsic or “backward” demographic behavior on the part of the Tibetans, as is often implied by some Chinese demographers.19 Family planning policies usually receive most of the attention, although from the perspective of the politicized debates on population in Tibet, their influence is in fact counterintuitive given that birth limits are leniently applied toward minorities. As discussed above, fertility rates rose in the TAR at the beginning of the reform period, even while they continued to fall in China due to the implementation of the one-child policy. Even into the 2000s, the official birth limit for minorities in China has been two children rather than the onechild limit applied to Han Chinese, and even this two-child limit has been applied only to urban dwellers in many parts of Tibet.20 These limits have nonetheless contributed to fertility declines, particularly with increased enforcement since the mid-1990s, as analyzed by Goldstein et al. (2002) and Childs et al. (2005, 347). As a result, fertility transition among Tibetans has accelerated since the late 1980s, albeit with a lag behind the Han Chinese and thus more rapid rates of natural population increase. These insights are often greeted with incredulity by exiled Tibetans given their own allegations that the PRC is practicing forced sterilizations and abortions on Tibetan women. Indeed, falling fertility might be taken as evidence of overzealous family planning, although many of these allegations conflate birth limits with forced sterilizations and abortions, whereas the former do not necessarily imply the latter.21 The latter were investigated by Goldstein et al. (2002), who concluded that their fertility data from thirteen villages in the TAR suggest no general program of forced birth control (2002, 25). However, many of the allegations of abusive forced birth control refer to Tibetan areas outside the TAR, albeit mostly with reference to the 1980s and early 1990s. In this regard, my own qualitative field insights in the Tibetan areas of Qinghai tend to support the conclusion that cases of forced sterilizations and abortions do not represent a generalized policy approach but were probably related to events in specific localities or during specific brief periods of policy zealousness.22 Moreover, the overall rates of



Population Foundations of Marginalization in Tibet 93

natural population increase for Qinghai have been consistently similar to those of the TAR, both the highest in the country since the 1990s alongside those of Xinjiang and Ningxia, and, as noted above, fertility rates in the Tibetan autonomous prefectures of Qinghai were similar to those of the TAR in the 2000 census. It is therefore likely that family planning has also been relatively lenient in many minority areas of Qinghai, similar to the findings of Goldstein et al. mentioned above, albeit leniency varies from locality to locality. For instance, Tibetans whom I interviewed in 2004 from Tsolho (Ch. Hainan) Prefecture in Qinghai or from some Tibetan counties in Gansu reported very strict implementation in these areas, with fines of 1,000 yuan for a third child, and 3,000 yuan and imprisonment for a fourth. Meanwhile, Tibetans interviewed in nearby Chentsa (Ch. Jianza) County in Qinghai did not necessarily report any penalties besides token fines of 100 yuan for a fifth child. Birth controls seemed very lax or even ignored in many of the remote pastoral areas of Qinghai, such as in Golok and Yushu Prefectures, where large young families with five or more children were still common, apparently with little or no penalty. The point here is not to make a judgment about the abusiveness of Chinese family planning policies in Tibet but, rather, that there is nothing surprising or sinister in the demographic trends of the Tibetan areas. Relative to the rest of China, slower fertility and crude birthrate declines since the 1980s have been consistent with a more lenient family planning regime. Specific cases of abusive family planning methods do not seem to have had much of an impact on these broader trends,23 although, as noted above, stricter implementation of birth limits in the 1990s definitely bolstered the speed of the decline. Rather, per recent theoretical work on demographic transitions, sustained mortality decline would seem to be the key remote causal factor determining fertility transitions (for instance, see Dyson 2001). Although more research on this issue is needed in the case of Tibet, we can plausibly suggest that both the timing of the transition in Tibet and its subsequent lagging with the rest of China up until the late 1990s were both due to mortality declines in the 1960s and 1970s that, although sustained, were much inferior to the rest of China. While birth control also definitely influenced trends, it is useful to reflect more broadly that the sharpest reductions in fertility in China were witnessed in the 1970s before the introduction of the one-child policy (Lee and Wang 1999; Hussain 2002a). Fundamentally, this was probably due to sharp mortality decline from 1950 to 1970 (in comparison to India, for instance), rather than the coerciveness of family planning policies (see Hussain et al. 2006). Moreover, along with family planning measures, other proximate factors would have moderated the timing and speed of the fertility transition in Tibet once unleashed, such as various social and

94

Chapter 3

economic “multi-phasic responses” to population growth (per Davis 1963), stagnating agrarian economies, increasing costs of raising children, rising education levels among Tibetan women, increasing access to and knowledge of contraception, and urbanization.24 Indeed, these insights from modern demographic theory help to explain the finding of Childs et al. (2005) that demographic trends among Tibetans in the TAR and Tibetan refugees in South Asia are remarkably similar in both timing and magnitude despite the fact the two groups live under completely different political, economic, and social conditions, and that the exile community is generally pro-natal in its attitude toward family planning. In other words, the seemingly contradictory perspectives on population in Tibet can in fact be understood as various aspects of these transitional demographic processes. Minority or Majority? Spatial Delineations and Distributions Population growth aside, the exile assertions that Tibetans are becoming a minority in their homeland, or have already become a minority in certain regions, must also be taken with a grain of salt. There are two spatial issues that must be considered. One is the precise delineation of “Tibet” and the other is the rural-urban distribution of populations within these delineated areas. In the first case, the delineation of Tibet in the borderlands where eastern Tibet meets western China proper is critical to any ethnic population count because of the sudden changes in topography, ethnicity, and population density that occur within frontier counties. Key in this regard is the northeast corner of Qinghai, that is, Xining and Haidong districts. These two districts account for about 2.8 percent of the provincial land area, versus the five Tibetan autonomous prefectures (APs) and one Tibetan-Mongolian AP that account for the remaining 97.2 percent.25 The exile government usually delineates these two districts as part of their definition of Tibet, reflecting the fact that much of it was historically composed of Tibetan farming and pastoral areas, albeit intermixed with Muslim farming areas and some Han Chinese garrison settlements, as discussed in the previous chapter. Indeed, the current Dalai Lama was born in this region, albeit it was controlled by the Muslim warlord Ma Bufang at the time. Even today, both Xining and Haidong contain a patchwork of Tibetan counties, townships, and religious centers, and they accounted for about one-quarter of the total Tibetan population of Qinghai in 2000 (calculated from Qinghai Tabulation 2002). Nonetheless, regardless of the legitimacy of historical claims, both districts were the main destinations of Chinese migration to the province, particularly following the creation of Qinghai in 1928 and especially during the 1950s and 1960s. As a result, today these two districts are relatively urbanized and account for two-thirds of the Qinghai population, largely Han with a Muslim minority.



Population Foundations of Marginalization in Tibet 95

Outside of this core northeast area of Qinghai, the bulk of the remaining Han population in Qinghai is concentrated in several strategic pockets of Chinese interest, such as the mining towns of northern and western Qinghai (Haibei TAP and Haixi TMAP). Both prefectures are very arid and largely composed of desert as well as a handsome supply of minerals and hydrocarbons. Haixi in particular includes the mineral-rich Tsaidam (Ch. Qaidamu) Basin, which became infamous in 1999 due to the “Qinghai Project” of the World Bank that came under a concerted attack by the Tibet Exile Government and Tibet Support Groups for supporting population transfer policies. Mineral resource development in these areas has resulted in the rapid expansion of several mining towns since the 1950s, particularly Germo (Ch. Golmud), populated mostly by Han, whose population dwarfs the indigenous and extremely sparse populations of pastoral Tibetans and Mongolians. Leaving aside the historically contested northeast corner and the highly mineralized north and west of the province, the rest of the province is essentially Tibetan. Tibetans accounted for 66 percent of the population of Malho TAP in the 2000 census, 63 percent of Tsolho TAP, 92 percent of Golok TAP, and 97 percent of Yushu TAP. The closely related Mongolian and Mongour (Ch. Tu) minorities in Qinghai, who are Tibetan Buddhist and mostly rural, accounted for an additional 17.5 percent of Malho TAP, giving a Tibetan-Buddhist total of 84 percent in that prefecture, even despite a substantial presence of Muslims in Chentsa County. It is clear that indigenous Tibetan Buddhists are a large majority in these prefectures. In Sichuan, Tibetans accounted for 54 percent of Ngawa (Ch. Aba) TQAP and the closely related indigenous Qiang minority for another 18 percent, while Tibetans accounted for 78 percent of Kardze (Ch. Ganzi) TAP. In Gansu, Tibetans accounted for 51 percent of Gannan TAP. The Han share in Gannan was close to the Tibetan share, at 42 percent, reflecting the proximity of the prefecture to Lanzhou, the capital city of Gansu and major urban center. It was only in Dechen (Ch. Diqing) TAP in Yunnan where Tibetans were in a definitive minority, although they were also the largest nationality, at 33 percent of the prefecture total, with the Lisu minority in a close second at 28 percent (all data calculated from county-level tabulations). Debates on population shares rarely distinguish the point that, in most cases, indigenous minorities remain the majority in these prefectures, overwhelmingly so in the rural areas, similar to the TAR. Precise delineation therefore becomes critical for a proper evaluation of the effective Chinese population within the definitively Tibetan areas. The exercise is complicated by the fact that borderland counties often have mixed zones within a single county. In such cases, a higher population density of non-Tibetans in the lowlands of a county can appear to overwhelm the lower population density in the definitively Tibetan high-

96

Chapter 3

lands, even though the two zones usually remain quite segregated. This scenario applies to much of Haidong in Qinghai as well as to the essentially Han-populated lowlands of the counties in eastern Kardze TAP and Ngawa TQAP in Sichuan. For instance, the epicenter of the devastating earthquake in Sichuan on 12 May 2008 was in Wenchuan County, located in Ngawa TQAP, although the casualties were mostly Han Chinese in the lowlands of this county.26 Considered in this way, Tibetans are definitely not a minority in most Tibetan areas. If anything, they are a large majority, albeit one that is predominantly rural and thus not very visible. This leads us to the second spatial issue. It is clear from the 2000 census that the main population schisms boil down to sharp differences in urbanization rates rather than differences in population shares. Tibetans were overwhelmingly rural in all of the five Chinese provinces that incorporate Tibetan areas, with 87.2 percent living in rural areas overall (see table 3.1). Tibetan urbanization rates ranged from 8.6 percent in Qinghai, 9.1 percent in Gansu, and 10.5 percent in Sichuan to 15.2 percent in the TAR and 20 percent in Yunnan. In contrast, the urbanization rate among the recorded Han was 79.5 percent in the TAR and 44.7 percent in Qinghai, both rates higher than the national average of 36.9 percent. The Hui Muslims were 79.7 percent urban in the TAR, where their presence is small, and 29.7 percent urban in Qinghai, where they account for about a sixth of the population. The Qinghai Hui were less urbanized than the Qinghai Han but much more than the Qinghai Tibetans. The extremely high rates among the Han and Hui in the TAR reflect their lack of indigenous base in the rural areas, which is discussed further in the next section. Furthermore, the Han and Muslim presence is more concentrated in cities than in towns, whereas urban Tibetans are more concentrated in towns than in cities, particularly outside the TAR. This is shown in table 3.1, which focuses on the TAR, Qinghai, and Sichuan (accounting for 88 percent of Tibetans in China). The town-city dimension reflects the fact that most urban jurisdictions in the Tibetan areas are classified as towns, with Lhasa and Shigatse in the TAR being the notable exceptions. It is also probable that rural Tibetans tend to migrate first to local towns rather than more distant and expensive cities, except in cases related to tertiary education, which is concentrated in the capital cities of each province. Yeh and Henderson (2008, 25) note that this pattern also exists with Lhasa Municipality itself, where Han in-migrants were concentrated mostly in the city center according to the 2000 census data, whereas their numbers were very limited in the urban towns and rural townships of the municipality. It is clear from table 3.1 that Tibetans have remained predominantly rural. Rates of change are not shown; temporal comparisons of urbanization rates below the national level are rendered very difficult if not meaningless



Population Foundations of Marginalization in Tibet 97

Table 3.1: D  istribution of population by selected nationality in the Tibet Autonomous Region (TAR), Qinghai and Sichuan Provinces, 2000 census Percent of group located in:

Share of total population in:

Rural

Town

Rural

Total population

80.6%

11.1%

8.3%

--

--

--

Tibetans

84.8%

9.4%

5.8%

97.6%

78.3%

65.0%

City

Town

City

TAR

Han

20.5%

35.5%

44.0%

1.5%

19.4%

32.1%

Hui Muslims

20.3%

25.6%

54.0%

0.1%

0.8%

2.2%

67.7%

11.7%

20.7%

--

--

--

Qinghai Total population Tibetans (+M & Mong)

91.4%

6.7%

1.9%

37.6%

17.4%

3.5%

Han

55.3%

13.6%

31.1%

44.2%

62.9%

81.3%

Muslims (Hui & Salar)

70.3%

12.8%

17.0%

18.1%

19.1%

14.3%

a

Sichuan Total population

72.9%

12.3%

14.8%

--

--

--

Tibetan (+Qiang)b

89.5%

9.5%

0.9%

2.3%

1.5%

0.1%

Han

72.0%

12.5%

15.5%

93.8%

97.1%

99.0%

Yi

94.1%

4.4%

1.5%

3.3%

0.9%

0.3%

Tibetans in China

87.2%

8.7%

4.1%

--

--

--

Source: calculated from Tabulation (2002, tables 1-6, 1-6a, 1-6b, and 1-6c). a: row measures urbanization rates of Tibetans only, but includes the Mongours (Ch. Tu) and Mongolians in the measure of shares in order to evaluate the weight of indigenous Tibeto-Buddhists in rural, town, and city. b: row includes the Qiang in shares; the Qiang are only 2 percent more urbanized than Sichuan Tibetans.

given that urban definitions are quite different in each of the five censuses in China (see Yixing and Ma 2003). Indeed, Yeh and Henderson (2008, 22) calculate that 30 percent of the reported “urban” population growth in Lhasa Prefecture between the 1990 and 2000 censuses was simply a matter of the reclassification of existing township populations as urban, mostly in 1999, through the administrative conversion of rural townships into urban towns. However, with these qualifications in mind, rapid urban growth in the Tibetan areas up to 2000 was disproportionately filled by Han (and to a lesser extent by Muslims in the case of Qinghai), relative to their overall population share. Current inter-provincial migration (or intra-provincial in the case of Tibetan areas outside the TAR) accentuates these differences given that the destination of most Han and Muslim migrants to the Tibetan areas is urban, as discussed in the next section. Although the Han and Muslim population shares may be small in these Tibetan areas, this share is concentrated in the urban areas, precisely where it is most visible.

98

Chapter 3

Moreover, minority dominance in the rural areas is notable. The combined total of minority nationalities in Qinghai (predominantly Tibetan and Hui Muslims) gives them a majority in the rural areas, even before isolating minority areas from the core Han areas. Given demographic momentum among both Tibetans and Muslims, this pattern is likely to continue. In the TAR, the contrast between the rurality of Tibetans and the urbanity of the Han and Hui is extreme. Therefore, the key issue is not whether the population balance has shifted toward Tibetans or outsiders, but that the latter have dominated urbanization, as demonstrated by Yeh and Henderson (2008) in the case of Lhasa Municipality. Overall, it might actually be the case that Han population shares in Tibetan areas outside the TAR have been decreasing since the 1980s. This is not the case in the TAR, where the share has been definitely increasing, although from very low levels (reportedly from 6 to 8 percent, as noted in the introduction). The reasons for this apparent incongruity with exile claims are explored in the following section.

Population Shares Reconsidered Regardless of misinterpretations, the exile (and predominant Tibetan) view of population invasion harbors legitimate concerns and the view is partially valid, but only with respect to Tibetan urban areas. It is only in the main Tibetan cities and towns where it can be argued that Han are outnumbering Tibetans, or at least matching their number. Nonetheless, the sheer Tibetanness of the Tibetan rural areas itself refutes the claim of overall population swamping through simple arithmetic: even if the Han became dominant in urban populations, in most cases their urban dominance would not even come close to compensating for the rural dominance of Tibetans. This would hold even if estimates of the Han population were significantly underestimated. Ironically, the perception of population swamping is essentially an urban-centric assessment, even while the Tibetan areas remain some of the most rural of China. Yet this is precisely why the issue is so problematic given that it reflects perceptions of exclusionary dynamics experienced in urban areas. Migration Returning to the historical symbols of population invasion, important features differentiate the experience of the Tibetan areas from Inner Mongolia, Xinjiang, and the northeast core of Qinghai. Han migration to Inner Mongolia was largely agrarian and took place as early as the eighteenth century, at a time when rural wealth was still coveted by migrants in China.



Population Foundations of Marginalization in Tibet 99

Mongolians were already in a minority position in Inner Mongolia by the end of the nineteenth century, when population transfer for the purpose of land colonization became official policy under the Qing dynasty. The agrarian focus of such policies meant that Han migrants settled in the countryside and became dominant in rural as well as urban populations. Thus, by 1947 Mongolians accounted for only about 14 percent of the population of Inner Mongolia. This share dropped further to 11 percent by 1964 during the peak of Maoist transfer policies. The share slowly regained ground to 13.5 percent by 1985 (Burjgin and Bilik 2003, 55-56). As noted in the introduction, the share reached 17 percent in the 2000 census. Similarly, the large-scale population transfers that turned the share of Han from 6 to over 40 percent of the population in Xinjiang and from around 50 percent to over 60 percent in Qinghai took place during the heydays of Maoism from the 1950s to the 1970s (re Qinghai, see QSY 2005, tables 4-3 and 4-5). In contrast, rural or agrarian population transfer never took hold in the Tibetan rural areas, in large part due to the failures of agrarian colonization in the 1950s and 1960s. In particular, attempts to convert high-altitude rangelands into large-scale collective farms led to significant desertification in several Tibetan areas.27 In the case of Qinghai, Goodman (2004b, 38688) notes that large transfers from 1956 to 1959 were matched by an almost equally large exodus following the famine of the Great Leap Forward. Similarly, in-migration during the Cultural Revolution was balanced for the most part by out-migration over the same period. Thus, the Chinese plans in the 1950s to reach a population of ten million people in Qinghai by 1967 were singularly unsuccessful; by 1978, the population of Qinghai had reached only 3.65 million (QSY 2005, table 3-1). This demonstrates the importance of differentiating between rhetoric and reality in PRC “population transfer” policies. Due to similar experiences throughout Tibet, Tibetan rural areas have remained mostly Tibetan, with few exceptions outside strategic or borderland locations.28 Conversely, the Han in most Tibetan areas are concentrated mostly in towns and cities. Similarly, the destination of most new Han migrants is urban (or peri-urban in the case greenhouse agriculture). There is little or no incentive for this to change given the poverty and harshness of the Tibetan rural areas (in the eyes of the lowland Chinese), and the fact that most land is already fully utilized if not being taken out of production, with little potential for commercial forms of farming or ranching given the ecological sensitivity of the region (reinforced by state policies of ecological restoration since the late 1990s). This is a critical difference with Inner Mongolia, Xinjiang, and even Xining and Haidong in Qinghai, where Han populations were well entrenched in the rural areas by the advent of the reform period. Furthermore, the concept of transfer (i.e., organized large-scale movements of population from one region to another) would apply only in a

100

Chapter 3

general sense up to the end of the Maoist period in 1978, when population movements were controlled and closely managed by the state. Since the beginning of the reform period, this control has been gradually liberalized, with the result that the principal trend in population movements has been away from rural areas and poor peripheral regions, and toward towns, cities, and the coastal areas of China. As in most countries of the world, this trend has been reinforced by the declining economic importance of agriculture, rapid population growth, and the consequent underemployment of rural labor. In other words, agriculture is no longer a coveted target of migration. Population movement since the reform period, particularly since the early 1990s, has been more or less voluntary, implying that policies must rely on incentives rather than command to instigate migration. In this regard, the trend since the 1980s has been for a net population outflow from the impoverished western regions of China. The Tibetan areas, being some of the most impoverished, are not strangers to this outflow. Indeed, the propensity to out-migrate would tend to be stronger among Han given that they are not indigenous to Tibet, they are more urbanized and mobile, many (or their parents) were forced to move there in the first place, most maintain connections with other areas of China, and many complain of the harshness of the climate and altitude in Tibet. Thus, by nature they are more apt to (re)integrate into mainstream Chinese society. Current Tibetan refugee migration to India can also be understood in this context, given that many of such Tibetans migrate out of a perception that better fortunes might be forthcoming in exile.29 The only factors counteracting this overall migration trend are the heavily subsidized incentives and opportunities for Han, Tibetan, and Muslim workers, staff, and businesspeople to come to and remain in Tibet. Nonetheless, net outflow has been the rule for much of the reform period despite incentives. For instance, registered net migration to the TAR was negative in every year from 1981 to 1992 besides three (Iredale et al. 2001, 151). The greatest outflows took place in 1981-1982 due to a change in cadre policy, which allowed many Han cadres to return home and increased the representation of local Tibetan cadres (Huang 1995). Outward flows were no doubt fueled by recession in the TAR up to the early 1990s and net inflows would have resumed only sometime in the mid-1990s following the heavily subsidized recovery of the TAR economy. In Qinghai, registered net outflows started in 1980 and remained consistently negative from 1986 until 2006, with peaks occurring between 1988 and 1993 (QSY 2009, table 3-3). The slightly positive net inflows in 2007 and 2008 reflect the rising intensity of subsidization in that province under the Eleventh Five-Year Plan, as discussed in the next chapter, although they remain basically insignificant at 0.04 percent of the total population, versus a natural rate of population



Population Foundations of Marginalization in Tibet 101

increase in 2008 of over 0.8 percent. In both the TAR and Qinghai, it can be assumed that the bulk of the outflow was Han.30 These statistics refer to officially registered moves (i.e., they are recorded by the provincial Public Security Departments, which are in charge of residency registrations). They therefore relate to changes in the non-military31 population with permanent or other forms of formal residential status, whereas many migrants operate under a variety of temporary statuses.32 It might therefore be argued that the above data on net migration do not reflect reality given that the majority of Han in the Tibetan areas are temporary migrants, even by official measures. Indeed, on top of the difficulties of measuring temporary migrants in China (as elsewhere in the world), the political sensitivity of such migrants in the TAR adds additional problems given the incentives for the government to hide or disguise the prevalence of Han in-migration. The difficulties are compounded by the fact that the nationality and the residency status data are tabulated separately in Tabulation (2002) without cross-reference. Nonetheless the residency status of the Han in the TAR can be extrapolated by comparing this source to data provided by the TAR Public Security Department (PSD), suggesting that a little over half of the Han residing in the TAR on 1 November 2000 were temporary residents.33 However, the tenuousness of this extrapolation begs for further analysis. A certain degree of creative interpolation is required to analyze these data further. This can be done by comparing annual population surveys with the 2000 population census. The population surveys are conducted with permanent resident households and exclude temporary resident households. This introduces a bias in the surveys depending on whether a province experiences a net inflow or outflow of population. Populations in inflow provinces are underestimated given that migrants are not usually registered as permanent residents. Populations in outflow provinces are overestimated, given that temporary migrants usually maintain permanent status in their source provinces, which exaggerates the baseline residency records used for population survey estimates. In contrast, the 2000 census went to great lengths to record temporary residents and represented the most accurate portrait of effective provincial populations up to that time.34 Keeping in mind the difficulties of measuring temporary migrants in China, the discrepancy between the two sources can nonetheless help to infer net inflows or outflows in a way that was not intended in either source. According to the above logic, the 2000 census population count should exceed the 1999 survey count (plus population increase) in net inflow provinces, and vice versa in net outflow provinces. In several of the strategic and sparsely populated far western provinces such as the TAR or Qinghai, the two trends of inflow and outflow overlap and create a churning effect. Migrants from other provinces, whether state

102

Chapter 3

organized or “spontaneous,” usually come for short periods and do not necessarily change their registration status to the host province, even if residency rules are considerably more lax than in the coastal areas. On the other hand, this inward flow overlaps with the out-migration of previous temporary in-migrants and, to a lesser extent, locals. Although out-migration receives less attention, it is a considerable source of concern for local authorities, particularly with respect to more skilled or educated out-migrants, as indicated to me in several interviews with local officials in Qinghai. During a period of increasing government spending, inflows potentially outweigh outflows, in which case the surveys would underestimate the effective population and would be lower than the census population count. The results of this inferred analysis are presented in table 3.2. According to this comparison, Guangdong in the southeast stands out as the quintessential inflow province, whereas Guizhou, the poorest province of China, is the most extreme case of an outflow province. Both are intuitive results. Xinjiang also stands out as a definite inflow province, supporting claims of large-scale in-migration, which clearly corroborates with the 2000 census data for Xinjiang. These data show that 1.41 million out of a total of 2.83 million temporary residents (or “migrants,” per census definitions) in Xinjiang were from outside the province (i.e., their permanent household registration was in a province other than Xinjiang at the time of the census), or a total of 7.6 percent of the total population measured by the census, versus only 4.2 percent in the TAR. In comparison, both the TAR and Qinghai appear relatively stable. In both cases, the difference between the 2000 census and the 1999 survey is less than 1 percent above the rate of natural population increase in 1999 (see the third-to-last and last columns in table 3.2). Measurement errors could account for much of this difference. Indeed, Qinghai recorded the largest estimated census error at 7.4 percent (second-to-last column, meaning that the final adjusted census population, after a post-census survey was conducted, was 7.4 percent greater than the original census results for Qinghai). If the original census tabulation is considered (column C), Qinghai definitely appears as a net outflow province. Despite the evident visibility of in-migration in the Tibetan areas, these observations point to the fact that such migration involves considerable churning, rather than sustained inflows as in Xinjiang. As noted above, the common rebuttal is that Han migrants are underestimated in the Tibetan areas due to the sensitivity of their presence, particularly in the TAR. For instance, the 2000 census recorded only 158,570 Han in the TAR (Tabulation 2002, table 1-6), a number that is treated with disbelief by most observers. Yet despite a similar sensitivity, an underestimation of the Han population in Xinjiang is not apparent in the census data, which unequivocally show a large inflow from outside the province. Also,



Population Foundations of Marginalization in Tibet 103

Table 3.2: Selected province-level population statistics, c. 2000 Population (in millions) according to A 1999 survey

B

C

2000 2000 adjusted census census tabulation

Discrepancies

RNPI

D

(percent)

(per 1000)

2001 survey

(B-A)/A (B-C)/C

1999 survey

Inflow provinces Guangdong

72.70

86.42

85.23

77.83

18.9%

1.4%

9.92

Shanghai

14.74

16.74

16.41

16.14

13.6%

2.0%

-1.10

Xinjiang

17.74

19.25

18.46

18.76

8.5%

4.3%

11.80

Fujian

33.16

34.71

34.10

34.40

4.7%

1.8%

5.21

5.43

5.62

5.486

5.63

3.5%

2.4%

12.32

Stable provinces Ningxia TAR

2.56

2.62

2.616

2.63

2.3%

0.2%

15.80

Yunnan

41.92

42.88

42.36

42.87

2.3%

1.2%

11.66

Qinghai

5.10

5.18

4.82

5.23

1.6%

7.4%

13.90

Inner Mong.

23.62

23.76

23.32

23.77

0.5%

1.9%

7.24

Gansu

25.43

25.62

25.12

25.75

0.7%

2.0%

9.17

Shaanxi

36.18

36.05

35.37

36.59

-0.3%

1.9%

6.13

Sichuan

85.50

83.29

82.35

86.40

-2.6%

1.1%

6.78

Guizhou

37.10

35.25

35.25

37.99

-5.0%

0.0%

14.24

National

1259.09

1265.83

1242.61

1276.27

0.5%

1.9%

8.77

Outflow provinces

Note: See text on categorization of inflow, stable, and outflow provinces. Column B, the “2000 adjusted census,” represents the official estimated population after adjustments were made following the post-census survey. Column F indicates the estimated census error. Sources: calculated from CSY (2000, table 4-3); CSY (2001, table 4-3); and CSY (2002, tables 4-3 and 4-8).

the TAR aside, an inflow of Han into western provinces such as Qinghai is not necessarily treated as a subject to be avoided by local officials. On the contrary, often it is interpreted as an indication of success in attempts to attract skilled labor and to stem outflow, thus supporting the development goal of building human resources. In the case of the TAR, the Han census count was obviously an undershot. Arguments of statistical manipulation aside, there may be several reasons why the Han count was low. Temporary migration is not necessarily one of them. In fact, as noted above, the 2000 census did attempt to record all types of migrants, down to those who had not yet resolved their residency status. The only migrants who would not have been counted were either those who had hid from the census takers (which is more of an issue in the coastal areas given official belligerence toward migrants, but less in the west

104

Chapter 3

where migration is encouraged), or else the common omissions related to the difficulties of tracking migrants versus permanent residents. On the other hand, timing is critical. The census was taken on 1 November when most of the seasonal Han migrants would have already left for the winter. The figure of 158,570 may therefore represent a reasonable count of the number of non-military Han who were actually residing yearround in the TAR in 2000. Also, military personnel, who are not included in this count, would represent a significant proportion of the visual presence of Han in Lhasa and in other cities and towns. Conversely, informal evaluations by tourists and journalists are made mostly during the summer months, when the Han presence is swollen by tourism. The number of tourists visiting the TAR subsequently skyrocketed from 1.6 million in 2005, to 1.9 million in 2006, 4 million in 2007, and 6.82 million in 2010—and most of these were domestic Chinese, most of who visited during the summer months.35 Although there has been a definite increase in temporary Han migration to the TAR since the mid-1990s, particularly during the summer months in parallel with tourism, such migration is not stable. It is to be differentiated from the case of Xinjiang in that Han migrants by and large do not settle in the long term. For instance, surveys taken in the TAR in the mid-1990s indicated that Han migrants were disproportionately male and did not tend to bring their families, leaving them at home and concentrating on economic activities in the TAR for a temporary period, perhaps five to six years on average (Iredale et al. 2001, 156-58). These insights were reconfirmed in a 2005 survey by Ma and Lhundrup (2008). Furthermore, my own field observations from Lhasa suggest that many Chinese migrants in the TAR are seasonal migrants, particularly since the massive boom in tourism started to take off in the early 2000s. This seasonal pattern of migration has been further facilitated by the Qinghai–Tibet railway, which opened in 2006 and substantially reduced the costs of traveling to Lhasa (indeed, this explains the jump in tourist numbers in 2007). Given these characteristics, Han migration would tend to be very susceptible to changes in social or economic conditions, such as the March 2008 protests in and around Lhasa, which subsequently caused a sharp drop in tourist numbers (PD 2008). This probably had the knock-on effect of reducing the tourism-related seasonal migration flows as well. Conversely, rapidly increasing subsidies, as has been the case since the mid-1990s, obviously encourages a trend of net inflows. Indeed, despite contentions that the railway would accelerate inmigration, increased inflows are more likely related to ongoing policies of subsidization rather than the easing of transport per se, whereas the easing of transport in the absence of subsidies could quite well have the opposite effect of encouraging outflows. However, the question remains whether net inflows could be sustained in the long term if and when subsidies stabilize



Population Foundations of Marginalization in Tibet 105

or social stability worsens. The underlying trend would suggest that inflows face an uphill battle against outflows. Furthermore, structural transformations in the regional economy could exacerbate out-migration of both Tibetans and Han from certain counties, even while increasing in-migration to others. For instance, local governments in the Tibetan areas of Sichuan, Qinghai, Gansu, and Yunnan are considerably more fiscally austere than in the TAR. They are often left with few fiscal means to stem an outflow of labor in the event of a collapse of a key industry, such as following the 1998 moratorium on forestry activities declared by the national government in response to flooding on the Yangtze, which had a debilitating impact on many Tibetan counties in Sichuan and Yunnan. Therefore, it cannot be assumed that the migration scenario of the TAR represents the norm in the Tibetan areas outside the TAR, given that the former is primarily a response to massive levels of subsidization. Migration versus Demographics As noted above, most of the rural areas of Tibet remain overwhelmingly Tibetan to this day, along with pockets of several other closely related ethnicities. For instance, 97.6 percent of the TAR rural population in the 2000 census was Tibetan. Even if the Han population of the province were underestimated, the underestimation would relate mostly to the urban areas and would not particularly influence this rural share. Even in eastern Tibetan areas that are relatively close to large cities such as Xining, Lanzhou, or Chengdu, rural populations in the decisively Tibetan areas also remain predominantly Tibetan. This was consistently confirmed to me in my fieldwork and can be shown through a county-by-county examination of the 2000 census, which is beyond the scope of this chapter. Because fertility and birthrates are considerably higher in the rural areas, the Tibetan population holds the upper hand in the demographic momentum. This demographic advantage is amplified by their low rates of urbanization, given that rural dwellers have more lenient birth limits, and by their younger age structure—31.2 percent of the TAR population was under the age of fifteen in the 2000 census, versus 22.9 percent in China (CPSY 2002, table 2-6). As these cohorts enter their reproductive years with the two-child limit presumably in place, they will sustain higher rates of natural population increase relative to the Han for the next twenty to thirty years at least. In contrast, the few Han who settle permanently in the Tibetan areas mostly have an urban Chinese demographic profile, that is, one child. Those who migrate on a temporary or seasonal basis are disproportionately male (besides sex workers) and of working age, and they do not tend to bring their families along during their working sojourns (see Ma

106

Chapter 3

and Lhundup 2008). While these last attributes amplify migrant demand for alcohol and sex workers, they considerably lower their rates of natural population increase. In other words, sustained net inflows of Han would be required merely to maintain their population share, let alone increase it. For instance, taking the rate of natural population increase for the TAR in 2003 and assuming that the Chinese component of this rate was zero (which is probably a generous assumption), a sustained net inflow of about 2,000 Han per year would have been required merely to maintain the Han population share (calculated from CSY 2004, table 4-3). Such net inflows would be easily sustained so long as the economy is booming, and they were definitely exceeded between 2000 and 2010 given the slight rise in the share of Han Chinese in the population of the TAR, as discussed further below. However, if the boom were to taper off, the balloon of Han migration could equally deflate. Thus, despite the visibility of in-migration to urban areas, Tibetan population shares in certain cases may be increasing rather than decreasing, particularly outside the TAR. For example, the share of Tibetans in the Qinghai population has been slowly edging upward during the reform period, from 18.5 percent in 1978 to 21.9 percent in 2000, and growing at about the same rate overall as the Hui Muslims (QSY 2005, table 4-5).36 The Tibetan population share in the TAR fell between the 1990 and 2000 censuses, due to the reasons discussed above and the fact that the Han have been increasing from a very small base. The general patterns are shown in table 3.3, which focuses on changes between the 1990 and 2000 censuses. Tibetans are not differentiated from other minorities. The category of “minorities” in the TAR is almost entirely Tibetan. In Qinghai it is about two-thirds Tibetan, in combination with Mongours (Ch. Tu) and Mongolians, and about one-third Hui and Salar. It Table 3.3: Changes in Han and minority populations, 1990 and 2000 censuses TAR

Qinghai

Sichuan

Gansu

Yunnan

China

Tibetans

1.3%

1.4%

1.1%

--

Annual change in population (%; non-cumulative) Total (%)

1.7%

1.5%

0.6%

Minority (%)

1.5%

2.2%

1.9%

1.8%

1.5%

1.5%

1.8%

Han (%)

6.5%

0.9%

0.5%

1.3%

1.4%

1.0%

--

Share of minorities in total population (%) 1990 census

95.9%

42.2%

4.4%

8.3%

33.4%

8.0%

0.41%

2000 census

93.9%

45.6%

5.0%

8.7%

33.4%

8.4%

0.43%

Share change

-2.0%

+3.4%

+0.6%

+0.4%

0.0%

+0.4%

+0.02%

Sources: calculated from CPSY (2001, tables 2-13, 2-14, and 2-21).



Population Foundations of Marginalization in Tibet 107

is about one-third Tibetan in Sichuan; one-fifth in Gansu and a small fraction in Yunnan and in China. Outside the TAR, minorities either increased or maintained their shares throughout the 1990s. While some of this might be due to ethnic status switching,37 more generally the results are not surprising. As discussed above, outflows and demographics would tend to dominate over inflows in these more fiscally austere and peripheral provinces, particularly among the more mobile and educated urbanites, who tend to be Han rather than Tibetan or Muslim. Notably, the annualized population change of Tibetans in the TAR measured by the two censuses is lower than that of the minorities in the other provinces, particularly in Qinghai, although it corresponds to the NPIR of the TAR shown in figure 3.2. This is suggestive of even higher birthrates among both Tibetans (especially pastoral Tibetans) and Chinese Muslims in Qinghai and Gansu up to the 2000 census—and among Tibetans and other minorities in Sichuan—which corroborates with the previous discussion of this chapter. Overall, these results are consistent with the argument that the TAR has been the only province where the Tibetan population share is decreasing (up to 2000), albeit from a very high level. Tentative Previews from the 2010 Census The surveys after the 2000 census were presumably refined with the findings and lessons of the 2000 census and thus major discrepancies between the 2009 population survey data reported in CSY (2010) and the 2010 census data reported in CSY (2011) do not appear as evidently as they appeared in the earlier data analyzed in table 3.2 above. Rather, the surveys in the 2000s appear to have accounted for migration flows more accurately than those of the 1990s, to the extent that Sichuan was reported as having a shrinking population throughout the decade (implicitly due to migration) despite having a positive rate of natural population increase. Nonetheless, a similar extrapolation can be performed on the few data from the 2010 census that have been made available so far (at the time of finalizing this book).38 By comparing the data on population growth rates between the 2000 and 2010 censuses with the rates of natural population increase (NPI) estimated by annual population surveys throughout the decade and by the 2010 census, it is possible to make rough but arguably robust measures of the patterns of net migration for each province over the same period. This is based on the understanding that populations change by only one of two causes—natural increase or net migration—hence, we can deduce the effect of net migration by deducting the known effect of natural increase from population growth. This approach bypasses the need to use the more precise but also more problematic data on residency status and also takes

108

Chapter 3

advantage of the consistencies in enumerating migrants in both the 2000 and 2010 censuses, in contrast to the 1990 census, which was inconsistent with the 2000 census in certain respects, resulting in many discrepancies.39 The 2010 census thereby provides the first opportunity to perform this exercise with some degree of accuracy. Two sources of data are used for this comparison. The first is the annualized rate of population growth between the two censuses, presented in CSY (2011, table 3-12). These population growth measures include net migration, given that both censuses enumerated migrants in principle. In practice, the enumerations undoubtedly led to numerous errors and omissions with respect to migrants, although these data are nonetheless the best available to date. The 2000 population reported in this table—and generally in official sources—is the official estimated population after adjustments were made following the post-census survey, which estimated a census error of 1.9 percent, that is, the estimate of the national population was 1.9 percent higher than the original census tabulation, as reported in Tabulation (2002), after corrections were made based on the post-census survey results. Accordingly, the results of this exercise are dependent on the accuracy of the corrections to the 2000 data, particularly for the provinces where large census errors were recorded, such as Qinghai, which had a census error of 7.4 percent.40 However, for better or worse, we must adopt these official data because they have been used as the basis for subsequent measures and estimates—indeed, the corrections are probably explained by more accurate estimations of migrant populations in the post-census survey than in the original census. The second source is the rates of natural population increase (NPI) estimated by annual population surveys throughout the decade and by the 2010 census. For the purpose of comparing the NPI rates with the annualized growth rates, an average NPI rate is calculated for each province by using the NPI rate recorded in each annual survey from 2001 to 2009 (and reported in the subsequent CSY of each respective year), in addition to the NPI rate recorded in the 2010 census and reported in CSY (2011, table 3-4). When the NPI rate is subtracted from the overall population growth rate, a significant positive discrepancy (where population growth is substantially greater than the rate of natural population increase) would imply net population inflows, whereas a significant negative discrepancy would imply the opposite. For instance, if a province grows by 2 percent a year but its NPI rate (births minus deaths) is only 1 percent a year, this implies that the net rate of in-migration amounts to 1 percent of the population per year. Or, if a province experiences no population growth but has an NPI rate of 0.5 percent a year, this effectively means that there is a net out-migration from the province of 0.5 percent of its population a year. As noted above, this calculation relies on the fact that the standard of the 2000 census was



Population Foundations of Marginalization in Tibet 109

apparently maintained in the 2010 census with regard to including all de facto residents in the census enumeration, including those with the most temporary residency status. The results are shown in table 3.4. Column C (the “net migration rate”) is the particular column to focus on as it reflects (indirectly) the degree to which a province is a net inflow or net outflow province after accounting for natural population increase. Column D shows the proportion of growth that is accounted for by net in-migration (or outmigration, if negative). Column E shows the absolute increase in population over the ten-year period, as reported in CSY (2011, table 3-12), and column F shows the absolute number of in (or out) migrants over the ten-year period estimated by this exercise. The results are presented for a greater selection of provinces than in table 3.2, in descending order of net migration rates. The results are predictable and broadly confirm the previous analysis of table 3.2, although there have been some interesting shifts in a few cases. The strongest population growth rates for the whole decade were observed in the coastal growth poles, for example, Beijing (41.9 percent), Shanghai (37.5 percent), Tianjin (29.3 percent), Guangdong (20.7 percent), and Zhejiang (16.4 percent). Population growth was also strong in three sparsely populated and heavily subsidized western autonomous regions that are known to have the highest NPI rates in the country in combination with substantial degrees of in-migration, namely, Tibet (14.6 percent), Xinjiang (13.3 percent) and Ningxia (12.1 percent). There was moderate population growth ranging from 1 to 10 percent in most other provinces, with faster growth occurring in the more eastern (or coastal) parts of each region, and slower growth in the more interior (or western) parts of each region. There were six cases of population decline, including the major engines of population outflow located in southwest and south central China, namely, Chongqing (-6.6 percent), Hubei (-5 percent), Sichuan (-3.4 percent), and Guizhou (-1.4 percent), as well as two smaller cases from east and northwest China, namely, Anhui (-0.6 percent) and Gansu (-0.2 percent). In terms of absolute numbers, the largest growth occurred in Guangdong (17.88 million people) whereas the largest decline occurred in Hubei (-3.04 million people). Regarding net migration flows, the results can be depicted in absolute and relative terms. In terms of absolute numbers, the largest net flows were out of Sichuan (-5.4 million people), Hubei (-4.8 million), Anhui (-4.1 million), Guizhou (-3.4 million), and Chongqing (-3.0 million), and into Guangdong (+10.7 million), Shanghai (+6.1 million), Beijing (+5.5 million), Zhejiang (+5.4 million), and Tianjin (+2.7 million). The leading two inflow provinces in terms of the rate of net migration from 2000 to 2010 as a proportion of the 2000 population were Beijing followed by Shanghai, where almost all population growth was due to net in-migration. In Beijing, net inflows throughout the decade were equivalent

110

Chapter 3

Table 3.4: Net population inflow or outflow from 2000 to 2010, selected provinces A

B

C

Percent change per year

Region

2010 pop. Pop. million growth persons rate

Avg. NPIR

National

1,339.72

0.57

0.57

0.00

3.56

0.18

3.38

D % share

Net migra- Net migr./ tion rate growth (A-B) (C/A)

E

F

1,000 persons 10-year pop. growth

10-year net migr. (D*E)

0.6

73,890

464

94.9

5,790

5,498

Inflow provinces (C > 0.1%) Beijing

19.61

Shanghai

23.02

3.24

0.10

3.14

96.9

6,280

6,082

Tianjin

12.94

2.60

0.18

2.42

93.1

2,930

2,725

Guangdong

104.30

1.90

0.76

1.13

59.7

17,880

10,679

Zhejiang

54.43

1.53

0.44

1.08

70.9

7,660

5,430

TAR

3.00

1.37

1.12

0.25

18.5

380

71

Xinjiang

21.81

1.26

1.10

0.16

12.6

2,560

324

Ningxia

6.30

1.15

1.05

0.10

8.6

680

58

0.60

0.01

1.2

2,180

27

Stable provinces (C between +/- 0.1%) Fujian

36.89

0.61

In. Mong.

24.71

0.39

0.40

-0.01

-3.1

950

-29

Shaanxi

37.33

0.35

0.41

-0.06

-16.8

1,280

-214

0.70

0.81

-0.11

-16.2

3,090

-499

Outflow provinces (C < -0.1%) Yunnan

45.97

Qinghai

5.63

0.83

0.98

-0.15

-17.5

450

-78

Hunan

65.68

0.20

0.53

-0.34

-170.7

1,280

-2,191

Henan

94.02

0.16

0.54

-0.39

-245.4

1,460

-3,592

Guangxi

46.03

0.25

0.80

-0.55

-218.7

1,140

-2,486

Sichuan

80.42

-0.35

0.30

-0.65

186.4

-2,870

-5,353

Gansu

25.58

-0.02

0.64

-0.66

3751.2

-40

-1,679

Anhui

59.50

-0.06

0.63

-0.69

1151.5

-360

-4,134

Hubei

57.24

-0.52

0.29

-0.81

156.7

-3,040

-4,769

Guizhou

34.75

-0.14

0.82

-0.97

672.2

-500

-3,385

Chongqing

28.85

-0.69

0.32

-1.01

146.8

-2,050

-3,015

Sources: calculated from CSY (2011, tables 3-4 and 3-12) and equivalent in earlier yearbooks.

to 39.8 percent of its 2000 population and accounted for 95 percent of population growth up to the 2010 census. In Shanghai, net inflows were equivalent to 36.3 percent of its 2000 population and accounted for 97 percent of population growth.



Population Foundations of Marginalization in Tibet 111

The two leading outflow provinces were Chongqing, where despite the apparent success of its now-famed “model,” net population outflow throughout the decade was equivalent to 9.8 percent of its 2000 population and was almost 50 percent greater than the rate of population decline, followed by Guizhou in the southwest and the poorest province of China, where net outflows were equivalent to 9.7 percent of its 2000 population and were more than five and a half times greater than the rate of population decline. Guizhou actually registered only a marginally shrinking population between the two censuses; the strong outflows are explained by the fact that it had one of the highest NPI rates in the country. Similarly, the population of Gansu was stable between the two censuses, whereas an average NPI rate of 0.64 percent led to an annual net outflow of approximately 0.66 percent of its population, proportionately about the same as Sichuan. From the perspective of this and the previous measures in table 3.2, Gansu is one of the provinces that fell from having a relatively stable population in 2000, with population growth roughly correspondent to natural increase rates and, hence, with balanced migration, to having a strong net outflow by 2010, despite rapid economic growth and increasingly intensive subsidization over the decade, as discussed in the next chapter. While these outliers are somewhat predictable, particularly at the net receiving end, the results also offer a revised understanding of the place of various middle-ranking provinces within these flows, particularly with regard to provinces that had higher or lower than national average rates of natural increase. For instance, Qinghai Province grew 0.83 percent a year between the two censuses, but given its average NPI rate of 0.98 percent a year over this same period, which was the fourth highest in the country, the province actually experienced an effective net outflow of population at 1.5 percent of its 2000 population over these ten years. Eight other provinces also registered positive population growth at the same time as net out-migration. In total, fifteen out of thirty-one provinces experienced net population outflows between the two censuses according to this measure (not all are shown in figure 3.4), versus only six that experienced negative population growth. Three exceptions to the general western pattern of net outflows were the TAR, Xinjiang, and Ningxia, which had the three highest average NPI rates in China and also experienced moderate net in-migration (as informal observations would suggest). Inversely, several eastern provinces that were not particularly outstanding in terms of population growth, such as Jiangsu, Liaoning, and Heilongjiang, became much more evident as destinations of migration given that their NPI rates were similar to Beijing and Shanghai and among the lowest in China. The most striking shifts in this selection of provinces, in comparison to table 3.2, are the divergent trends in the TAR and Qinghai—two of our provinces of particular interest. The TAR predictably moved from being a

112

Chapter 3

relatively stable province in 2000 to a province with a moderate rate of net inflow, at a quarter of a percent per year combined with an average rate of natural population growth of 1.12 percent (the highest in China) leading to an overall population growth rate of 1.37 percent per year. This placed its rate of in-migration higher than that of Xinjiang, albeit from a much lower base. In absolute numbers, 380,000 people were added to the population of the TAR over the ten-year period, of which about 71,000 were from net migration according to these measures. This might again appear to be an undercount of net in-migration. Indeed, as implied by the data presented by Xinhua (2011a), the Han population alone increased by approximately 86,707 people over the decade, reaching a total of 245,277 people in 2010, and most of the additional Han were presumably migrants, whether they were registered as temporary or permanent residents (i.e., there was probably no positive natural increase among the Han in the TAR). Xinhua (2011a) reported that these population data represent permanent residents, not temporary migrants, although we have reasons to believe that this assertion is erroneous and that the numbers do include temporary migrants, as has been the standard method of census reporting since at least 2000 and as reported in CSY (2011).41 This will only be clarified once the government releases the detailed census tabulations. Net in-migration would have included other nationalities as well, notably Chinese Muslims and Tibetans from outside the TAR (although many of the latter might have been forced to leave the TAR following the protests in 2008). Hence, if the extrapolation in table 3.4 is relatively accurate, the only plausible explanation for the discrepancy between the number of net in-migrants versus the increase in the number of Han (and for the rising Han share in the face of a relatively rapid NPI rate) is that the net inflow of Han Chinese was compensated for by a net outflow of local Tibetans—whether to exile or else to other parts of China, including students being sent to “inland” high schools, or else graduates working in other provinces.42 Given the magnitude of the numbers of Han Chinese involved, this explanation is quite plausible and would confirm the suggestion in this section that migration flows in the TAR involve considerable churning, including substantial outflows of young and educated Tibetans. Obviously, the 2010 head count of Han Chinese in the TAR—at 245,277 people per the data provided by Xinhua (2011a)—would also appear to be hugely underreported, even more so than in 2000. Nonetheless, the same qualifications apply as with the 2000 census data. The 2010 census was again taken on 1 November, and the addition of the railway in 2006 as a cheaper travel option for migrants would have enhanced the seasonality of migration in the TAR. Hence, the Han head count needs to be understood as referring to the number of non-military, non-tourist Han who were



Population Foundations of Marginalization in Tibet 113

actually residing in the TAR during the off-peak tourist season, rather than the boom-time population of the tourist season, when most observers visit the TAR. Along these lines, the new Han head count might well be reasonable for the period of the year, or at least not hugely underestimated. Even if underestimated, it is nonetheless interesting that TAR does show up through these indirect measures in table 3.4 as having experienced a greater degree of net inflow over the ten years up to 2010 than in 2000 (as reflected in table 3.2), as would be expected from the developments during this period and numerous qualitative observations. The results also correspond to the increase in the Han share from 6.1 percent in 2000 to 8.2 percent in 2010. Conversely, Qinghai actually slipped from being a stable province in 2000 to a moderate outflow province in 2010, despite an even higher rate of economic growth in the 2000s than the TAR and the second highest and also intensifying degree of subsidization, as discussed in the next chapter. Although Qinghai had one of the highest NPI rates at 0.98 percent a year on average between 2001 and 2010, the population grew by only 0.83 percent a year, implying a net outflow of 0.15 percent a year, or about a fifth of the natural increase over the ten-year period. Obviously, one reason for this result might be related to the large census error in the 2000 census, that is, if the population was actually closer to the original 2000 census tabulation (column C in table 3.2), then the rate of population growth up to the 2010 census would have been greater, corresponding to the rate of NPI. However, presuming that the adjustments to the 2000 census results were done in a fairly rigorous manner, then it is more plausible that the case of Qinghai demonstrates the rule argued in this section—that despite rapid economic growth, Qinghai, like Gansu and Sichuan, has increasingly come to resemble the norm of other western provinces in that migration outflows have tended to dominate over inflows, as would be expected of a peripheral region increasingly integrated into the national economy. Indeed, as analyzed in the next two chapters, the character of recent growth in Qinghai, unlike in the TAR, has been strongly rooted in capital-intensive productive sectors such as mining, hydroelectric power generation, or steel and aluminum production, in which case, growth based in these sectors would have generated relatively little employment per unit of output. We do not yet know the disaggregation of these population data by nationality, although based on the logic of the argument of this section, it is plausible that out-migration was stronger among the Han than among Tibetans and Chinese Muslims, thereby raising the population shares of the latter, particularly given their much higher NPI rates than the Han in Qinghai. In this sense, it is plausible that the pattern of population change in the TAR with respect to the share of Han continues to be an exception among the Tibetan areas in China.

114

Chapter 3

The Cases of Xunhua County and Malho (Ch. Huangnan) Prefecture It is worth taking a closer look at the dynamics in Qinghai through an examination of two borderland cases: Xunhua Salar Autonomous County (SAC) and Malho (Ch. Huangnan) Tibetan Autonomous Prefecture (TAP) in Qinghai. These two jurisdictions are particularly interesting because, as discussed in the previous chapter, they are on the edge of the northeast core of Qinghai, with relatively decent roads connecting them to Xining within a three-to-four-hour bus ride. Both have experienced significant development and urban growth since the mid-1990s. Salar-owned private industry has driven growth in Xunhua, whereas hydroelectric dam construction has driven growth in Malho.43 Given the locations and circumstances of Xunhua and Malho, presumably both would be on the front line of any impending population invasion. Precisely for this reason, these two jurisdictions were chosen as “most likely” case studies for my fieldwork. Analysis of these two cases was facilitated by population statistics provided to me during interviews with officials in the local statistics department of each jurisdiction in 2004.44 Xunhua is part of the Haidong District even though it is on the south side of the Yellow River (whereas the rest of Haidong is on the north side). Although a Salar Muslim county, Xunhua contains a significant minority of Tibetans. This Tibetan minority are concentrated in four townships in the highlands of the county, the townships effectively operate as Tibetan autonomous townships and account for almost half of the county land area. All the Xunhua Tibetans whom I interviewed in 2004 reported that the Tibetan townships were almost entirely Tibetan in ethnicity; the Salar and Tibetan rural populations remain segregated despite their very close proximity. Both groups were mostly agrarian up to the 1980s and continue to have very low levels of schooling, both among the lowest in China (see chapter 6). Malho is a predominantly Tibetan prefecture with four counties, also on the south side of the Yellow River and bordering Xunhua and Haidong. The northernmost county of Chentsa (Ch. Jianza) contains a Muslim minority (mostly Hui), who account for about one-quarter of the county population and about 70 percent of the total Muslim population in the whole prefecture. The prefecture seat is Rongwo (Ch. Longwu) Town in Rebgong (Ch. Tongren) County to the south. Rebgong contains small communities of Hui and Salar Muslims, concentrated in Rongwo Town and a few neighbouring villages, who comprise about 5 percent of the county population. The two other counties to the south, Serkog TAC and Sogwo Mongolian AC, are both pastoral with almost no settled Muslim or Chinese presence outside the single county town of each prefecture. Rather than population invasion, the opposite has occurred in these two jurisdictions, correspondent to the more general analysis above. The shares



Population Foundations of Marginalization in Tibet 115

of the two main nationalities in Xunhua (Salar and Tibetan) were relatively stable within the overall total population, which increased from 89,445 in 1985 to 116,332 in 2003. The Salar gained in share, from 57.6 percent of the total population in 1985 to 62.1 percent in 2003 (or from 51,535 to 72,253 people), while Tibetans dropped from 24.6 percent to 23.4 percent over the same period (from 21,961 to 27,173 people), albeit both shares were stable from 1999 onward. The Hui share fell from 9 to 8.2 percent and the Han share from 7.8 to 5.9 percent. The faster growth of the Salar is to be expected. The Salar, a small nationality based mostly in this county, are known as being particularly pro-natal and adept at avoiding family planning regulations, as recounted to me by a Salar professor in Xining in an interview in June 2004, as well as by several other Salar and Tibetan scholars and officials. In addition, the cultural and economic revival of Xunhua in the 1990s, based on very successful textile factories, contributed a pull factor for Salar return migration. Conversely, the slower growth of the Tibetan population probably represents some out-migration, such as young Tibetans migrating out of the county to neighboring Rebgong or nearby Xining for education or work.45 Rongwo in particular holds a special pull given that it is a cultural and education center for local Tibetans, who feel more protected there, just as the Salar feel more protected in Xunhua than in Rebgong. The Hui and Han populations and shares fell between 1999 and 2000, by over 20 percent in the case of the Han head count (from 8,587 to 6,684 people). County officials (one Tibetan and one Han) explained to me that this specific drop had been due largely to out-migration to the Tsaidam Basin (i.e., to the developments related to the controversial World Bank “Qinghai Project” in the late 1990s). Xunhua had been nominated as one of the impoverished out-migration counties of eastern Qinghai for this project, and it was mainly these Hui and Han who left for the occasion. The county officials also explained that the relatively slow growth of the Han population earlier in the 1990s had been due to out-migration, but there had been some return migration later in the 1990s due to the difficulties of finding work elsewhere. There had been little migration of non-Xunhua Han into the county over these years, although some non-Xunhua Hui had come from neighboring counties or from Gansu Province to work in the factories. Ultimately, out-migration was the dominant trend for all groups except the Salar. There was also significant out-migration among the Salar, although this had been tempered by the presence of Salar-dominated economic opportunities. The case of Malho illustrates a similar case of Tibetan dominance, correspondent to the general arguments made above for Tibetan areas outside the TAR. The overall total population of the prefecture increased from 157,278 in 1985 to 215,384 in 2003. The Han head count fell between

116

Chapter 3

1985 and 1995 (from 15,280 to 14,554 people), reflecting the pressures of out-migration. The Han head count started to rise thereafter, but surpassed its 1985 level only in 2000, when it reached 16,223 (and 16,822 by 2003). The return to a slightly positive rate of Han population growth from 1995 onward was probably due to in-migration stimulated by various development drives, such as the building of several very large hydroelectric projects and, more generally, the OWC from 2000 onward. However, the positive growth in the Han population between 1999 and 2003 did not outweigh the faster growth of the Tibetan population in those same years, which increased from 101,745 in 1985, to 136,566 in 2000 and 142,390 by 2003. As a result, the Han share of total population fell from 10.1 percent of the prefecture population in 1985 to 7.8 percent in 2000 and 2003, versus a Tibetan increase in share from 64.7 percent in 1985 to 65.9 percent in 2000 and 66.1 percent in 2003. Most of this loss in Han share and gain in Tibetan share occurred in the ten years up to 1994, after which both shares were relatively stable. Mongolians also gained in share, from 12.9 percent in 1985 to 13.6 percent in 2003. This Mongolian population has been indigenous to the region for hundreds of years, is mostly pastoral, concentrated in Sogwo (Ch. Henan), the southernmost county of the prefecture, and is assimilated into the surrounding Tibeto-Buddhist culture. It is therefore expected that they would have a population pattern very similar to the Tibetans in this prefecture. Together, the two groups accounted for a 2.1 percent increase in the population share of Malho between 1985 and 2003, reaching almost 80 percent of the prefecture population by 2003. The inclusion of the Mongour—another indigenous Tibeto-Buddhist group of this region—would place this rising share above 80 percent. The Hui and “Other” population shares stabilized from the mid-1990s onward at about the same share as in 1983. “Other” represents a combination of the Mongour and the Muslim Salar and Bao’an. The two Muslim groups have been settled in Rebgong for well over a century. Overall, both the Hui and “Other” experienced slight decreases in share from 1999 onward, despite popular Tibetan opinion that there were proportionately more and more Muslims in this prefecture during these years. Notably, the head count of “Other” even fell slightly between 2002 and 2003 (from 11,277 to 11,202 people), which might indicate the out-migration of Salar families to Xunhua or Xining following the anti-Muslim boycotts organized by Tibetans in 2003 (see chapter 8). Several cases of such Salar out-migration were reported to me during fieldwork in 2004. In other words, the predominant trends in Xunhua and Malho have been threefold. First, the Han shares of total population have fallen in both cases, presumably due to out-migration, particularly in the period up to the mid-1990s when this region was experiencing economic recession. Second, the dominant nationalities of both jurisdictions increased their shares of



Population Foundations of Marginalization in Tibet 117

total population. Third, the population shares of the indigenous minorities of both jurisdictions (the “minority minorities”) were maintained or fell slightly, possibly due to out-migration to their respective cultural or educational centers (Xunhua Tibetans to Rebgong or Xining, and Rebgong Salar to Xunhua or Xining). These observations support the general argument made above with respect to migration and demographics (abstracting from military presence, about which we have no data). Urban Exclusions All indications seem to point to the fact that the appearance of increasing Han or Muslim presence in Tibet—especially in eastern Tibet outside the TAR—represents more specifically an increased presence of migrants in urban areas, and a changing, more market-oriented character of migrants compared to the past, even while Tibetans might be maintaining or even increasing their overall population dominance. Concerns of population invasion aside, this situation is still problematic precisely because urban areas represent the centers of economic and political power, the dominant sources of economic opportunities in the new economies of the Tibetan areas, and hence crucial points of convergence for local urbanizing rural Tibetans as well as migrants from outside the Tibetan areas. Hence, the real issue is economic and political dominance rather than population dominance. More precisely, regardless of whether or not out-of-province migrants are flooding the local population, they are definitely flooding urban employment opportunities. This can be succinctly illustrated through an extrapolation of the 2000 census data for the TAR. While permanent and temporary Han residents accounted for 6.1 percent of the TAR population in 2000, they accounted for almost 20 percent of the town population and one-third of the city population. Given that they are concentrated mostly in the economically active age groups (Ma and Lhundup 2008, 14), we can safely assume that 80 percent of them were economically active. Indeed, most temporary migrants would eventually go home if they did not find employment. It is also safe to assume that they were working mostly outside agriculture and/or in urban areas, given their urbanization rate of 80 percent (albeit some were working as greenhouse vegetable farmers in suburban areas—see Yeh and Henderson 2008). Accordingly, their 6.1 percent population share in 2000 was in effect equivalent to more than 38 percent of total employment outside the primary sector (both urban and rural), and more than 55 percent of total urban employment (employment data from CSY 2001, table 5-4). Given that the Han were heavily concentrated in the city center of Lhasa, as analyzed by Yeh and Henderson (2008), their share of employment in this location would have been even greater (and less in more marginal towns). In other words, the Han population share

118

Chapter 3

is magnified many times over in terms of its share of urban and especially city employment. Considering that the census probably underestimated the actual Han Chinese migrant population, particularly during the summer months, this magnification in urban employment is also underestimated. Similarly, even the conservative measures of the 2010 census show an increase in the Han share from 6.1 percent to 8.2 percent. If we apply the same assumptions as above, this 8.2 percent share would have still been equivalent to 40 percent of total urban employment in 2010, despite the rapid transfer of local Tibetan labor out of agriculture (as discussed in chapter 5) and a more than doubling of the number of urban employed, from 226,000 people in 2000 to 531,000 people in 2010 (2010 data from CSY 2011, table 4-2). Under these circumstances, the increasing share of the Han in the overall population, even though from a small base, placed a hugely disproportionate pressure on the expansion of urban employment opportunities, particularly considering that the Han tend to dominate the most lucrative of these opportunities. Again, the likely underestimation of the number and share of the Han in the 2010 census emphasizes the gravity of this situation.

Conclusion: Urban Fault Lines of Migration and Exclusion In summary, Tibetans have been in demographic transition since the 1950s or 1960s, as have most other nationalities in China or, indeed, in Asia. While they lag behind the Han in this transition—perhaps by ten or twenty years, depending on the demographic variable considered—the lagging is partially due to differences in family planning policies since the 1980s and other conjunctural issues since the 1950s. However, their demographic trends show enough similarities with the Han to clearly dismiss simplistic theories that Tibetans are either poor because they are demographically backward, or else they are demographically backward because they are poor. Rather, they are poor and they are relatively well advanced along the demographic transition, to the extent that fertility rates are now below replacement levels. Nonetheless, the demographic transition does clarify some of the key sociological aspects of modernity facing Tibetans. Relatively rapid population growth is not solely the result of Chinese rule, although the extension of primary public health care in the 1950s and 1960s obviously contributed to the falling death rates that typically precipitate demographic transitions. Nor are falling fertility and birthrates solely the result of repressive birth control policies, given that these are also more generally linked to the remote cause of falling mortality rates and are accentuated by stagnating rural economies and urbanization.



Population Foundations of Marginalization in Tibet 119

Urbanization in turn sets the stage for understanding the crux of modern ethnic tensions within the Tibetan areas. The contemporary meeting points between nationalities are in the towns and cities, as they were in the past, although with the important difference that modern demographic and economic transitions place far greater pressures on urban areas than historically would have been the case. In the process, rising population density and rapid urban expansion can give the visual impression that the more urbanized or economically successful groups (Han or Muslims) are becoming more dominant even while they are only maintaining or even losing their share in the overall population. In this way, perceptions may be deceptive even while they fuel resentment and signal other underlying dilemmas of considerable gravity. In the Tibetan areas, these relate to the interaction between population and economic transitions within a context of rapid economic growth that is heavily skewed toward the urban areas. Yet this observation of urban bias tends to lead to the conclusion that the appearance of an ethnic bias in Tibet is merely a reflection of spatial bias in the unbalanced development of the region, as argued by Sautman and Eng (2001). Indeed, in line with the Chinese government, they argue that out-of-province migration brings a net economic benefit to Tibetans. As support, they refer to research from Canada on immigration proving this point in the Canadian context, albeit they ignore research on the migration of non-aboriginals into aboriginal regions in Canada, which would have been the much more appropriate comparison for the case of Tibet. There are a variety of problems with this argument. These are not necessarily related to the idea of spatiality. Spatiality is involved in complex ways in the patterns of development and social stratification in Tibet, as discussed by Yeh (2013), but not necessarily in simplistic urban-rural binary ways. Rather, Han Chinese dominance in the urban areas of Tibet is crucial because urban areas are the portals of political and economic power and thereby serve as the basis from which new stratifying and subordinating tendencies become articulated and are emanated in the rapidly changing regional economy. In particular, educational, political, and socio-cultural biases operating in urban employment, in a context where illiteracy is still very high among even permanent urban Tibetan residents, are arguably the dominant factors producing marginalization in contemporary development in Tibet. The ways that these biases are played out in urban labor markets has also led to acute forms of intra-urban inequality. In this respect, even though spatial polarization might distance rural dwellers from the portals of prosperity, the critical arenas where ethnic exclusion is played out in Tibet, for both rural and urban dwellers, are primarily urban. This is precisely where the political and economic levers of power are controlled, along with access to strategic or lucrative economic opportunities. Dynamics within

120

Chapter 3

the urban areas thereby structure the ethnically discriminatory patterns of development at both the lower and upper ends of the urban labor hierarchy. In this light, it is not correct to say that the urban bias is the cause of discrimination and marginalization in Tibet, although it definitely structures the way that discrimination and marginalization manifest, in contrast to other minority nationality regions in China, such as Xinjiang or Inner Mongolia, where rural populations are more ethnically mixed and minority populations are more likely to possess higher levels of schooling. Moreover, if locals in these Tibetan urban areas were able to use administrative means to protect themselves against migrant competition, as has been the case in other provinces in China, then perhaps this situation would be less contentious (although migration is also very contentious in other parts of China, for other reasons). However, local Tibetans do not have any of these means at their disposal given that their regions are effectively ruled by non-Tibetan outsiders, appointed by either Beijing in the case of the TAR or by provincial capitals in the case of Tibetan areas outside the TAR. In this context, Tibetan locals are left to play on a field that is unfortunately far from level, as is analyzed in detail in the rest of this book with respect to economic policy (next chapter), employment and inequality (chapter 5), and the education-employment nexus (chapter 6), making it clear how the average local Tibetan is easily out-competed by the average Han Chinese migrant. Thus, while probably misconceived, the perception of population swamping can be seen as a reactive lens through which locals interpret their concerns with modernity within a context of stark political disempowerment. In particular, the view voices a legitimate concern that in-migration might exacerbate economic exclusion among Tibetan locals in their towns and cities, particularly at a crucial stage of their population and economic transitions when such exclusions will no doubt heavily influence the longterm trajectories of ethnic stratification. Nevertheless, the logic of peripheral development would tend to counter the thesis of population swamping. If anything, population outflow, particularly among the educated, is an important development constraint in these areas. However, this symbiotic conflation of in-migration and exclusion seems to provide the discursive superstructure through which Tibetans understand their contemporary impasse within disempowered development.

Notes   1.  His view shows little change from previous official exile statements on population transfer and control, as in TGIE (1993; 1996).   2.  See the interview with Samdhong Rimpoche, then the prime minister of the Tibetan Government in Exile, in Mahalanobis (2003).



Population Foundations of Marginalization in Tibet 121

 3. See Kwok (2008) for a report on a government-organized foreign press conference in Beijing following the Tibetan protests in March 2008, when Tibetan researchers from the China Tibetology Research Centre insisted that claims of population invasion were exaggerated.   4.  It is not clear whether this reference to “permanent residents” is a typo, as often happens in these press releases of population data on Tibet, given that the 2000 and 2010 census population head counts in principle include all people de facto residing in a locality, including temporary migrants and those who have not resolved their status at the time of the census. If the head count does include only permanent residents, then all the data on population shares and rates of increase would be erroneous because the head count would not be comparable to that of the 2000 census, which did include temporary migrants. It would also be in transgression of the way populations are reported for the other provinces. Hence, the reference must be a typo given that the same head count in reported in CSY (2011). See more details in footnote 40.   5.  At the time of finalizing this book late 2012, the 2010 census had not yet been made publicly available beyond a few summary tabulated data in the CSY (2011).   6.  For instance, in PRC (2001a, Question 20), the government argues that after 1951, “helping the Tibetan people to develop their economy and improve their living conditions became the common concern of the central government and the Chinese people, and it is an important facet of China’s modernization drive.” Later in Question 56, it argues that “with the constant improvement in health care and living standards of Tibetan people, the long stagnant population growth of Tibet has seen a sharp increase.” Moreover, in Question 55, it contends that “[t]hese days Han residents in Tibet are mostly technicians, workers, teachers, medical professionals and officials from other provinces, municipalities and autonomous regions of China.”   7.  PRC (2001a, question 17); “With the support of the central government and people of other areas, Tibet has developed its pillar industries through utilization of its rich resources.”   8.  For more examples of these official arguments, see PRC (2001a, questions 17, 18, 20, 54, 55, 56, and 58).   9.  See Ma and Pan (1992); Xi et al. (1992); Andersen et al. (1995); Guo (1996); Ma (1996); Sun and Li (1996); Marshall and Cooke (1997); Iredale et al. (2001); Goldstein et al. (2002); Childs (2000; 2001a; 2001b; 2003; 2004; 2006; 2008); and Childs et al. (2005). 10.  See PD (2002) for an example of how these data are asserted in the official Chinese press. 11.  For instance, in PRC (2001a, question 22), the population of “Tibet” is said to have been one million in 1953. This estimate was in fact supplied by the Tibetan government in Lhasa, at the time headed by the Dalai Lama. This 1953 population of one million is then compared to the TAR population in 1964 of 1,250,000, implying a substantial population growth of 250,000 people (questions 55 and 56). Yet it is specified in question 22 that this population of one million excluded the prefecture of Chamdo, which the answer stipulates was placed under the jurisdiction of Tibet (i.e., TAR) in 1956. Therefore, the population of one million does not refer to the TAR but to a smaller territory, given that the TAR includes Chamdo.

122

Chapter 3

In the same question, they further specify that the 1953 population of the region including Chamdo was 1,270,000. This is the appropriate comparison to the 1964 population and surprisingly reveals that the population of the TAR, properly defined with the inclusion of Chamdo, actually fell by 20,000 people between the 1953 and 1964 censuses. Obviously, even this observation, if it is to carry any significance, depends on whether any of these estimates can be considered reasonably accurate. The estimate provided by the Tibetan government in 1953 was not based on a proper census but was the result of informed guesswork. Thus, most of the statements of the PRC on the nature of population changes in the TAR up to the 1960s are similarly based on such guesswork. See Childs (2008, 216-21) for an excellent discussion of these estimates produced by the Chinese government and ideologically constrained demographers in China, as well as by the Tibetan Government in Exile (with reference to the mythical figure of six million Tibetans). 12.  This is not to suggest that Beijing sees population growth as positive for development in Tibet. Rather, from another perspective, it argues that current poverty in Tibet is exacerbated by population growth and environmental degradation (e.g., see PRC 2001a, question 22). 13.  For instance, an Asian Development Bank report on Western China implicitly refers to Tibetans as a “pre-transitional” minority (ADB 2003, 268) and asserts that demographic transitions are caused by level of socioeconomic development, in apparent oblivion of contemporary demography. 14. See Childs (2008, 214-21) for an excellent discussion of historical demographic estimates for the TAR. He adeptly debunks the common hypothesis in the Chinese literature that the Tibetan population was shrinking, or at best stagnant, up to the 1950s due to exceptionally high death rates and low levels of fertility. 15.  See the discussion on mortality during this period in the previous chapter. 16.  This can be roughly represented by the fact that crude death rates fell below ten per thousand in the TAR and other western provinces by the middle to late 1960s. See figure 3.2 for data on the TAR and QSY (2005, Table 4-1) for data on Qinghai. On lags with China, life expectancy for all of China was 68.6 years in the 1990 census and 71.4 years in the 2000 census, whereas it was 59.6 years and 64.4 years respectively for the TAR (CSY 2007, table 4-6). UNICEF reported that (official) infant mortality stood at 53 per 1,000 live births in the TAR in 2000, and maternal mortality at over 400 per 100,000 live births, versus 35 per 1,000 and 56 per 100,000 in China overall in 2000 (UN 2004). Unofficially, several INGO health workers in the TAR whom I interviewed estimated that rates of infant and maternal mortality are much higher, perhaps more than double these official rates (also see Harris et al. 2001), placing them in the range of many sub-Saharan African countries. The official rates place the TAR within a range similar to Bhutan in 2000 (56 and 420 respectively), or lower than Nepal (64 and 740 in 2000), according to data in UN (2002). 17.  These data include urban areas and non-Tibetan populations. Alternatively, Childs (2008) analyzed village-level survey data collected by Goldstein et al. (2002) from an entirely rural Tibetan sample. These data show similar patterns, besides a peaking of fertility rates at 6.3 in 1986, after which they fell rapidly from 1989 onward. He emphasizes that the speed of the fertility transition took place simultaneously in both the rural areas of the TAR and the exile community, falling in an



Population Foundations of Marginalization in Tibet 123

almost identical manner, both from 6.3 in the late 1980s to below replacement by 2000 (237-39). 18.  For instance, the very low birthrates do not corroborate with the official fertility rates for this period, as discussed above, and the low death rate of 0.5 percent (versus 1 percent for China in 1965) does not make any sense given the socio-economic circumstances of the TAR at the time as well as much higher subsequent trends in the 1970s. 19.  For an analysis of the standard discourse by (officially sanctioned) demographers in China, see Childs (2008, 221-31). 20. See this argument in PRC (2001a, question 57). For general background on minority family planning policy, see Zhang (1990); Zou (1991); Yang and Liu (1992); Yang and Zhu (1993); and Greenhalgh (2003). 21.  For examples of these allegations, see TIN (1994) or TSGUK (no date). 22.  For instance, official statistics for Qinghai report a sharp drop in crude birthrates between 1980 and 1985, from 21.1 to 14.2 per thousand, but the rate then increased to 24.9 in 1986 and remained at that level until the late 1990s (QSY 2004, table 4-1; data not available from 1981 to 1984). The sharp drop in 1985 might reflect an aggressive application of family planning measures in the early 1980s, thus explaining the reports of abuse from the exile community in the 1980s and early 1990s. Alternatively, the data might simply reflect reporting problems. 23. Similarly, forced sterilizations and abortions implemented in India under Indira Gandhi in the 1970s also had very little apparent impact on overall levels of fertility, birthrates, or natural rates of increase in India (Livi-Bacci 2001, 154). 24.  Many of these sociological factors are discussed in Goldstein et al. (2002), Childs et al. (2005), and especially Childs (2008, 233-52). 25.  Minority autonomous areas (including Muslims areas and counties outside of prefectures) account for 98.9 percent of the land area of Qinghai. 26.  Communications from various contacts working in the area indicate that the mostly rural Tibetan and Qiang highlands in the earthquake zone were relatively unscathed by the earthquake. 27.  While there is debate about the causes of desertification in China (e.g., see Williams 1997), these mostly address the claim that pastoralism has been causing environmental degradation. It is generally accepted that agrarian land colonization schemes caused the degradation of high-altitude pastures in some Tibetan areas, particularly in Qinghai. 28.  For instance, agrarian colonization policies were practiced with some success in Dartsedo (Ch. Kangding) County, Ganzi TAP in Sichuan as early as the Qing dynasty, similar to Inner Mongolia (see Coleman 2002; van Spengen 2002; and Tuttle 2005). As a result, the county now contains a strong rural Han presence. Nonetheless, the population becomes indisputably Tibetan after the first pass west of Kangding Town, with almost no Han in the rural areas (confirmed with locals during fieldwork in the area in 2004). 29.  This observation comes from my field experience with Tibetan refugees in India and Nepal from 1995 to 2001 and from interviews with returnees during fieldwork in Tibet in 2003 and 2004. 30.  Iredale et al. (2001) argue that minority populations in China have been significantly less mobile than the Han. Their mobility started to increase in the 1990s, although much later than the Han.

124

Chapter 3

31.  Provincial population statistics do not include resident military populations, which are considerable in both the TAR and Qinghai. For instance, estimates for the military in the TAR in the 1990s ranged from anywhere between 40,000 to 200,000 or more (UNPO 1997, 70). Taking an estimate of 130,000, there was about one soldier for every 20 residents in 2000, versus one for every 4,500 residents in China overall. This is obviously related to the vast border area of the TAR alongside India. 32.  China stipulates a de jure approach to residency status, meaning that residency is recorded by officially registered status regardless of the amount of time spent in a location, in contrast to standard international definitions, which consider de facto residency. The Chinese approach is equivalent to an institutional norm of citizenship, such that a person can reside in a place for years without being considered permanent in the census so long as they have not made an official change in status. 33. Although the nationality and the residency status data are tabulated separately in Tabulation (2002), the residency status of the Han in the TAR can be extrapolated by comparing the total Han count in the 2000 census (158,570) with the nationality population records reported in the Tibet Statistical Yearbooks (e.g., TSY 2005, table 3-4). While the latter do not indicate any source, they presumably draw from the Public Security Department (PSD), which is the department responsible for residency registration, given that the numbers match those reported in CSY (2002, table 2-9), which sources its data from the PSD. By comparing this PSD data for 2000 to the 2000 census tabulation on residency status (Tabulation 2002, table 2-2), the PSD count for the TAR (2.51 million) falls between the permanently residing population (2.39 million) and the population both permanently residing and residing for more than six months but registered elsewhere (2.587 million). In other words, the PSD data probably uses a cut-off criteria that lies somewhere between long-term temporary residency and permanent residency. Moreover, these (apparently) PSD data in TSY (2005) are broken down by nationality, showing a more or less unchanging population of Han from 1985 to 2000, at 72,122 in 2000. Accordingly, we can presume that this is the population of permanent or longterm resident (non-military) Han in the TAR, and that the remaining 86,488 Han counted by the 2000 census were temporary residents. I would of course invite the government to be more transparent in order to avoid such convoluted calculations. Notably, when government officials claimed earlier in the 2000s that the Han population in the TAR had barely increased in recent years, as they often did in foreign press conferences, they were usually referring to these PSD data. See some mention of this in Ma and Lhundup (2008, 7). 34. See Yixing and Ma (2003). For an alternative perspective, see Goodkind (2004). Obviously, we can expect that the forthcoming 2010 census will add even further accuracy. 35.  Data on tourist numbers are compiled from Xinhua (2006), PD (2008), and Tibetinfonet (2011). Tourists would have exceeded the total population of the TAR of about 2.8 million in 2007. 36.  All ethnic shares remained constant after 2000, suggesting that the later data had been estimated simply on the basis of the 2000 census. 37. Some authors argue that part of the increases in minority populations in China have been related to the switching of ethnic status by Han in order to take advantage of various preferencial policies for minorities, particularly between the



Population Foundations of Marginalization in Tibet 125

1982 and 1990 censuses. See, for example, Zhang (1990), Zou (1991), Yang and Liu (1992), Yang and Zhu (1993), and Attané and Courbage (2000). However, most of these studies focus on more central ethnic groups, such as the Manchu, that have been much more assimilated into Han mainstream culture, thereby facilitating switching. The phenomenon was probably less pronounced in Tibetan areas than elsewhere in China. 38.  The detailed tabulations of the 2010 census had not yet been released as of the time of finalizing the manuscript in late 2012. The delay was in standing with the 2000 census, which took two years to be publicly released. 39.  See Fan (2005) for an excellent discussion of the changing spatial and temporal definitions of migrant populations between the 1990 and 2000 censuses. In addition, the 1990 census was conducted in the summer while the 2000 census was conducted in November, which would have been a strong source of discrepancy in provinces experiencing substantial seasonality in their migration flows. 40. There are some reasons to doubt the accuracy of the adjustments. For instance, several scholars and officials whom I interviewed in Qinghai in 2004 suggested that the adjustments in Qinghai—at least at the sub-provincial level— were in part the result of political negotiations. However, it is possible that these negotiations were based on fairly accurate post-census survey estimates at the aggregate provincial level. 41.  See note 4 for further discussion of this point and my reasons for assuming that Xinhua (2011a) erroneously suggested that the head count was for permanent residents, although this can only be confirmed once the detailed tabulation is released. However, as further evidence, the PSD data discussed in footnote 33 above show a slight increase in the Han population count after 2000, from 72,122 in 2000, to 105,379 in 2003, 93,306 in 2004, and then 123,558 by 2008 (TSY 2009, table 3-4; this table was not reported in TSY 2010). These numbers are still far below the Han count in the 2000 census of 158,570, let alone the Han count in the 2010 of about 245,277, and the increase of around 50,000 Han in the PSD data is far below the increase implied in Xinhua (2011a) of 86,707. Corresponding to the logic of footnote 33, we can presume that these PSD data refer to permanent (or long-term) residents, in which case the 2010 data must include temporary migrants. 42.  CTIC (2010a) reports that “1,900 graduates (10% of the graduates) from Tibet found jobs in inland areas [in 2009], and in the past two years, the proportion is about 8% to 9%.” This trend took off in 2006 with the reform of the employment system in the TAR, as discussed in chapter 6. Similarly, according to CTIC (2010b), the total enrolment of Tibetan students in inland high schools was 20,000 in 2010. Xinhua (2011b) indicates that about one-quarter of these students return to the TAR after graduation (i.e., 10,000 out of a total of 40,000 since the project started in 1985). If we assume that the 20,000 enrollment in 2010 was divided equally over the three years of high school and that the annual intake was thus about 6,667 students a year in 2010, although less in previous years given the intensification of the program over the decade, and that one-quarter of these students return to the TAR after graduation, we can speculate that these two channels of outflow of local Tibetans from the TAR could have amounted to at least 30,000 people between 2000 and 2010.

126

Chapter 3

43.  See chapter 8 for further details on the economic development of these two jurisdictions. 44.  Population statistics are generally considered to be fairly accurate in these less-populous jurisdictions. Unlike larger cities, they have limited points of transportation access and typically have only one main town per county. As a result, local governments usually have a good idea of who is entering or leaving their jurisdiction. Also, the government officials interviewed were not concerned with the implications of an increasing or decreasing Han population share. To the contrary, their paramount concern was to attract investment and skilled labor from other parts of Qinghai and China. 45.  Such out-migration was explained to me in interviews with ten Xunhua Tibetan high school students in Rebgong, one Xunhua Tibetan university student in Xining, and three Xunhua Tibetan high school teachers in Xunhua and Rebgong, all in June and July 2004.

4 Instituting Economic Growth and Marginalization in Tibet

As noted in the introduction, the development discourse of the CPC with regard to Tibet treats the particular mode of integration of the Tibetan areas into the national economy through the implementation of western development strategies as if there is no alternative. There is silence on the alternative modes of integration that might well have been possible within the context of CPC rule. This mode of integration that has dominated since the late 1980s has essentially accentuated the disempowering and assimilationist dynamics of development, even while providing for strong economic growth and the rapid buildup of certain types of infrastructure, that is, those conceived principally within a broader center-periphery relationship, wherein connections to the rest of China have been prioritized over widening and deepening the articulations of locally oriented infrastructure. In this context, marginalization can proceed apace with economic growth and improving livelihoods (livelihoods understood according to standard materialist conceptions). Whereas marginalization is commonly conceived as impoverishment and, hence, improving livelihoods as a refutation of the possibility of marginalization, development in Tibet since the 1990s suggests the need to rethink this common conception. This need for a rethinking is not new. For instance, marginalization in the contemporary Tibetan context is comparable to the way that increased levels of (modern) material consumption might have accompanied the integration of aboriginal communities in northern Canada over the course of the twentieth century, despite their progressive marginalization from the political and economic levers of power driving the development of these regions. Moreover, integration of aboriginals through the vehicle of the 127

128

Chapter 4

reservation system and residential schools, and associated forms of social dislocation and alienation, remain poorly accounted for in most conventional measures of development and not at all in measures of economic growth. This type of understanding is essential if we are to understand how rapid growth in the Tibetan areas has exacerbated the unresolved contestations with Chinese rule, as well as the possible range of alternatives that are conceivable in the current situation. This being said, direct policy analysis is extremely difficult to conduct in Tibetan areas given the high degree of sensitivity regarding policy matters, the lack of access to policy makers, and the inability to conduct surveys anywhere besides a pocket of select rural areas by vetted scholars (particularly in the TAR). Hence, this chapter takes an approach to analyzing these dynamics by examining the macro-structural outputs and outcomes of economic development in these areas as indirect de facto representations of the actual implementation of development strategies and related policies. Obviously, outcomes cannot be attributed entirely to policy implementation alone. They are also influenced by other factors such as business cycles, external economic environment, or technological or socio-economic transformations within wider economies. However, the sheer preponderance of policy-directed subsidies in the Tibetan areas allows for a particular methodological opportunity, in the sense that the connection between policy and outcome is much clearer to infer in comparison to other areas of China where endogenous or global economic factors would play a much greater role in determining economic outcomes. The focus on these outputs and outcomes is undertaken through an analysis of the structural economic changes in the Tibetan areas over the last twenty years and the heavily subsidized sources of these changes—namely, government expenditure and investment—followed by a discussion of the patterns of ownership underlying these sources. These dimensions are sufficient to sketch out the particular characteristics of growth and structural change in Tibet that are very distinct from those of all other regions of China, including western China. The subsequent chapter then focuses more precisely on changes in the labor/employment structure parallel to these macroeconomic changes, as well as household incomes and inequality, with the purpose of identifying the possible spaces where discrimination and/or exclusion might be operating in a systemic manner in regional and, in particular, urban labor markets. Chapter 6 unpacks the theme of urban exclusions further by looking at the education-employment nexus in the context of these changes, with the notable insight that these processes of marginalization and exclusion are not necessarily associated with material standards of living or with lower levels of socio-economic status. In terms of particular characteristics, the exceptional subsidy deluge in the TAR surged to record-high levels, particularly following the widespread



Instituting Economic Growth and Marginalization in Tibet 129

protests that occurred across all Tibetan areas in 2008 and surpassing the peaks of subsidization in the Maoist period. By 2010 (the latest data available at the time of writing), direct budgetary subsidies from the central government to the TAR local government exceeded 100 percent of the TAR gross domestic product (GDP) for the first time ever. This was about four times the average per capita rural household income in the TAR in 2010. Similarly, investment in fixed assets—most of it also probably subsidized—reached 91 percent of the TAR GDP in 2010. These data indicate the extent to which subsidies have been disconnected from conditions in the local economy, even though the minor portion of subsidies that does trickle down into Tibetan rural areas is nonetheless large enough, given the overall deluge, to cause a rapid degree of socio-economic transformation, as discussed in the following chapter. Within a context of continued political disempowerment of Tibetan locals, this centrally directed channeling of massive amounts of subsidies and subsidized investments (relative to the local economy) through the government itself or else through Chinese corporations and other entities based outside the Tibetan areas has accentuated the already highly externalized orientation of wealth flows in the economy. Subsidization strategies result in a form of what I have called “boomerang aid” (e.g., see Fischer 2009b). The TAR represents a quintessential example. Subsidies largely return to their sender while debilitating indigenously oriented forms of wealth creation and accumulation in most sectors of the economy outside agriculture. This reinforces a situation of extraordinary inefficiency and extreme dependence, as well as strong cultural, linguistic, and political biases stemming from characteristics of the dominant cultural and political group controlling the subsidies, such as Chinese fluency, Chinese work cultures, and connections to government or business networks in China proper. From this perspective, despite almost twenty years of intensive development efforts, the TAR remains locked into the institutional norms guiding the subsidization of this politically sensitive region since the Maoist period. As a result, recent development strategies have not altered in any significant way the long-term trend of very intense and very inefficient subsidization, with economic growth largely reflecting the intensification of subsidies. In contrast to every other province in China, these norms appear to have been predominantly guided by security and ideological concerns—the muchtouted tourism and mining activities still pale in comparison to government administration and related activities that are the main targets of the subsidization strategy and which do not appear to carry any particularly coherent economic rational. Rather, the recent phase of intensive subsidization has completed two principal tasks first envisaged during the Maoist era: integration into China and consolidation of the very visible hand of the state in the structuring of most aspects of the economy toward its strategic

130

Chapter 4

goals. As a result, the economy of the TAR can be aptly described in structural terms as having become a peripheral subsidiary of the central government and related interests. Local development dynamics (and people) are increasingly captive to the discretion of these central interests, particularly given their rapid transition away from the traditional bases of subsistence in the rural economy. These insights from the TAR possibly predict the course of future developments in other Tibetan areas outside the TAR given the stated intention of the Fifth Tibet Work Forum to start prioritizing all Tibetan areas, as noted previously. The disempowered and externalized aspects of this “development model” imposed on the region have thus exacerbated the discriminatory and assimilationist tendencies in the economy, intensifying exclusionary pressures despite rising living standards. These exclusionary pressures are most evident in the competition between local Tibetans and non-Tibetan migrants from other parts of China, as elaborated further in the following chapters. The main point of this chapter with regard to the discrimination question is to establish an understanding that, regardless of intention, discriminatory tendencies in the economy are more generally produced through the structural and institutional dynamics underlying development and occur alongside considerable improvements in the material consumption of local Tibetans. This argument is an important distinction from the contention that ethnic inequalities in the TAR and other Tibetan areas are mostly the result of spatial inequalities, that is, that Tibetans are mostly rural and poor, hence ethnic inequalities reflect urban-rural inequality, as argued by Sautman and Eng (2001) or Hu and Salazar (2008). Rather, inequalities differ from elsewhere in China in that they are instituted and structured through the perverse special treatment accorded to the Tibetan areas by the CPC, which results in a progressive appropriation of ownership in the local economy by outsiders, placing local Tibetans at a disadvantage despite the profusion of subsidies. The degree to which discrimination is specifically or systematically intended within these strategies is perhaps less important than the fact that the result is effectively discriminatory. In any case, the strong local Tibetan perception of discrimination offers important insights into the explosion of tensions in March 2008 and beyond, in addition to (and in interaction with) political factors. The method of this and the two subsequent chapters is comparative, in the sense of comparing the TAR with three other provinces (Gansu, Qinghai, and Sichuan) and China as a whole. The four provincial cases are chosen with purpose. Sichuan is the most important source of emigration to the TAR and to the Tibetan areas of Sichuan and Qinghai (as well as the most important source of emigration for the country as a whole, as noted in the previous chapter). Gansu is the second most important source, especially to the Tibetan areas of Qinghai and Gansu.1 Economic and education data in



Instituting Economic Growth and Marginalization in Tibet 131

these two provinces and the national average therefore give a broad indication of the average economic conditions and skill levels in the sources of emigration to the Tibetan areas. Sichuan also represents the political and economic center of southwest China and an important economy in its own right. Gansu is also an important center in northwest China, particularly for Qinghai, and it is also the second-poorest province in China, with topographic and population density characteristics more similar to the Tibetan areas than the densely populated provinces of southwest China (with the exception of the TAR, which is much more northwestern than southwestern in terms of population density and arid topography). Hence, Gansu serves as a good comparator for the Tibetan areas in many respects, even though it is also very different in other respects. Sichuan is half composed of Tibetan areas, although the population weight of these areas is so small that it renders them insignificant in the survey data. Tibetan areas account for about 10 percent of Gansu although, again, the Tibetan population weight is insignificant in the total population. In contrast, Qinghai is almost entirely composed of Tibetan areas, and Tibetans represent a substantial minority in the rural surveys of the province, mixed together with Chinese Muslim and Han Chinese, although the urban surveys of this province represent mostly a Han Chinese, and to a lesser extent Chinese Muslim, experience. The macroeconomic data of Qinghai are also more representative of its role as an industrial and resource extraction extension of Gansu rather than a Tibetan area per se. The provincial data from Qinghai therefore offer a mixed picture of Tibetan and non-Tibetan realities. The TAR is exceptional in terms of the politicized security focus and the extent of fiscal privilege that it receives (compared to other Tibetan areas and the rest of China) although, as noted in the introduction, it is otherwise similar to other Tibetan rural areas outside the TAR in terms of socio-economic structure, human development indicators such as health and education, population density, topography, and ecology. Hence, while exceptional in terms of the fiscal priority it receives, it can otherwise be used as a province-level proxy to broadly represent a rural Tibetan experience in comparison with other provincial and national cases. It can also be used as a proxy to represent the rapidity of social and economic change occurring across Tibet, which was perhaps sooner to start in the TAR, but that most Tibetan areas have been following since the early 2000s, with a variety of localized particularities. Indeed, with the signal in the Fifth Tibet Work Forum in 2010 to broaden the remit of “Tibet policy” to include all Tibetan areas, not just the TAR, even the urban experiences of the TAR might increasingly bear relevance to the Tibetan areas outside the TAR.2 There are obvious reasons for the very different economic starting points of these four provincial cases, particularly with respect to the share and structure of industry and mining. The remoteness and low population

132

Chapter 4

density of the TAR and other Tibetan areas has made them very unconducive to industrial activities outside natural resource extraction, as with most remote regions around the world (although remoteness and climate have made even mining very expensive and hence limited in the TAR). Nonetheless, the salience of such comparisons is with respect to the rapid changes that have been instituted by regional development strategies, especially with respect to the tertiary sector, and which are not necessarily related to topography or demography. The comparisons are not meant to be taken as normative judgments on the patterns of change (indeed, we might deem it preferable that the TAR remains largely unindustrialized), but simply to gain insight into how these patterns might have been shaping people’s lives, livelihoods, and economic opportunities in comparison to the more commonly understood patterns in China. In particular, we are interested in how policy-induced changes have either exacerbated pre-existing differences (or else have caused divergence in previous similarities), or how, despite obvious differences, the changes have brought about aspects of convergence and similarity across otherwise dissimilar cases. The chapter is divided into three sections. The first reviews the economic takeoff of the TAR and other western provinces from the mid-1990s onward. The second examines the heavily subsidized sources of this takeoff through an analysis of government expenditure and investment. The last section explores further the externally dependent and heavily importdependent nature of the economic model in the TAR, alongside the institutional characteristics of externalized ownership that underly this model. The conclusion returns to the implications of ethnic discrimination.

From Bust to Boom: Takeoff into Subsidy-Sustaining Growth in Tibet The speed of recent economic growth in the Tibetan areas has been phenomenal, definitely by international standards and even by recent Chinese standards. For instance, the per capita GDP of the TAR measured in constant value (i.e., discounted for the rate of inflation) started to grow faster than 10 percent a year from 1997 onward. From 1997 to 2010, the nominal aggregate GDP of the TAR increased by more than six and a half times, versus a general rate of CPI price inflation of only 23 percent between these years (hence most of the change was “real,” not inflationary). The GDP of Qinghai, the province with the next-highest proportion of Tibetans in its population and 97.2 percent composed of Tibetan autonomous areas, increased by about the same multiple. In comparison, the economy of China increased by five times, which was perhaps the fastest (and definitely the



Instituting Economic Growth and Marginalization in Tibet 133

largest) experience of sustained rapid economic growth the world had ever seen. Other western provinces in China grew at similar rates as China. In the TAR and in Qinghai, which together account for about three-quarters of the Tibetan population and Tibetan autonomous areas in China, economic growth had been almost one-third faster. This had not always been the case. As discussed in chapter 2, the TAR and Qinghai had been among the worst cases of economic lagging in China from the beginning of the reform period in the late 1970s until the mid-1990s, together with Ningxia and Gansu, two other northwest provinces with important concentrations of Chinese Muslims (along with Qinghai). Given that this lagging had real implications in terms of economic stagnation and the possibility of falling living standards, these realities were not taken lightly. In response, several decisive policy initiatives were taken by Beijing from 1994 onward in order to propel the TAR economy back toward the per capita national average from which it had been lagging. As noted in the introduction, these included: the national 8-7 poverty reduction plan in 1994, complemented in the TAR by the Third Tibet Work Forum; the focus on western development in the Ninth Five-Year Plan (1996–2000); and the “Open the West” campaign (OWC; xibu da kaifa) announced in 1999, which complemented the Tenth Five-Year Plan and was supported in the TAR by the Fourth Tibet Work Forum in 2001. To a certain extent, such prioritization of the west was already planned as early as the Seventh Five-Year Plan (Yang 1997, 28). The tax reform of 1994 and related fiscal reforms in the two subsequent years also facilitated such initiatives by crucially shifting a large proportion of tax revenues from the provinces back into the hands of the central government, allowing the latter to finance national redistributive strategies. Thus from the mid-1990s onward, spending and investment increased in the western region. Economic growth rates (in real per capita terms) picked up and in many instances surpassed national growth rates (see figure 4.1).3 Higher than national average per capita GDP growth rates started in the TAR as early as 1996 and in Qinghai by 1998. The reversal of the lagging was also aided partly by bumper crops in 1996, which coincided with a brief rise in the terms of trade of agricultural goods (relative to industrial goods) lasting from 1994 to 1997. Together these had a disproportionately positive impact on the more agrarian western provinces, inverse to the situation from 1985 to 1995 discussed in chapter 2.4 Hence, rapid growth was already well under way in both the TAR and Qinghai on the eve of the OWC. The OWC then maintained rapid growth (in the TAR) or accelerated it further (in Qinghai). The buoying of the west (see figure 4.2 for the real per capita GDP values) was matched with a tempering of the precocious takeoff of southeast

Sources: calculated from CSY (2011, tables 2-1, 2-15 and 9-5) and equivalent in previous CSYs.

Figure 4.1: Real per capita GDP growth rates, selected provinces, 1995-2010

Sources: calculated from CSY (2011, tables 2-1, 2-15 and 9-5) and equivalent in previous CSYs.

Figure 4.2: Real per capita GDP, selected provinces, 1995-2010 (constant 2010 Yuan)

136

Chapter 4

China, as shown in figure 4.3. Among the western provinces, the TAR accelerated faster than the rest in the late 1990s and early 2000s through a huge increase in subsidies and subsidized investments, but then fell back into the pack by the mid-2000s (despite continuously increasing subsidies, as discussed below). In nominal terms, the per capita GDP of the TAR reached 17,319 yuan in 2010, versus 29,992 yuan for China as a whole. Several other western provinces grew at a similar, if not faster, pace from the early to mid-2000s onward. For instance, the real per capita GDP growth rates of Sichuan surpassed 20 percent in 2004 and 18 percent in 2010, reaching a per capita GDP of 21,182 yuan by 2010. If accurate, these growth rates of Sichuan are especially phenomenal given the provincial population of over eighty million people as per the 2010 census. Figure 4.3 reveals several differences among western provinces in terms of their changes in per capita GDP relative to the national average, albeit all were growing rapidly throughout this period. The TAR fell from 80 percent of the national average per capita GDP in 1991 to 48 percent by 1995 (it was in recession throughout these years), but then quickly reclosed the gap with the national average from 1997 to 2002, when it plateaued at 64 percent of the national average per capita GDP. After 2005, the TAR started to lag again as the rest of China was galloping under the boom-time conditions that prevailed before the onset of the global financial crisis in 2008, and its per capita GDP fell back to 58 percent of the national average by 2010. Other western provinces, such as Qinghai, Sichuan, or Gansu, more or less maintained a steady gap with the national average throughout this period, with some variations. Xinjiang fell from an above national average per capita GDP down to the level of Qinghai by 2010. Yunnan lagged from a lower position, converging with the TAR and Gansu (and Guizhou from below) and became the second-poorest province for the first time in 2007 and then again in 2010. Others made significant gains, particularly from 2005 onward, such as Ningxia or Shaanxi (the latter not shown). Inner Mongolia, which was provisionally included as “western” under the OWC in 1999, was catapulted after 2002 to one of the highest levels of per capita GDP in the country, surpassing that of both Fujian and Guangdong by 2010. This was probably due to its close integration into the north China hub developing around Beijing and Tianjin and to the intensive development of mining. Overall, the shuffling of rank orders among the western provinces was minor, besides the relative takeoff of Inner Mongolia (which was never really western proper). The lagging of the 1980s and early 1990s was stalled as western China joined the national growth experience, although the region generally remained poorer than the national average. Moreover, although the growth of the TAR economy had been phenomenal, in relative terms it more or less remained in—or fell back into—its position as one among a cluster of the four (relatively) poorest provinces

Sources: calculated from CSY (2011, tables 2-1, 2-15 and 9-5) and equivalent in previous CSYs.

Figure 4.3: Provincial to national real per capita GDP, 1995-2010 (constant 2010 Yuan)

138

Chapter 4

in China (shared with Gansu, Guizhou, and more recently with Yunnan). This was despite the intentions-publicly-stated by government officials and matched with subsidies—to raise the per capita GDP of the TAR to the national average within this time period. The TAR diverged in this sense from the experience of Sichuan and Ningxia after 2004, up to which point the per capita GDP of these provinces were more or less on par with each other. This re-emerging (relative) lagging of the TAR occurred despite the exceptionally high and increasing levels of subsidies received by the TAR over these years, as discussed later in this chapter. The discrepancy is suggestive of the particular character of growth in the TAR economy, including a declining capacity of subsidies to maintain economic growth at the same rate as in other provinces with much smaller degrees of subsidization. Indeed, this problem has been debated among Chinese scholars and policy makers since the 1980s, as noted by Yan (2000) and Dreyer (2003) and epitomized by the groundbreaking contribution by Wang and Bai (1991 [1986]). The problem, however, appears to have continued to intensify following the implementation of the OWC and subsequent strategies since 2000. The structural transformations that accompanied rapid growth in the TAR (and that are plausibly representative, to a certain degree, of transformations in other Tibetan areas outside the TAR) offer some insight into this character of growth, particularly in comparison to other provinces. Rapid growth in the TAR was dislocated from productive sectors, particularly the primary sector (mostly farming and herding)—which is predictable in a context of industrialization in the wider economy—but also from manufacturing and mining within the secondary sector. The primary sector was the largest sector in value-added GDP terms up to 1996, when it accounted for 42 percent of GDP in the TAR and employed just over three-quarters of the labor force (almost all Tibetan). In contrast, the tertiary sector (a combination of government and party administration, social services such as education and health, trade, hotels and catering, transport, and other services) accounted for 41 percent of GDP and only 19 percent of employment (much of which would have been non-Tibetan). Between 1997 (when the tertiary sector surpassed the share of the primary sector for the first time) and 2010, the aggregate GDP increased by 559 percent in nominal terms,5 whereas the primary sector increased by 136 percent over the same period, thus decreasing in GDP share to 31 percent by 2000 and 13.5 percent by 2010, even while still employing 53 percent of the labor force in 2010. The secondary sector increased 867 percent between 1997 and 2010, albeit from a small base. The secondary sector increase was due largely to construction, which increased 1,308 percent, increasing in share from 11.5 percent in 1997 (or from a brief previous peak of 17 percent in 1995) to 24.5 percent in 2010. This was not matched, however, by “industry” (i.e., manufacturing and mining activity), which increased 389 percent (less than overall GDP), hence declining in GDP share from 10.6



Instituting Economic Growth and Marginalization in Tibet 139

percent in 1997 to 7.8 percent in 2010. The tertiary sector increased by 791 percent, increasing in GDP share from 40 percent in 1997 to 54 percent in 2010. Figure 4.4 gives a summary of these structural changes in the TAR (i.e., sectoral shares of GDP rather than nominal GDP values). Besides the sharp drop in the share of the primary sector, there are two other observations worth highlighting in figure 4.4. One is the divergence between construction and industry/mining from 1997 onward and especially from 2002 onward (2003 was the year that the construction of the Qinghai—TAR railway entered the TAR on a large scale). The value-added of construction surpassed that of the primary sector in 2006 and was over three times the value-added of industry and mining by 2010, despite the often-reported surge in mining activities in the TAR that accompanied the opening of the railway in 2006 (and was speculated by many to have been one of the major motivating factors for the railway, although these GDP data put that claim into question). The value-added of construction was only a fraction of industry and mining in every other province of China, despite the evident construction boom that had been sweeping the nation over these years. For instance, construction activity amounted to only 4.6 percent of the GDP of Beijing in 2007, when it was apparently in the grip of a construction frenzy in preparation for the 2008 Olympics. By 2010, when the housing market was often said to have been overheating in Beijing, Shanghai, and other coastal cities, construction activity had fallen to 4.4 percent of the GDP of Beijing and only 4 percent of the GDP of Shanghai. The GDP shares of construction in the TAR have been without parallel. Figure 4.4: Structure of the TAR GDP, 1995-2010

Sources: calculated from CSY (2011, table 2-15) and equivalent in previous CSYs. Data for 1995 is from TSY (2003, Table 1-12).

140

Chapter 4

While the increase in construction in the TAR was disassociated from other secondary sector activities, it was closely associated with the tertiary sector. The tertiary sector became by far the largest sector of the TAR from the late 1990s onward, surging in particular from 1997 to 2002, when it increased in GDP share from 40 percent to 55 percent in as little as five years. It also contributed almost the entirety of GDP growth in the TAR in certain years, contributing 80 percent of GDP increase in 1996, 87 percent in 2002, and 73 percent in 2005 (despite the ongoing railway construction in that year). In most other years it contributed well over half of GDP increase, except in 2003 and 2009 when huge surges in construction activity dominated the GDP increase (e.g., the railway in 2003). This means that the sector was not only large (in terms of GDP share) and growing fast, but was also rapidly expanding its share of GDP in these years, in contrast to agriculture, which was large but contributing little to GDP growth (and hence shrinking in share). In China and most other western provinces, the largest sectoral contribution generally came from industry and mining throughout this period. Even in the next most subsidized western provinces such as Qinghai and Gansu, industry re-emerged as the leading contributor to growth after a short surge of subsidized construction and tertiary activities in the late 1990s and early 2000s, suggesting that much of the construction activity was supporting industrial restructuring. This comparative perspective can be summarized by examining these structural changes across the five cases mentioned above— the TAR, Qinghai, Gansu, Sichuan, and the national average—in four figures: one for the primary sector, two for the secondary sector (one for construction and one for manufacturing and mining), and one for the tertiary sector. The comparison of trends in primary sector shares of GDP is presented in figure 4.5. The degree and speed with which the TAR converged with the national pattern is striking. The TAR started at a share that was more than double the national average and almost 50 percent higher than that of Sichuan in 1995, but then fell below the share of Sichuan from 2004 onward and even below that of Gansu from 2009 onward (note that the primary shares in Qinghai and Gansu were lower in the beginning of this period than in the more densely populated southwest region because of the legacy of heavy industrialization and urbanization in these two provinces, combined with the impoverished agrarian sector in the more densely populated—largely Han—parts of these two provinces). The pace of economic structural change in China is already considered to be rapid by international comparisons. The pace in the TAR (and presumably other Tibetan areas) has been astounding, albeit this has been due largely to the accentuated subsidized boom in the non-agrarian parts of the economy. Nonetheless, as discussed in the next chapter, such structural changes in the TAR GDP have been paralleled by equally dramatic shifts in the employment structure of the province and, hence, in the livelihoods of ordinary Tibetans in rural areas.



Instituting Economic Growth and Marginalization in Tibet 141

Figure 4.5: Primary sector share of GDP, selected provinces, 1995-2010

Sources: same as figure 4.4. Shares are calculated in current values.

The decomposed GDP shares of the secondary sector are presented in figures 4.6a (for industry and mining) and 4.6b (for construction). As noted above, the trends reflect the intensive industrial restructuring that was started in earnest from the beginning of western development strategies in the mid-1990s and in particular under the OWC from 2000 onward. Much of the subsidization strategies in these western provinces were focused on modernizing the antiquated industrial base left over from Maoist interior industrialization strategies of the 1960s and 1970s. In most cases, intensive subsidization and construction activity bolstered the leading role of industry within a few years. This is especially apparent in the remarkable industrial takeoff in both Qinghai and Sichuan from 2002 onward, where the share of industry and mining surpassed the national average from 2007 onward in Qinghai and 2009 onward in Sichuan. Gansu represents perhaps one of the most intensive cases of industrial restructuring, with a significant decline in industrial GDP share from 1995 to 2001 despite starting at the national average share, probably due to a particularly inefficient legacy of large state-owned heavy industries. In China as a whole, secondary industry (including mining, but only as a very minor share) was generally the largest sector driving growth throughout the 1990s and 2000s, amounting to over 40 percent of GDP throughout. These observations are confirmed by the shares of construction shown in figure 4.6b. Huge surges in construction as shares of total GDP ocurred over these years in Gansu, Qinghai, and Sichuan, peaking in 2003 in all three cases and then converging back toward the national norm thereafter. As with industry, the national trend in the construction GDP share was more or less

142

Chapter 4

Figure 4.6a: Industry and mining GDP share, selected provinces, 1995-2010

Sources: same as figure 4.4. Shares are calculated in current values.

Figure 4.6b: Construction GDP share, selected provinces, 1995-2010

Sources: same as figure 4.4. Shares are calculated in current values.

stable throughout this period at around 6 percent of GDP, despite the evident construction boom taking place in China and talk of property bubbles in coastal China, as discussed above. Such booms and bubbles were evidently riding off strong performance in other parts of the economy, such that they produced no obvious structural shifts in GDP shares. In contrast, the divergence between industry and construction in the TAR is unique. Early surges of construction activity in the late 1990s and early 2000s, similar to those of Qinghai in GDP share, did not lead to a revival of secondary industry, which continued to linger in the range of 7 to 8 percent



Instituting Economic Growth and Marginalization in Tibet 143

of GDP. Instead, the construction share in the TAR continued to increase, surpassing the heights reached during the boom related to the construction of the railway in the TAR in 2003 and 2004, and reaching almost 25 percent of GDP in 2010—a height never before observed in any western province, not even in the TAR. That such an increase was completely divorced from the share of industry—increasingly so for a period of fifteen years—belies the government claims that a restructuring of the local economy, as in other western provinces, was necessary to place the economy of the TAR on a more sustainable footing. The perpetual discourse of economic restructuring in the TAR appears to be much less oriented toward creating a sustainable local economic structure but, instead, toward other objectives that are quite distinct from those guiding regional development policy in other western provinces. These objectives concern security priorities above all. Similar to the case of construction, the contribution of the tertiary sector to GDP growth since the mid-1990s has also been somewhat exceptional in the TAR (see figure 4.7). Nationally, the role of the tertiary sector gradually increased to just above 40 percent of GDP in the 2000s, albeit some of the increase might have been due to corrections to national accounting data following a study in 2005 that demonstrated that tertiary sector activity in China had been underestimated up to that point. Various western provinces followed similar patterns, although at lower tertiary sector shares. Qinghai, which like the TAR had proportionately larger tertiary and construction sectors than other western provinces and even China in the late 1990s, subsequently fell below the national norm in the early 2000s, becoming proportionately almost identical to Sichuan in Figure 4.7: Tertiary sector share of GDP, selected provinces, 1995-2010

Sources: same as figure 4.4. Shares are calculated in current values.

144

Chapter 4

the last half of the decade. The disproportionate role of the tertiary sector in Qinghai in the late 1990s appears related to the intensive industrial restructuring that was taking place at that time. It then dissipated with the industrial takeoff in 2000s. In contrast, the tertiary sector in the TAR started from a fairly normal share in 1995 and then became the leading sector of growth for much of the next fifteen years, stabilizing at a share of around 55 percent of GDP from 2005 onward, or well over ten percentage points more than the national average share. The divergence with the secondary category of industry was very similar to construction, suggesting that construction in the TAR was much more linked to tertiary rather than productive activities over these years. This disconnect is often explained by the government as an aspect of its new strategy for the TAR to move away from past strenuous efforts to establish an industrial base and to concentrate instead on tertiary industries (such as tourism) as the new lead pillars of growth (possibly with mining and energy as well). Indeed, the overwhelming role of the tertiary sector begs for further analysis of its composition, although this has become more difficult to do after 2003 given that the National Bureau of Statistics (NBS) ceased publishing detailed decompositions of the provincial tertiary GDP data in both national and provincial statistical yearbooks, other than for a limited number of six categories including transport, trade, hotels and catering, finance, real estate, and the composite category of “others.” “Others” includes the major categories of social services, health, education, and government administration, which typically accounted for up to 50 percent of tertiary sector activity in most western provinces as of the 2003 data. Hence, the lumping together of these categories limits our ability to analyze important dynamics within the provincial tertiary sectors. It is even difficult to know the precise trends in the proportions of various tertiary subcategories in the national data because the reporting of some of these categories was changed in the mid-2000s.6 Nonetheless, the national shares of these subcategories within the tertiary sector appear to have been generally quite stable. Nationally, the share of “government and party agencies” (or “public management and social organizations” since 2004) gradually rose from about 7 percent of the tertiary sector in 1995 to about 10 percent in 2010.7 Transport, trade, and hotels and catering remained at fairly stable shares throughout this period, transport and related activities at just under 20 percent of the tertiary GDP, and trade, hotels, and catering at around 25 percent. Health care, social security, and social welfare accounted for around 3 to 4 percent of tertiary GDP during this period, and education for around 6 to 7 percent. Considering that the tertiary sector was a fairly stable share of national GDP from the early 2000s onward, these tertiary subcategories would have also been fairly stable with respect to their share of overall GDP. The patterns were similar in most western provinces



Instituting Economic Growth and Marginalization in Tibet 145

up to 2003, when the data ceased to be reported for many subcategories. For instance, the structure of the tertiary sector in Sichuan was quite similar to the national average. Both Qinghai and Xinjiang had substantially higher GDP shares of “government and party agencies” than the national or general western provincial norm. For instance, the share of government agencies in Qinghai increased strongly between 1995 and 2001, when it stabilized at just over 18 percent of the tertiary sector. In this respect, the composition of the tertiary sector in the TAR up to 2003—as well as the degree of change in this composition—again contrasted with the rest of China. While the share of government and party agencies in the tertiary sector of the TAR was always the highest in China, at around 20 percent in the mid-1990s, it surged in 2000 and 2001 to over 26 percent, becoming the largest component of the tertiary sector in those two years. Considering that the tertiary sector in the TAR had surged to almost 50 percent of GDP in 2001, government administration accounted for over 13 percent of total GDP in the same year, or almost twice the value of mining and industrial activity and close to the value of construction activity. Government administration had effectively become the engine of growth in the opening years of the OWC in the TAR, growing 68 percent in 2000 and 28 percent in 2001. By 2003 it stabilized at 11 percent of GDP and was surpassed by what might be called social services broadly defined (i.e., health, education, and social services). In comparison, government administration accounted for 7.5 percent of total GDP in Qinghai in 2003, and only 2.3 percent of total GDP in China (the national share increased to 4.4 percent by 2009, as calculated from CSY 2011, table 2-11). The unusually large and rapidly growing category of government administration in the TAR (as well as in Qinghai and Xinjiang, where it also grew rapidly in 2000 and 2001, in Xinjiang by 63 percent in 2001) probably indirectly reflects—in part—the relatively large military and/or security presence in these provinces and possibly a strengthening of this presence in the opening years of the OWC. If the military were included in the GDP statistics (which they are not) they would mostly tend to show up as part of the government administration category. Conversely, the sudden sharp increase in the government administration category of the GDP might imply a military buildup given the corresponding need for civilian administration to support such activities. The increase in the initial years of the OWC therefore probably indicates that an expansion of the control apparatus of the state was seen as a precondition to subsequent spending and investment in the TAR. This is of course a matter of speculation, as military activity is one of the most closely guarded secrets in China. However, the inference is supported by observations of expanding military facilities around Lhasa and along the highway leading to airport at the time. The Qinghai-TAR railroad project can also be best understood in this light, rather than any argument

146

Chapter 4

concerning its economic potential or viability. As in the past, military concerns probably guide much of the TAR development policies.8 It is difficult to judge these precise structural dynamics since 2003 given the lack of detailed reporting of provincial tertiary GDP data after 2003. Nonetheless, indirect indicators suggest that government administration continued to play a leading role throughout the 2000s, probably more than even tourism, which was nonetheless skyrocketing in the 2000s.9 This is indicated by the residual category of “other” reported in CSY (2005) onward, which does not include the categories related to tourism (such as hotels and catering, or transport and trade) and, judging from earlier data, would be dominated by government administration and various social services (e.g., health, education, and social welfare). This residual category accounted for 60 percent of the TAR tertiary sector in 2010 or 32 percent of total GDP, versus 17 percent of total GDP in Qinghai and 16 percent in Sichuan. A substantial proportion of this share of “other” most likely continued to be driven by government administration (or public management) in the TAR given indications provided by the TAR government expenditure data (presuming that most of the GDP value of government administration comes directly from government expenditure). For instance, the share of TAR government expenditure spent on government administration rose from 13.8 percent in 2003 to 16.1 percent in 2006 (calculated from CSY 2007, table 8-15 and the equivalent in CSY 2004), which is an increasing share of a rapidly increasing nominal (yuan) value of government spending. Government spending on government administration ceased to be reported as an identifiable category from CSY (2008) onward (presumably it was absorbed into other categories), although in comparison TAR government spending on education rose from 9 percent of total expenditure in 2003 to 10.3 percent in 2006 and then 11 percent by 2010. TAR government spending on medical and health care rose from 3.6 percent of total expenditure in 2003 to 4 percent in 2006 and 5.8 percent in 2010. Given that these mild increases in share were minor in comparison to the increase in the tertiary GDP share of the “other” category from 45 percent in 2004 to 60 percent in 2010, it is likely that the share of government administration in total government expenditure would have at least stabilized at around 16 percent, if not continued to increase in line with rapidly rising salaries in the public sector (which are the main cost factor of such expenditure). Given that the share of the tertiary sector in the total GDP remained at around 55 percent (the level it reached in 2002) and that government expenditure as a proportion of GDP in the TAR rose from 64 percent in 2006 to over 100 percent in 2009 and 2010 (as discussed below), it is very likely that the category of government administration (or public management) would have maintained a share of at least 13 percent of total GDP up to 2010, if not more. From this extrapolation, although tourism had been



Instituting Economic Growth and Marginalization in Tibet 147

booming throughout this period and undoubtedly fueled the urban private sector, tourism would have nonetheless played a secondary role in driving overall GDP growth next to government administration (and related military and paramilitary activities not included in these statistics). Indeed, despite the hype about the railway and tourism in the TAR, it is actually surprising that the tertiary categories of transport, trade, and hotels and catering (the main categories under which tourism would be reported) were actually falling in share over these years and were lower than the typical shares for these categories in other western provinces or in China as a whole. For instance, despite a surge from 2002 to 2004, probably related to the railway construction, transport and related activities fell from 13.5 percent of the tertiary GDP in 1996 to 8 percent in 2010 (although some of this reduction might have come from changing methods of calculating this category in the mid-2000s). We can presume from the above analysis that this share was far lower than the share of government administration (23 percent in 2002), and lower than even education and related activities (12 percent in 2002). Similarly, hotel and catering services amounted to just over 10 percent of the tertiary GDP in 2007, before the impact of the protests in spring 2008 that slowed tourism for a season. Despite the subsequent rebound in tourism, the share of hotel and catering fell to 5.7 percent of the tertiary sector by 2010. The share of wholesale and retail trade also fell throughout this period, from 30 percent of tertiary GDP in 1996 (including hotels and catering), to around 17 percent in 2010. Real estate had made large gains in share up to 2008, when it exceptionally accounted for 14 percent of tertiary GDP, perhaps as a spin-off of various housing projects by the government (see Yeh 2013), although it fell back down to 5 percent in 2010. Finance made great gains after 2008, rising from 3.4 percent of the tertiary GDP in 2008 to 9.9 percent in 2010, perhaps because one of the main levers that the government used to respond to both protests and the threat of economic slowdown was through the financial system (as elsewhere in China). However, bank loans were a very minor source of funding for investment in the TAR, as analyzed in the next section, so if at all accurate, this increasing share of finance must represent financial intermediation directed toward consumption and housing. Finance reached a previous one-off spike in tertiary GDP share of 14 percent in 2004; this might have been similarly related to urban housing projects, as with the parallel increases in real estate, or else to the financial needs associated with rapid wage increases among urban staff and workers. In sum, most of the growth generated in the TAR over these years derived from an alternating sequencing between tertiary activities (dominated by government administration) and construction (dominated by large construction projects such as the various components of the Qing-

148

Chapter 4

hai–Tibet railway). Both of these drivers were determined mostly by policies of subsidized spending and investment dictated in Beijing and, to a much lesser extent, supported by various rich coastal provinces in China. In this sense, growth in the TAR has been instituted largely through these policies, rather than being at least partly rooted in endogenous domestic processes as in most other provinces. Given the enormous weight of these instituted sources of growth in the local economy, as discussed in the next section, changes in the economic structure of the TAR have been much more radical and volatile than elsewhere in China, including the next most resembling province of Qinghai.

Sources of Subsidy-SustainED Growth in the TAR The magnitude of these drivers relative to the local economy in the TAR is worth emphasizing. The extremely high and increasing magnitude of direct subsidies in the TAR—from an already high level in the mid-1990s—almost defies logic. They increased so fast that they came to exceed total GDP in 2010 for the first time since the creation of the TAR in 1965. Such a high level of direct budgetary subsidies relative to local GDP has never been observed before in the TAR, even in the Maoist period. Indeed, the extent to which subsidies rebounded following a short and slight respite in 2006, and particularly since the protests in 2008, exceeded even my own expectations in my earlier work (e.g., see Fischer 2005a; 2009b). Moreover, this measure of direct subsidies does not include indirect subsidies, such as subsidized investments, which also rose rapidly as a proportion of GDP. To a certain degree, this extremity is unique to the TAR and is one of the principal differences between the TAR economy and the economies of other Tibetan areas in other provinces. However, these patterns in the TAR could well be replicated to a certain degree in other Tibetan areas under the Twelfth Five-Year Plan given that the government signaled in the Fifth Tibet Work Forum in 2010 that it would start focusing on all Tibetan areas and not just those of the TAR. These two aspects of subsidy (direct and indirect) are discussed below according to budgetary expenditure and investment. Subsidized Government Budgetary Expenditure For clarification, “direct subsidies” here refer to direct budgetary support from Beijing to the provinces in order to finance provincial deficits (provincial government revenues minus expenditures). Such deficits effectively proxy for fiscal transfer payments given that, up until recently or with some exceptions in coastal China, Chinese provincial governments have had a limited ability to borrow in order to finance public expenditures, such as



Instituting Economic Growth and Marginalization in Tibet 149

through bond financing or borrowing from banks, although this appears to have changed somewhat since 2008 in response to the global financial crisis in that year. Local state control over local financial institutions has nonetheless allowed for a substantial degree of indirect financing for the investment projects of state-owned enterprises (SOEs) involved in infrastructure construction and urban development, although such debt is presumably not calculated as part of government finances. Moreover, we can presume that these extra sources of local government financing were very limited in the TAR, for reasons stated in the section on investment below. This being said, the use of local government deficits as a proxy for subsidies is a blunt and approximate instrument given that much of provincial deficits are perhaps better understood as the result of the negotiated compromise between central and provincial authorities over the control of fiscal revenues, as in any country. The extent to which central revenues are themselves raised within a province is a key consideration to decipher the extent to which the provincial deficit can be used as a proxy for central subsidies. For instance, if the central government raises tax revenue in Fujian and then transfers part of that revenue to the Fujian provincial government, the transfer is not a subsidy, not in a regional sense. Rather, it is a representation of the institutional dynamics between various strata of fiscal authority overlapping within a particular jurisdiction.10 The central government transfer would only be a subsidy proper if it involves a transfer of revenue raised in another province or centrally. In the more affluent coastal provinces, local government deficits for the most part would represent a particular balance of power between central and local authorities with respect to control over locally generated revenues. However, much of the provincial deficits in western provinces would represent subsidies, in the sense of transfers of revenue raised outside the province. In principle, the share of subsidy in the total deficit would be greater in the poorer provinces given their lower capacity to raise revenue, and it would increase as the magnitude of deficit increases relative to the provincial economy. Indeed, all provinces in China were running deficits in the late 2000s, including Shanghai, Guangdong, and other coastal provinces. This was partly due to the government response to the global financial crisis and has since been one of the causes for concern about the sustainability of local government finances in China, in addition to the debt loads that local SOEs rapidly built up since 2008. However, at the aggregate level, provincial deficits were mostly covered by central surpluses. For instance, the total fiscal balance for all provincial governments was in deficit by 5.7 percent of national GDP in 2007, 6.9 percent in 2008, and 8.3 percent in 2009 and 2010. This was compensated for by a central government fiscal surplus, which increased from 6.2 percent of GDP in 2008 to 6.6 percent in 2010. Central and local fiscal authorities thereby play overlapping and complementary roles.

150

Chapter 4

In other words, the central government has come to play a major role in regional fiscal mediation, raising just over half of total central and provincial revenue in 2010 but spending only 18 percent of total central and provincial expenditure in 2010, whereas provincial governments spent 82 percent. Provincial governments have thereby become dependent on centrally coordinated transfer payments in order to balance their budgets, which in turn has become an important compulsion for local governments to seek revenues that they are mandated to collect and control, as argued by Tao et al. (2010). This situation started with the major tax reforms in 1994, when the sum of provincial government fiscal balances suddenly shifted from being balanced in 1993, with almost no net contribution from the center, to a deficit equivalent to almost 43 percent of expenditure in 1994, financed mostly through centrally directed revenues. This proportion of deficit to expenditure in the total provincial government fiscal balance remained more or less stable thereafter. For example, it was 45 percent in 2010.11 It is precisely this shift that allowed the center to start subsidizing the western region so intensively from the mid-1990s onward. In most western provinces since then, the ratio of deficits to local expenditures has been higher than in central and coastal China and was increasing since the late 1990s. The only exception in this regard is the TAR, where the deficit always amounted to almost the totality of local government expenditure, even before 1994 and without any significant change over time. These perpetual deficits in the TAR would have been mostly financed by central subsidies. We do not have access to precise data that would allow us to measure such fiscal subtleties, although it is safe to assume that very little of the deficit in the TAR would have been financed through revenue raised in the TAR by the central government, in particular considering the various incentives and tax waivers that the government offers in order to encourage businesses to move far westward to the Tibetan areas. The contribution of “net taxes on production” to the overall GDP, calculated according to the income approach of national accounting, has always been much lower in the TAR than in all other provinces, at 7.4 percent of total GDP in 2010, versus 14.6 percent in Qinghai, 16.2 percent for Gansu, or 15.6 percent for Sichuan (calculated from CSY 2011, table 2-16).12 Given that local government revenue in the TAR amounted to 7.2 percent of GDP in 2010, the revenue collected by the central government in the TAR therefore probably amounted only to an insignificant addition, perhaps only 0.2 percent of GDP if the comparison of these two measures were appropriate,13 versus a budget deficit equivalent to 101.4 percent of GDP. In contrast, local government revenue in Qinghai amounted to 8.2 percent of GDP in 2010, whereas net taxes on production amounted to 14.6 percent of GDP, implying by this comparison that central government revenue collection in Qinghai amounted to about 44 percent of the total revenue collected in Qinghai,



Instituting Economic Growth and Marginalization in Tibet 151

or 6.4 percent of GDP—only a small share of the local government deficit of 46.9 percent of GDP, but substantially more than in the TAR. Similarly, local government revenue in Sichuan amounted to 9.1 percent of GDP in 2010 whereas net taxes on production amounted to 15.6 percent, implying that about 42 percent of revenue in Sichuan was collected by the central government, or about 6.5 percent of GDP, which would amount to about two-fifths of the local government deficit of 15.7 percent of GDP in 2010. This comparison, while based on extrapolation across different measures of taxation and revenue, nonetheless gives a rough idea of the degree to which central government transfer payments have been raised through local taxation rather than being financed through subsidy. The residual thereby offers a more effective representation of actual subsidization, for example, about 9.2 percent of GDP in Sichuan, 40.5 percent of GDP in Qinghai, and 101.2 percent of GDP in the TAR. Accordingly, we can assume in the case of the TAR that the deficit serves as a rough but robust proxy for direct budgetary subsidies from the central government, in addition to subsidized investment. These fiscal dynamics—especially the exceptional fiscal deluge in the TAR—are clearly demonstrated in figures 4.8 and 4.9. Figure 4.8 measures the local government deficit of each province as a proportion of total expenditure (i.e., the degree to which expenditure is financed by central government revenues, whether or not subsidized). Figure 4.9 measures the magnitude of this deficit (subsidized to various degrees) as a proportion of GDP. The sharp rise in the measure of deficit to expenditure in 1994 in all provinces besides the TAR, as shown in figure 4.8, is a reflection of the tax reform in that year. Through this reform, the central government share of total central and provincial revenues shifted from 22 percent to almost 56 percent in one year, as discussed above. Subsequent increases in the western provinces above the provincial total norm is reflective of subsidization, first with the focus on western development in the Ninth Five-Year Plan, then under the OWC and the Tenth and Eleventh Five-Year Plans. In the case of Sichuan, the sharp increase in 2008 is explained by the Wenchuan earthquake in that year, which led to a large transfer of resources to the province by the central government for several years. Disaggregated government expenditure data for Sichuan clearly indicate that the increased expenditures were related to recovery and reconstruction activities.14 To a lesser extent, there was an increase in this ratio in 2009 in Qinghai, Gansu, and the provincial total, which probably reflects government efforts to bolster domestic growth in the face of the global economic crisis. In contrast, the TAR was exceptional in the degree to which an extreme level of subsidy dependence was maintained throughout the entire period shown. The share in the TAR was already very high in the early 1990s, at almost 100 percent, and then declined slightly throughout the 1990s as

152

Chapter 4

Figure 4.8: Local government deficit as a proportion of local government expenditure, selected provinces and provincial total, 1990-2010 (current value)

Sources: calculated from CSY (2011, tables 2-14, 8-7 and 8-8) and equivalent in previous yearbooks.

economic growth was resuscitated in this autonomous region. The effects of the 1994 tax reform barely register in this measure. The ratio consistently remained well above 90 percent throughout this period—a pattern that started in the late 1960s and has not abated with intensive development efforts since the 1990s. Because such expenditure was increasing rapidly as a proportion of GDP, the magnitude of subsidies in the local economy of the TAR has never been higher, as shown in in figure 4.9. Figure 4.9 clearly shows that it is only the TAR that is treated with an almost absurd degree of subsidization priority. The TAR government deficit and, hence, direct budgetary subsidies from Beijing, surpassed 100 percent in 2010, following a previous peak of 81 percent in 2002 and then a trough of 64 percent in 2006. This is the highest level ever observed for this measure, even in comparison to the peaks of subsidization during the Maoist period. Although government expenditure was heavily subsidized in the Maoist period—sometimes in excess of 100 percent (see Wang and Bai 1991 [1986], 73)—government expenditure (and hence subsidies) was proportionately lower relative to GDP.15 By 2010, government expenditure in the TAR was equivalent to almost 109 percent of GDP. Given that 93.3 percent of this expenditure was deficit/subsidy financed, the deficit/subsidies were equivalent to 101.4 percent of the TAR GDP. Ideally, a measure of subsidization should be combined with investment data (analyzed below), given that most investment in the TAR would have also been subsidized, whether directly or indirectly. Government expenditure data up to CSY (2007, table 8-15) included a category of “expenditure



Instituting Economic Growth and Marginalization in Tibet 153

Figure 4.9: Local government deficit as a proportion of provincial GDP, selected provinces and provincial total, 1990-2010

Sources: calculated from CSY (2011, tables 2-14, 8-7 and 8-8) and equivalent in previous yearbooks.

for capital construction,” which reflected investment financed directly by the TAR government, most of which was effectively subsidized by Beijing given that most of the TAR government expenditure was subsidized. However, this category was discontinued in the expenditure reporting from CSY (2008) onward. In any case, it did not include investments in the TAR that would have been subsidized by the central or other provincial governments independent of TAR government expenditures (such as investment in the railway). For the sake of simplicity, by assuming that all investment in the TAR was subsidized and adding this to the deficit/subsidy measures above, albeit deducting expenditure on capital construction from the measure of government subsidy so as to avoid double counting with the investment data, this approximate measure of direct and indirect subsidies passed 100 percent in 2001 and rose to 123 percent of GDP by 2006. It would have been much greater by 2010 given the subsequent sharp rises in both deficit and investment as proportions of GDP, although the data has not been available since CSY (2007) to continue this particular calculation. The degree of similarity is also striking in the trends for the TAR observed in figure 4.9 compared with the trends in the GDP shares of construction and the tertiary sector, as shown in figures 4.6b and 4.7 above. In particular, the increasing magnitude of deficit/subsidies from 2006 onward is almost identical in trend to the increase in the GDP share of construction, suggesting a continuation of the trends discussed above after 2006, even though this occurred after the completion of the Golmud–Lhasa railway. The previous subsidy surge from 2001 to 2004 preceded this railway construction by

154

Chapter 4

one or two years, although the huge bulge in the proportion of expenditure and subsidies to GDP in the TAR from 2000 to 2004 and the shorter bulge in 2005, as shown in figure 4.9, were related almost entirely to government expenditure on capital construction, in the sense that if such expenditure were deducted from the total, the share of the deficit/subsidy to GDP would have remained more or less constant throughout these years at less than 50 percent. In addition, the surge in these years also appears to bear close correlation with the sharp increase in the share of tertiary sector from 2000 to 2002, as shown in figure 4.7. There was also a rising weight of the deficit (relative to GDP) in the three other western provinces shown in figure 4.9, particularly from 2008 onward. This would imply rising subsidies, as a proportion of both government expenditure and GDP, which in turn would be related to central government efforts to rebalance the economy toward a stronger domestic orientation in the face of the global economic crisis. Among these three other cases, Qinghai represents the second-highest level of subsidy dependence in China and in this respect it started to resemble the TAR with about a tenyear lag. The ratio of deficit to provincial government expenditure in Qinghai reached 85 percent in 2010, and the ratio of deficit/subsidy to GDP reached 47 percent in 2010. However, the structural differences between the economies of the TAR and Qinghai, as analyzed in the previous section, suggest that subsidies have had a much more productive focus in Qinghai. Moreover, a larger proportion of this deficit would have been funded by revenue raised locally by the central government in Qinghai (and even more so in other western provinces, as discussed above), although this local revenue-generating capacity obviously becomes quite limited at the levels of deficit observed in Qinghai by 2010. It remains to be seen whether such levels in Qinghai and other western provinces will prove to be a oneoff surge or else the beginning of a new and sustained pattern of regional redistribution as China strives to rebalance its economy toward its interior. The high and increasing deficits of various western provinces raise some concern as to whether these deficits are sustainable. However, this is less of an issue for the smaller provinces such as the TAR and Qinghai. For instance, the budget deficit of the TAR, at 101 percent of its GDP in 2010, accounted for only 1.5 percent of the total of all provincial government deficits in 2010 (albeit, for only 0.2 percent of the national population). The budget deficit of Qinghai, at 47 percent of its GDP in 2010, similarly accounted for only 1.9 percent of all provincial deficits (for 0.4 percent of the national population). Sustainability is more of a concern for large provinces such as Sichuan, with a deficit at 16 percent of its GDP in 2010 accounting for 8.1 percent of all provincial deficits, for 6 percent of the national population, or Xinjiang, with a deficit of 22 percent of its GDP accounting for 3.6 percent of all provincial deficits, for 1.6 percent of the national population. None-



Instituting Economic Growth and Marginalization in Tibet 155

theless, such deficits are sustainable in principle if they are counterbalanced by central fiscal surpluses, together with a sustained political commitment by the central government to regional redistribution. Rather, it is the intensity of subsidies in the less populated provinces that has consistently raised political attention in China. There are two common concerns. One regards equity, leading to perceptions that Tibetans are being pandered to and spoiled even though most subsidies probably never reach the average Tibetan. The other concern regards efficiency, that is, the increasing degree of subsidization that is required to sustain a consistent rate of growth. Equity of Subsidization In terms of equity, there is no doubt that the TAR (although not necessarily Tibetans) receive much more in subsidies per person than other provinces in western or central China. For instance, per capita subsidies (deficit) in the TAR reached 17,105 yuan in 2010, versus a per capita GDP of 16,904 yuan, as shown in figure 4.10.16 In comparison, the per capita local government deficit in Qinghai was 11,237 yuan in 2010, 4,355 yuan in Gansu, 3,352 yuan in Sichuan, and 2,481 for all provincial government. The ratio of the TAR with the provincial total—at 6.9 times in 2010—had not changed much since the late 1990s, except in 2002 when it jumped to about nine times and fell gradually thereafter. Keeping in mind that the deficit in the all-province total would include central revenues raised locally, the ratio nonetheless reflects that, despite the phenomenal rise in subsidies in the TAR after 2006, they had not risen relative to the all-province total per capita deficit, which was increasingly subsidized over this period as well. Hence, although the TAR remained disproportionately privileged (although not necessarily Tibetans in the TAR), the extent of this privileging remained more or less constant over time given that the subsidization of other provinces also intensified. With further reference to the data shown in figure 4.10, which are in constant value, it is interesting to note that the real (constant yuan) value of subsidies per person in the TAR did not actually increase in the second half of the 1990s, but only kept up with price inflation, as throughout China. Obviously, in nominal (current yuan) value, subsidies per person were increasing, from 575 yuan per person in 1990 to 1,364 yuan per person in 1995 and then to 2,116 yuan per person in 2000. After 2000, the real value of deficit/subsidies per person in the TAR increased rapidly, as they did to a lesser extent in other western provinces. The real per person increase was especially phenomenal after 2006 in both the TAR and Qinghai. The extremely close correspondence between rising per capita deficit/subsidies and per capita GDP in the TAR suggests that the increases of the latter were mostly a reflection of the former.

156

Chapter 4

Figure 4.10: Local government deficit per capita, constant 2010 rmb, selected provinces and provincial total, 1990-2010

Sources: calculated from CSY (2011, tables 2-14, 8-7 and 8-8) and equivalent in previous yearbooks.

However, the extent to which subsidization in the TAR was disconnected from most Tibetans is important to bear in mind with regard to arguments that Tibetans (rather than the TAR) have been disproportionately privileged. For instance, the per capita deficit/subsidy in the TAR in 1990 was 0.9 times the average per capita rural household income in the TAR in the same year, whereas it was 1.6 times in 2000. This reveals the extent to which the increasing value of subsidies had very little relation to actual wealth conditions in the rural areas of the TAR, where 85 percent of TAR Tibetans resided in 2000. After 2000, the divergence became even more accentuated. The ratio of per person subsidies to average per capita rural household income in the TAR rose to 4.1 by 2010, even despite strong growth in rural incomes after 2003. Per person subsidies in 2010 were even equivalent to over 1.1 times the per capita urban household disposable income. In other words, if the TAR government had simply divided these subsidies equally among all permanent residents of the TAR—urban and rural, Tibetans, Han, and others— and if these residents did nothing other than receive these subsidies, the net income of the average rural Tibetan in the TAR would have been more than four times higher than it was in 2010, and that of the average urban resident 10 percent higher than it was in 2010. Considering that subsidies in effect would have accounted for only a fraction of total rural incomes in the TAR (whether as direct transfer payments to households or else as subsidies for various rural development and construction projects that employed rural Tibetans), in addition to the traditional activities of farming and herding as well as new activities such as caterpillar fungus harvesting, it is clear that the degree of trickle down of subsidies to the rural areas and Tibetan households has been very limited relative to the sheer deluge of subsidies.



Instituting Economic Growth and Marginalization in Tibet 157

The actual uses of government expenditure help to further clarify the local disconnect in the TAR. Education has consistently been the largest category of expenditure across China, at 16 percent of the total general budgetary expenditure of all provincial governments in 2010. This was also broadly the case in most western provinces, even under the surging investment and construction that accompanied the OWC. Although the share of government expenditure in capital construction (i.e., government expenditure used for investment) increased sharply in the early 2000s in all western provinces with the initiation of the OWC, it still rarely exceeded expenditure on education until 2006, after which capital construction was no longer reported as a separate category of government expenditure.17 Other categories of expenditure were relatively stable as shares of total expenditure throughout this period, indicating that expenditure priorities more or less remained proportionate with increasing overall expenditure. Both Qinghai and the TAR were exceptional in the degree to which capital construction (i.e., investment) became the largest category of government expenditure in the late 1990s and then surged in the early 2000s, reaching a peak of almost 30 percent of total government expenditure in Qinghai in 2002 and 41 percent in the TAR in 2003. The share started to fall in both provinces after these peaks, although by 2006 it still remained at 29 percent of total expenditure in the TAR versus 10 percent on education, and 17 percent in Qinghai versus 11 percent on education. All provincial government expenditure on capital construction amounted to 9.6 percent in the same year, versus 14.7 percent on education. As discussed above, it is safe to presume that capital construction continued to dominate government expenditure up to 2010 in the TAR and possibly in Qinghai as well. The TAR was exceptional even compared to Qinghai in terms of the second-largest category of government expenditure. In the TAR, expenditure on government administration became the second-largest category of expenditure from 1996 onward, reaching heights of almost 22 percent of total expenditure in 1996, almost 18 percent in 2000, and 18 percent in 2004, even despite surges in capital construction expenditure in those years. In 2006, government administration accounted for 16 percent of total government expenditure in the TAR, versus 10 percent on education. Qinghai exhibited a similar prioritization up to the mid-1990s, but it fell in line with the western provincial norm in the 2000s, with its share of government expenditure on government administration at 9.3 percent in 2006, versus 11.4 percent on education. Most other provincial governments spent substantially less on their administration than on education. In the provincial government total, government administration accounted for 9.5 percent of expenditure in 2006, versus 14.7 percent for education. Again, this category was no longer reported from 2007 onward.

158

Chapter 4

The lower share of education expenditure in the TAR nonetheless amounted to a much higher nominal expenditure per person than in other provinces (as in all the other categories of expenditure, government administration included) given the much higher level of overall per person expenditure in the TAR relative to other provinces, as analyzed above. For instance, government expenditure in education amounted to 2,027 yuan per person in the TAR in 2010, versus 1,465 yuan in Qinghai, 892 yuan in Gansu, 672 yuan in Sichuan, and 883 yuan for the regional provincial total in China (see table 4.1 below). The difference in per person expenditure on education in the TAR—about double the regional provincial average in 2000 and about two and a half times in 2010—was not nearly as high as the per capita deficit/subsidy ratio of seven to one, supporting the claim that subsidies were disproportionately concentrated in capital construction and government administration.. However, the difference was substantial enough to support claims that the government was prioritizing education in the TAR (as well as in other western provinces, where the increase in education expenditure was faster than in the TAR over this decade, albeit from a lower level, as shown in table 4.1). Nonetheless, per capita non-monetary education inputs in the TAR— particularly the per capita number of teachers—were generally on par with other provinces in 2010 (although the TAR had a strong lag at the level of vocational education, similar to the 1990s). These inputs are shown in table 4.1, together with public expenditure in education and the wages of urban employees working in the education sector in 2000 and 2010 across the four provinces and national average (actual education outcomes will be discussed in chapter 6). The year 2000 is chosen because this was just before the enormous surge of subsidies that started in the TAR in 2001 under the OWC (albeit still after several years of increasing subsidization and rapid growth in the late 1990s). Two input measures are used: schools per million people and teachers per million people. The measure of teachers per person, as opposed to schools per person, is provided in order to account for shifts that have been taking place across China toward larger consolidated schools, which would reduce the number of schools per person—and even some staff through scale efficiency gains—without necessarily reducing the effective supply of education (quality of schooling is another very debatable question). This policy of school consolidation was started in some parts of Qinghai in 2009 (see Cencetti 2012). It was probably not yet strongly reflected in the data of the TAR by 2010, whereas the implementation of this policy appears to have been much more advanced in Sichuan, where the number of primary schools per million people, for instance, apparently fell from 520 to 115 between 2000 and 2010—a sharper reduction than nationally—whereas per million primary school teachers fell only marginally, from 3,981 to 3,802.



Instituting Economic Growth and Marginalization in Tibet 159

Table 4.1: Expenditure, costs, and infrastructure inputs in education, selected provinces, 2000 and 2010 TAR

Qinghai

Gansu

Sichuan

China

pc gov exp in education 2000

266

140

108

78

128

pc gov exp in education 2010

2,027

1,465

892

672

883

% change pc gov exp 2000-10

661%

943%

730%

764%

588%

average s&w wage in edu 2000

15,727

10,945

8,548

7,923

9,482

average urban wage in edu 2010

52,781

42,447

29,725

34,408

38,968

% change edu wage 2000-10

236%

288%

248%

334%

311%

Primary schools / million 2000

321

662

841

520

437

Primary schools / million 2010

291

318

453

115

192

PS teachers / million 2000

5,031

5,349

4,886

3,981

4,630

PS teachers / million 2010

6,300

4,722

5,488

3,802

4,193

Junior sec schools / million 2000

32

55

50

42

50

Junior sec schools / million 2010

31

56

62

50

41

2,958

2,544

3,251

2,545

2,630

Sen SS / million 2000

5.7

31.9

16.4

9.4

11.5

Sen SS / million 2010

9.7

21.1

17.7

9.3

10.5

SSS teachers / million 2010

1,057

1,341

1,467

999

1,133

SS teachers / million 2000

1,644

3,213

2,892

2,606

3,164

SS teachers / million 2010

4,015

3,885

4,718

3,543

3,763

Voc. SS teachers / million 2010

197

434

581

450

508

Higher ed instit / million 2000

1.5

1.4

0.7

0.5

0.8

Higher ed instit / million 2010

2.0

1.6

1.6

1.1

1.8

HEI teachers / million 2000

310

407

281

221

366

HEI teachers / million 2010

732

663

812

808

1,003

15+ illiteracy rate 2000

47.3%

25.4%

19.7%

9.9%

9.1%

15+ illiteracy rate 2010

32.2%

13.1%

10.6%

6.6%

4.9%

Share of pop 0-14 years old, 2000

31.2%

26.9%

26.9%

22.4%

22.9%

Share of pop 0-14 years old, 2010

24.4%

20.9%

18.2%

17.0%

16.6%

JSS teachers / million 2010

Sources: calculated from CSY (2011, tables 3-9/11, 4-14/16, 8-8, 20-27/28/29/30/31/32/33/34) and equivalent in CSY (2001).

160

Chapter 4

Overall, these data definitely reflect a catching-up of the TAR with the rest of China in terms of the supply of hard and soft education infrastructure, particularly at the secondary level, where the TAR was severely lagging in 2000 (for an analysis of 2001 data, see Fischer 2005a, 65-69). The primary level in the TAR was already on par in 2000 with the per capita national supply of teachers (although not schools), and the supply of per capita primary teachers increased even further by 2010 while it decreased in China, Sichuan, and Qinghai, leaving the TAR with a significantly greater number of teachers per person. However, this is correspondent to what we would expect for the much younger population in the TAR and, hence, the proportionately much greater need for primary schooling (24 percent of the TAR population was aged between 0 and 14 years in 2010, versus 17 percent in China and Sichuan, 18 percent in Gansu, and 21 percent in Qinghai). This is the extent of the privileging that can be discerned from these data, so to speak: to bring the TAR up to a national norm of per capita teacher supply, but not beyond it in any obvious way despite a much greater need due to a much younger population, much higher levels of illiteracy, and much lower levels of schooling. Catching up was especially evident at the secondary level, although again not beyond the national or western provincial norm and with a continued lag at the senior secondary and especially vocational levels. Secondary schools also appeared much more centralized in the TAR than in other provinces, perhaps due to the sparsely populated characteristics of the TAR and the reliance on boarding schools at the secondary level (and, at the senior secondary level, the lower rates of enrollment). For instance, the number of junior secondary schools per million people was 31 in the TAR (down from 32 in 2000), versus 56 in Qinghai (up from 55), 50 in Sichuan (up from 42), and 41 in China (down from 50). In contrast, the TAR actually had a greater number of junior secondary school teachers per million people, at 2,958 in 2010, versus 2,544 in Qinghai and 2,630 nationally, signifying a greater concentration of teachers (and students) per school (and a greater need given the younger population and the enforcement of nine years compulsary schooling). At the senior secondary school level, the TAR had fewer per capita schools and teachers in 2010 (9.7 schools and 1,057 teachers per million people) than all the other cases besides Sichuan. In contrast, the higher per capita number of senior secondary schools in Qinghai (21 per million people in 2010) probably reflects the duplication of schools due to the extensive minority education system,18 although this number had fallen sharply over the decade (from 32 schools per million people in 2000) due to the school consolidation policy and was not matched by a significantly greater per capita number of teachers. Junior and senior together, the severe lag in the per capita number of teachers in the TAR had been corrected by 2010 and was broadly in line with the national norms (4,015 secondary



Instituting Economic Growth and Marginalization in Tibet 161

school teachers per million people in the TAR, versus 3,763 nationally and 3,885 in Qinghai).19 The strong lag in vocational education in 2010 is reflected by the fact that the number of vocational secondary school teachers per million people in the TAR was only 197 in 2010, versus 434 in Qinghai, 581 in Gansu, and 508 nationally. As I have noted previously (Fischer 2005a, 65-69), this lag at the vocational level is worrisome considering the urgent need for skills training in the TAR, particularly in the face of strong competition with migrants from outside the province, yet this is precisely where the education system appears to be most under-prioritized. On the other hand, higher education continued to be oversupplied in the TAR relative to other provinces in 2010. This is possibly due to a scale issue, whereby a province is expected to maintain a minimum number of higher education institutes regardless of its population size. Nonetheless, the number of per capita higher education institutions in the TAR increased from 1.5 per million people in 2000 to two in 2010—the highest of the five cases listed in table 4.1 in both years. In absolute terms, this represents an increase from four to six higher education institutions, versus seven to nine in Qinghai, or eighteen to forty in Gansu. The number of higher education full-time teachers per million people also increased substantially, more than in Qinghai but not by as much as in Gansu, Sichuan, or China. From this perspective, the apparent privileging of the TAR in per capita government expenditure, whereby the TAR government spent almost two and a half times more per person on education in 2010 than the all-China provincial average, is not reflected by an equivalent privileging in the supply of education infrastructure inputs. Instead, the much higher expenditure reflects the higher costs of supplying education in the TAR relative to other locations, due largely to the policies that guide such servicing. The most substantial cost in this respect is wages (given that education is labor intensive). The average wages in the education sector of people employed in urban units in 2010 (or of urban staff and workers in 2000) is shown in the second set of rows in table 4.1. These do not account for wages in rural areas or in the urban private sector, although they are the only wage data available. Accordingly, such wages in the TAR were 1.7 times the national average in 2000 and 1.35 times in 2010. The wage difference probably accounted for a large share of the difference in education expenditure (even though wages did not increase as rapidly as the education expenditure in all of the cases in table 4.1).20 In addition to wages, another reason for the higher costs of per person education provisioning in the TAR and, to a lesser extent, in other remote western provinces is the fact that it is relatively more expensive to service a low population density region than a high population density region. This is because the catchment area of one school covers a much smaller population and thus the costs of providing school infrastructure per student are

162

Chapter 4

higher, particularly where subsidized boarding is supplied or required. For instance, most schools in the minority areas of Qinghai are four to five times more expensive than ordinary schools in China (Cheng 2003, 201). In addition to this, the education infrastructure in the TAR is top heavy in the sense that there are more higher education institutions per person and, within these institutions, lower student-to-staff ratios. Because higher education institutions are more expensive than primary and secondary schools, and a lower student-to-staff ratio accentuates this costliness, the top-heavy education structure in the TAR requires proportionately more funding per person than in the rest of China. After accounting for all of these considerations, it is not at all clear that education expenditures in the TAR are privileged beyond the structural and institutionalized costs of providing education, or else beyond the imperative of bringing education infrastructure up to a national norm, particularly after accounting for the younger population of the TAR and hence the greater per capita need. It is open to debate whether the structural costs are inherent to social provisioning in remote regions or else whether they are more specifically related to central government policies designed to accelerate the integration of the TAR into the rest of China, particularly through wage policies designed to attract human resources from around the country in an increasingly open employment and migration context. However, it is clear that a large portion of subsidies have simply bypassed these categories of social spending altogether and have focused on government administration and capital construction. Efficiency of Subsidization The efficiency concern (i.e., that subsidies have not been particularly efficient in producing GDP growth) has been particularly acute with respect to the TAR, as earlier highlighted by Wang and Bai (1991 [1986]) and Yan (2000), and more recently by Jin (2010). Dreyer (2003, 421) also discusses the public debates in China in the 1990s regarding the subsidy dependence and declining economic rates of return in the TAR. The efficiency concern is usually measured in terms of the elasticity of growth to subsidies (or the multiplying effect of subsidies on growth), calculated as the increase in GDP value associated with a per unit increase in the value of deficit/subsidies (this has been the conventional measure used in the Chinese debates cited above).21 A ratio of one (i.e., one yuan of subsidy increase matched by one yuan of GDP increase in the same year) represents no multiplier effect of subsidies on immediate growth, that is, the economy merely expands by the increase in the value of subsidies injected into it. In the context of balanced trade, this would be a minimum expectation, given that actually used investment in one year would show up as construction and other ac-



Instituting Economic Growth and Marginalization in Tibet 163

tivity in that year, and direct budgetary subsidies would similarly stimulate economic activity within the same year through salaries and other expansions in economic activity, including construction itself. A ratio below one represents a negative multiplier effect, that is, an increase in subsidies by one yuan is matched by less than a one-yuan increase in GDP. This occurs if the increase in subsidies creates inefficiencies elsewhere in the economy (or outflows from the economy altogether) such that part of the effect of the increase is canceled out. Ideally, subsidies should have a positive multiplier effect, such that an injection of one yuan produces more than one yuan of GDP growth as the yuan circulates through the economy. This measure must be treated with circumspection given that growth is due to many factors other than subsidies, such as investment, especially in the larger provinces where subsidies might play only a minor role in growth. Multiplier measures would tend to overemphasize the effect of subsidies in provinces such as Sichuan, where a very small value of subsidy might be associated with a large value of GDP growth, implying a very high multiplier effect even though most of the growth might not have been related to the subsidies. Indeed, this is also a problem in attempts to judge the growth effect of international official development assistance through similar measures, which are common in cross-country regression analyses of this nature. Nonetheless, this measure is probably useful to reflect the effectiveness of subsidies in the TAR given the predominance of subsidies in this province, although ideally the measure of subsidies should include subsidized investments, whereas data has not been available for this purpose since 2006, as noted above. The results of this measure are shown in table 4.2, calculated on a five-year basis, roughly matching the Five-Year Plans.22 The results show a declining multiplier effect of deficit/subsidies on growth across the three five-year periods in all of the cases shown. This is predictable given that it corresponds to the increasing weight of subsidies as a proportion of GDP. In Sichuan and the regional total, the results are largely reflective of the growing weight of subsidies from a relatively minor magnitude and not necessarily the declining effectiveness of subsidies, as explained above. Table 4.2: Ratio of the change in nominal GDP over change in deficit/subsidies, various provinces and the provincial total, 1995-2010 Five-Year Period

TAR

Qinghai

Gansu

Sichuan

Regional Total

1996-2000

2.81

3.12

5.41

11.79

17.42

2001-2005

1.13

3.32

5.31

8.78

13.50

2006-2010

0.75

1.62

2.70

4.68

9.48

Sources: calculated from CSY (2011, tables 2-15 and 8-8) and equivalent in earlier CSYs.

164

Chapter 4

Considering this qualification, it is interesting that in Sichuan during the last five-year period, when the deficit increased as a proportion of GDP from 8.6 percent in 2006 to 15.7 percent in 2010, a one-yuan increase in deficit was associated with an increase of 4.7 yuan in the GDP of this province—a substantial multiplier given the weight of the deficit in the local economy. The multiplier also declined but remained quite strong in Gansu, where the weight of subsidies was even heavier, and it was lower but still positive in Qinghai, where the deficit/subsidies reached 47 percent of GDP in 2010. In the TAR, where subsidies dominate the economy and thus growth, the multiplier declined from 2.8 during the Ninth Five-Year Plan (1996-2000) to 1.1 during the Tenth Five-Year Plan (2001-2005), when subsidies surged as a proportion of GDP. This measure then fell below one (i.e., a negative multiplier effect) during the Eleventh Five-Year Plan (2006-2010), meaning that for every one-yuan increase in subsidies, the GDP actually increased only three-quarters of a yuan. The government was effectively increasing the value of subsidies faster than the growth of value-added in the economy. Moreover, this measure does not include subsidized investments, which if included, would lower the multiplier effect even further and would also make it appear negative earlier. For instance, deducting expenditure on capital construction from the deficit in order to avoid double counting (as explained above) and then adding the remaining deficit to investment can give a rough overall assessment of total subsidies in the TAR economy (assuming that most investment in the TAR is subsidized). According to this measure, the multiplier of subsidies and investments to growth was negative (below one) from 1999 onward (calculated up to 2004 in Fischer 2007, 156), to the extent that in 2001 there was only 0.5 yuan of GDP increase for every one yuan of increase in subsidies and investment combined. In comparison, the equivalent multiplier in Qinghai, the next most subsidized province in China, fell below one for a few years in the late 1990s and early 2000s, reflecting the intense subsidization of western development strategies over these years. However, the multiplier returned positive (above one) after a few years, reflecting that such subsidization was directed toward a productive restructuring of the industrial base of the province, including strong expansions into mining and hydroelectricity, as discussed previously. Data is not available to evaluate whether this combined multiplier again fell below one in Qinghai during the surge of subsidies and investment under the Twelfth Five-Year Plan, although this outcome was possible, particularly after 2008. The negative multiplier effect of subsidies in the TAR (where we can reasonably talk of a multiplier effect of subsidies) effectively represents a return to the state of affairs during the Maoist period. Wang and Bai (1991 [1986],



Instituting Economic Growth and Marginalization in Tibet 165

68-73) point out that for every one-yuan increase in the value of output between 1957 and 1983 in the TAR, subsidies increased by 1.24 yuan, that is, a negative multiplier effect. They also recorded a similar—although less dramatic—trend in Qinghai. According to their data for the TAR, local revenue (or what they call “local financial income”) turned negative and the level of subsidy started to increase sharply in 1968. The proportion of central government subsidies to output value (equivalent to the subsidy/ GDP measure above, although not including the tertiary sector)23 increased from 31 percent in the 1950s, to 45 percent in the 1960s, 80 percent in the 1970s, and 97 percent in 1980-1983. Or, the negative multiplier effect that they measured from 1957 to 1983 was mostly concentrated from 1968 onward, and it would be much more accentuated if the calculation were correspondingly adjusted. Notably, they did not clarify that this intensification of subsidies occurred simultaneously with the late implementation of rural collectivization in the TAR, which started to be fully implemented only in 1969, as discussed in chapter 2. The timing was also in the midst of the Cultural Revolution and intensive interior industrialization strategies. The implication of this lacuna in their argument is discussed further in the next section. Similar patterns continued even in the context of reform. For instance, the nominal value of government expenditure almost doubled in 1984 (to 75 percent of GDP, and with more than 100 percent of expenditure subsidized),24 yet the GDP only briefly came out of recession and fell back into recession in 1986, as analyzed in chapter 2. Nominal government expenditure declined sharply in 1986 and then rose gradually thereafter, albeit slower than inflation. Overall, real government expenditure (indexed to the provincial CPI) passed its 1984 nominal value only in 1998. This helps to explain the recessionary trend in real per capita GDP from the mid-1980s to the mid-1990s, after which the combination of investment and subsidy increases served to buoy up real per capita GDP, although in an extremely inefficient manner. In sum, recent development strategies in the TAR have not altered the trend of very intense and very inefficient subsidization observed since the late 1960s in any significant way. This point will be explored further below, after a discussion of investment, the other major (also largely subsidized) source of growth in the TAR. Investment in Fixed Assets The other major source of subsidies in the TAR is investment in fixed assets (“investment” for short). Investment is also subsidized to a lesser extent in other western provinces. In principle, the extent of subsidization would decline with the size of the province, similar to the case of subsidies as a proportion of the fiscal deficit. In the TAR, the bulk of investment is

166

Chapter 4

financed by the central government or, to a lesser extent, by various coastal provincial governments (e.g., the railway was entirely subsidized by Beijing whereas provincial aid projects are entirely subsidized by various rich coastal provinces). Data is not available to be able to track whether investment funding originates from inside or outside a province. However, as an indirect indication, 62 percent of investment in the TAR was funded by “state budget” in 2010 (which could be the TAR government budget, the central government budget, or the budget of another province). Given that the entire TAR government revenue amounted to 7.2 percent of GDP in 2010 whereas investment in the TAR amounted to 91.2 percent of GDP in the same year, it is clear that most of these state budget funds for investment were from outside the TAR. In contrast, state budget funds provided only 5 percent of investment in China, 9 percent in Sichuan, and 14 percent in Qinghai. Only 2 percent of investment in the TAR was funded through domestic loans (an important source of local investment financing in many provinces), versus 15 percent in China, 16 percent in Sichuan, and 17 percent in Qinghai. The largest source of investment funds in China was “self-raised funds” (the largest component of which was probably reinvested earnings), which accounted for 63 percent of investment funds in China, versus 57 percent in Sichuan, 52 percent in Qinghai, and only 31 percent in the TAR (calculated from CSY 2011, table 5-5; see table 4.3 below). Similar to the state budget category, much of the self-raised funding in the TAR probably came from outside the province, such as a state-owned enterprise in Sichuan raising its own funds for a project in the TAR, whereas in other provinces self-raised funds would include a substantial local element.25 As with deficit/subsidies, the level of total investment to GDP in the TAR reached levels unparalleled anywhere in China in recent history, at almost 80 percent of GDP in 2006 and then rising even further after 2008, to 91 percent of GDP in 2010 (see figure 4.11). If the data on gross capital formation (GCF) from the expenditure approach of GDP accounting are used—which is reasonable given that these are the data used for GDP calculations rather than the investment data—gross capital formation in the TAR reached over 111 percent of the TAR GDP in 2010 (see figure 4.12). The fact that gross capital formation was so much higher than investment in fixed assets (IFA)—by the equivalent of about 20 percent of GDP in 2010—is difficult to explain with the available data, although the extremity of this discrepancy is nonetheless exceptional to 2010, and hence might be explained by some imaginative accounting in that year.26 In any case, both measures reveal the degree to which the government has intensified investment in the TAR even beyond the levels reached in the middle of the decade. Such levels of investment once appeared exceptional to the TAR and Qinghai, particularly in the early to mid-2000s, although they actually became



Instituting Economic Growth and Marginalization in Tibet 167

Figure 4.11: Investment as ratio of GDP, selected provinces, 1990-2010

Sources: calculated from CSY (2011, tables 2-1, 2-13 and 5-3) and equivalent in earlier CSYs.

Figure 4.12: Gross capital formation as ratio of GDP, selected provinces

Sources: calculated from CSY (2011, tables 2-1, 2-13 and 2-21) and equivalent in earlier CSYs.

the norm in western China after 2008. Even nationally, investment apparently surpassed 69 percent of GDP in 2010 and gross capital formation reached 49 percent of GDP. The difference with the TAR is that, as noted above, such levels of investment in the other provinces and nationally would have been locally financed much more than in the TAR. The earlier trends in the investment data are discussed elsewhere in my work (Fischer 2005a; 2007; 2009b), although some perspective is worth noting here. Investment in both the TAR and Qinghai started from more or less the same level as the national average in 1990 (in the TAR) and 1995 (in Qinghai). There was a one-year investment spurt in 1995 in the TAR,

168

Chapter 4

when investment reached 63 percent of GDP, but then the TAR resumed a level of investment to GDP that was more or less identical to Qinghai, both substantially higher than the national norm up to 2002. After 2002, levels of investment to GDP continued to climb in the TAR. They fell in Qinghai, converging back toward the (rising) national pattern up until 2008. After 2008, investment surged in all cases. The large spike in Sichuan in 2009 is probably related to earthquake reconstruction. Nationally, this measure of investment—at over 69 percent in 2010—shows the extent to which investment in fixed assets drove China’s response to the global financial crisis despite publicly-stated intentions to decrease dependence on investment. Nonetheless, the sharp increases in all of the western provinces—as with the levels of deficits shown earlier—are reflections of the effort of the central government to reorient aggregate demand in China toward domestic sources, in which western development strategies have played a key role. Similar trends are observed in the gross capital formation GDP data in figure 4.12, except that GCF was much lower than investment in the TAR from 1997 to 2003. This probably reflects a high degree of inefficiency underlying investment in these years, as I have previously analyzed in detail (Fischer 2005a, 73-82). The similarity between the TAR and Qinghai in terms of investment up to 2002 is not apparent in terms of gross capital formation; the TAR actually appears quite similar to the national trend in terms of GCF despite much higher levels of investment. Besides inefficiencies, this might also reflect that the heightened level of investment in fixed assets that was apparently occurring in the TAR over these years was in fact representative of higher wage levels or other costs being paid in such investment projects (similar to the case of subsidized expenditures versus actual provisioning of education, as discussed above), rather than an actual faster rate of capital formation (i.e., roads and railways constructed). Accordingly, similar investment levels in Qinghai at lower wages and costs led to more actual capital formation. This situation in the TAR was corrected from 2004 onward, when the railway construction reached full steam in the TAR and gross capital formation actually surpassed investment. This correction and the pattern of GCF’s exceeding investment in subsequent years besides 2008 might have been related to the massive plunge into international trade deficit in the TAR from 2003 onward, as discussed in the next section, although the exact relation is difficult to decipher. As with government expenditure, several unique characteristics sharply differentiate investment in the TAR from all other provinces in China, in this case including Qinghai. These are presented in table 4.3, which compares the composition of investment across the five cases in 2010. The comparison with 2000 is not presented, as it is in table 4.1 above, because the reporting categories were not all equivalent in the earlier data and many were not available in CSY (2001). Nonetheless, some comparisons



Instituting Economic Growth and Marginalization in Tibet 169

are made to data from 2004 (as reported in Fischer 2007) and to the 2000 data where available. First of all, in terms of the rural-urban distribution of investment, 87.5 percent of all reported investment in fixed assets in the TAR in 2010 was located in urban areas. This was higher than the national average despite Table 4.3: Composition of investment in fixed assets, selected provinces, 2010 TAR

Qinghai

Gansu

Sichuan

China

Rural/Urban distribution of investment, % of total Urban / total

87.5

82.6

88.9

84.3

86.8

Real estate / urban

2.2

12.9

9.5

19.8

20.0

Ownership structure of investment (by selected statuses of registration), % of total State-owned units

72.0

45.8

53.3

38.1

30.0

Limited liability

3.0

24.5

19.8

29.4

25.3

Private

4.0

9.5

10.2

13.3

21.8

Self-employed individuals

4.5

4.9

3.5

4.7

3.4

Other

12.3

1.6

3.6

2.2

2.8

Funds from HK / M / Taiwan

0.4

0.1

1.1

1.7

3.0

Foreign funded units

0.2

2.2

0.6

2.4

3.2

5.1

7.5

4.4

3.6

2.8

By selected main sectors, % of total Agriculture, ani. husb., for/fish Mining

4.4

7.3

4.3

3.3

4.0

Manufacturing

7.7

27.4

16.5

25.7

31.9

Prod./supply elect., gas, water

12.1

9.1

17.6

7.6

5.6

Transport, storage, and post

25.0

14.4

6.6

12.0

10.8

Real estate

9.9

17.3

16.0

24.1

23.3

Public management & soc. org.

12.5

3.2

9.7

1.8

2.0

State budget

62.1

13.6

17.1

9.4

4.7

Domestic loans

1.9

17.2

15.7

16.1

15.2

Foreign investment

0.3

0.5

0.6

0.7

1.6

Self-raised funds

30.6

51.9

55.2

56.7

63.4

Others

5.1

16.8

11.4

17.1

15.1

By source of funds, % of total

Sources: calculated from CSY (2011, tables 5-2, 5-3, 5-5 and 5-7)

170

Chapter 4

the fact that the TAR was one of the least urbanized provinces in China in 2000 (the urbanization data from the 2010 census was not yet available at the time of writing). Rural investment might be underestimated partly because these data probably refer to the place of registration of the unit making the investment, not necessarily the location where the investment is made. For instance, large-scale infrastructure projects such as the railway are most likely recorded as urban even if much of the construction takes place in rural areas. Nonetheless, these data effectively refer to the locations of control over investment, which are predominantly urban, in the TAR as elsewhere in China. The degree to which investment was registered as rural increased in Qinghai and Gansu since the 2000 and 2004 data, although it decreased in Sichuan and China. It presumably increased in the TAR given that investment was reported as 100 percent urban in the earlier data, although this was probably due to a lacuna in the reporting of rural investment in these data. There is no doubt a significant amount of investment in fixed assets by rural households in the TAR, as analyzed in chapter 7, although much of this is probably not recorded in these investment data given a minimum cutoff point of 500,000 yuan for investment in capital construction or innovation, or else 50,000 yuan for real estate development, rural collective investment, or individual investment (CSY 2003, 184). The comparable statistical source in TSY (2005) showed some rural investment in 2004, possibly due to a lower cutoff point, although only to the order of 0.16 percent of total investment from rural collective-owned units and 2 percent from rural individuals. The main point is that such rural-registered investments accounted for very little of the total volume of investment taking place in the TAR in 2000 and 2004. The increase in share by 2010 might reflect better reporting or else a threshold effect, that is, the fact that rural economic norms in the TAR had increased to such an extent that more rural investment activity was able to surpass the minimum cutoff threshold of 50,000 yuan. As a second characteristic of investment in the TAR, a much greater share of investment was made by state-owned units (SOUs) in comparison to other provinces. In 2010, 72 percent of investment in the TAR was made by SOUs, versus 46 percent in Qinghai, 53 percent in Gansu, 38 percent in Sichuan, and 30 percent in China. This SOU share in the TAR fell since 2001, when it accounted for 95 percent of total investment (a share that SOUs had maintained throughout the 1990s in the TAR; see Fischer 2005a, 72), and also since 2004, when SOUs accounted for 84 percent of total investment.27 However, the fall in the state-owned share in the TAR after 2004 was not compensated by a rise in private or limited-liability forms of ownership, which were the next most prominent forms of investment after state-owned units in most other provinces and China as whole. Instead, investment in the TAR was characterized by a rise in the vague category



Instituting Economic Growth and Marginalization in Tibet 171

of “other,” which had risen from 3.4 to 12.3 percent of total investment between 2004 and 2010. It is not clear what “other” refers to because most other forms of ownership units are accounted for in the data, including state owned, collective owned, cooperative, joint, limited liability, shareholding, private, self-employed individual, funds from Hong Kong, Macau, and Taiwan, and foreign funded. Moreover, “other” ownership units in other provinces were marginal in 2010, at 2.8 percent of total investment in China. If we assume that this “other” category was somehow related to state activities, such as various ambiguously defined para-statal units, it is notable that the share of investment accounted for by state-owned and “other” units combined changed only marginally in the TAR from 87.8 percent of total investment in 2004 to 84.3 percent 2010. In contrast, limited-liability forms of ownership amounted to only 3 percent of total investment in the TAR in 2010, whereas they had become the second most prominent source of investment in most other provinces, at almost 25 percent of total investment in Qinghai, for instance. Third, the three largest sectors of investment in the TAR in 2010—as in 2004—continued to be “transport, storage, and post” (at 25 percent of total investment in 2010 versus 36 percent in 2004), “public management and social organizations” (13 percent in 2010 versus 14 percent in 2004), and “production and supply of electricity, gas, and water” (12 percent in 2010 versus 10 percent in 2004). The shares in these three categories were substantially greater than in the rest of the cases shown, except in Gansu, where the share of investment in public management was also relatively high and the share in the production and supply of electricity, gas, and water was higher than in the TAR. In the case of transport, railway construction would have accounted for a large part of this category in 2004 and 2010. In contrast, the largest sector of investment in all other cases was manufacturing (at 27 percent of total investment in Qinghai in 2010 and 32 percent in China, and at similar shares in 2004). The next largest sector of investment in most other cases was real estate (at 23 percent of total investment in China in 2010 versus 10 percent in the TAR). Investment in “agriculture, forestry, animal husbandry, and fishery” in the TAR increased from 2.8 percent of total investment in 2004 to 5.1 percent in 2010, and in mining from 0.6 percent in 2004, when mining investment was very minor, to 4.4 percent in 2010. Despite some reports suggesting that mining is a major motive for western development strategies in the TAR, the increase of investment in mining merely brought the investment share of this sector more or less in line with the national norm, unlike in Qinghai, where mining interests are more evident, or in Xinjiang (not shown). Where mining accounted for 17.8 percent of total investment. Education and health are not shown in table 4.3 because they are fairly marginal categories of investment, unlike their fiscal importance as major

172

Chapter 4

categories of government expenditure. This is because the bulk of costs involved in expanding education and health are wages, not fixed assets. Nonetheless, we would expect public management to be similarly marginal in the investment data for precisely the same reasons, as it was in most cases besides Gansu. For instance, public management and social organizations accounted for 3.2 percent of total investment in Qinghai and 2 percent in China. However, this category in the TAR continued to display similar patterns as in 2004, when it accounted for 14 percent of investment, or twice the investment in education and fourteen times the investment in “health, social security, and social welfare,” versus Sichuan, where investment in public management (called “government administration” in 2004) was only half of investment in education and double the investment in health (see Fischer 2007, 145). Despite the fall in the share of investment in public management in the TAR to 12.5 percent by 2010, investment in this sector was almost five times the investment in education in 2010 and almost sixteen times the investment in health, social security, and social welfare, still in stark reversal of the investment priorities in other provinces such as Sichuan, where investment in education was still greater than in public management. From the perspective of these characteristics and the sheer weight of investment in the TAR economy, in contrast to the very limited capacity to finance such investment locally, it is clear that most of the funding, ownership, and priorities guiding such investment came from outside the TAR.

Explaining Economic Absurdity in the TAR Turning back to the bigger picture, the TAR economy has been growing rapidly—although no longer as rapidly as some other western provinces— through a deluge of subsidized government expenditure (at 109 percent of GDP in 2010, most subsidized) and investment (at 91 percent of GDP in 2010, probably most subsidized). The combination of these two amounted to 200 percent of GDP in 2010, that is, the amount of money that the TAR government was spending in the TAR in 2010 or that various entities were investing in the TAR in 2010 was equivalent to double the value of economic activity of the TAR in that year. As noted above, this combined measure includes some double counting given that some government expenditure is spent as investment, although the data to account for this have not been available since 2006. In 2006, when TAR government expenditure was 69 percent of GDP and investment was 79 percent, the combined total minus government expenditure on capital construction amounted to 128 percent of GDP. The equivalent was surely much higher in 2010, quite probably in excess of 150 percent of GDP. The only other province that would have come close to the same level of subsidization was Qinghai, where the



Instituting Economic Growth and Marginalization in Tibet 173

local government deficit reached 47 percent of GDP in 2010 and investment reached 75 percent of GDP, although given that investment in Qinghai was only 13.6 percent financed through state budgetary appropriation, the overall level of subsidization was most likely well below 100 percent of GDP. It is only in the TAR where such absurd levels of subsidization have occurred. With such intensive subsidization, that there would be high rates of per capita GDP growth is not at all surprising. Rather, it is the apparent sheer inefficiency of such subsidization that is striking, as discussed above. In previous publications I placed more emphasis on the institutional reasons for inefficiencies in the TAR, focusing primarily on the aspect of ownership, whereby most aspects of the economy outside the primary sector are based on forms of ownership located outside the province, whether state or private. Externalized ownership results in a variety of perverse dynamics and outcomes, including: the tendency to contract construction or other work to out-of-province companies for reasons of project efficiency, regardless of concerns for economic efficiency or else broader (local) development objectives (versus national development objectives);28 inefficient, poor-quality construction projects ill conceived for local needs and with little or no accountability once construction projects are completed; the tendency to marginalize Tibetan participation in project work to the lowest skill levels; and the pilfering of subsidies before they even arrive in the province by the actors controlling the externalized flows. I have detailed these perverse dynamics and more in works (2005a, 74-82; 2007, 161-65) based largely on my fieldwork from 2003 to 2005. I will not repeat the analysis here in any more detail. The relevance of these points notwithstanding, the much more straightforward reason why government expenditure and investment could be so much larger than GDP is simple: a large proportion of subsidies were spent on imports from abroad or from elsewhere in China. This import intensity might itself be the product of the institutional dynamics listed above, insofar as externalized patterns of ownership would also tend to accentuate the demand for and use of imported goods and services. Moreover, efforts by the government to improve “efficiency,” understood as the conversion of investment into more and better capital formation, might imply even further intensification of the use and dependence on external goods and services. Hence, this aspect will be elaborated here, as it speaks to the rapid rerouting of the productive and consumption structure of the TAR in ways that will be very difficult to reverse, particularly within a context of urbanization as is occurring in the TAR. To start with the international dimension, a large proportion of government expenditure and investment was effectively being spent on imports from abroad, whether directly or indirectly. This is revealed by the expenditure approach of GDP accounting. According to the expenditure approach,

174

Chapter 4

net imports (i.e., imports in excess of exports) are deducted from GDP calculations (i.e., GDP = consumption + investment + net exports), whereas they are not deducted from the government expenditure or investment data if these are spent on imports. Hence, the expenditure or investment data can appear inflated in comparison to the GDP data to the extent they are associated with a trade deficit, particularly if they are the cause of the trade deficit. This has been the case from 2003 onward in the TAR and to a lesser extent in Qinghai from the late 1990s onward. According to the expenditure approach, the trade balance for the TAR suddenly fell from a slight surplus in 2002 to a massive deficit equivalent to 48 percent of GDP in 2003 and then 71 percent in 2004. There was a correction after 2004, although the trade deficit remained at a level around 50 percent of GDP and then plunged again in 2010 to 76 percent of GDP (see figure 4.13).29 The two periods when the trade account plunged into severe deficit (2003-2004 and 2010) both corresponded to strong increases in the ratio of investment to GDP (as shown in figure 4.11) and especially in the ratio of gross capital formation to GDP (as shown in figure 4.12). The re-descent in 2010 was also strongly associated with the sharp increase in government expenditure to GDP (figure 4.9), suggesting a continued use of government expenditure for investment and capital formation. In 2003-2004, the trade deficit is probably explained by the import of high-tech goods and related services for the railway construction that took off in the TAR in 2003 and peaked in 2004, such as rail carriages from Bombardier in Canada, IT goods and services from Nortel, also in Canada, and the subcontracting of Japanese Figure 4.13: Trade balance (net exports) as percentage of provincial GDP, 1993-2010

Sources: calculated from CSY (2011, table 2-21) and equivalent in earlier CSYs.



Instituting Economic Growth and Marginalization in Tibet 175

and other international engineers. The variations in the trade deficit since then probably also reflect the import-intensive input needs of various large infrastructure projects in the TAR, such as the continued extensions of the rail network. The trade deficit of Qinghai also fell to 40 percent of GDP in 2003—probably also linked to the railway and related constructions—but after this point the trade deficit of Qinghai remained more stable, gradually decreasing to around 30 percent of GDP later in the decade. Assuming that overall subsidies to the TAR, including government expenditure and investment, were in the range of about 150 percent of GDP in 2010 (probably a fair, if not underestimated, guess), then deducting the trade deficit of 76 percent of GDP from this amount would leave an effective rate of subsidization of about 75 percent of GDP. Obviously, not all subsidies would have been spent on imports, but the net monetary result is the same: government subsidies amount to a direct and immediate monetary injection into the local economy (if they are not siphoned off before arriving in the province), while the trade deficit amounts to a direct and immediate monetary drain out of the local economy (or at least in the accounts of those state and corporate entities controlling both subsidies and import purchases—both transactions might simply amount to a credit and debit balance sheet entry in terms of their relevance to the local economy of the TAR). This aspect of national accounting explains much of the apparent absurdity of the surge of subsidies in the 2000s, although not all of it. Notably, total subsidies (direct and indirect) in the TAR probably started to exceed GDP from 2001 onward, as noted above, even though the province recorded a trade surplus in 2001 and 2002. In other words, the overwhelming weight of subsidies in the TAR economy is more endemic than simply a decade-long spending spree by the government on the latest high-tech railway and related gadgetry. The more inherent and intractable dimension of this problem essentially follows the same logic, except with respect to the rest of China rather than the international economy. Similar to international trade, the TAR most likely runs a very large trade deficit with other provinces in China, particularly with Sichuan, which supplies much of the interprovincial goods and services to the TAR. Data is simply not available to measure this interprovincial trade balance, although it is actually the simplest and most straightforward explanation for the various macroeconomic attributes of the TAR and is in accordance with earlier analyses made by Wang and Bai (1991 [1986]), as discussed further below. It became impossible to measure interprovincial trade balances since the breakdown of state control over such trade in the early 1980s. The problem is similar to the oft-noted anomaly that provincial GDP growth rates in China are often all higher than the national GDP growth rate (whereas we would expect some to be higher and some to be lower given that the national GDP growth rate would be a weighted average of all provincial GDP

176

Chapter 4

growth rates). The response to this by statisticians knowledgeable about national accounting in China is that the anomaly is not necessarily due to deception but to a technical problem of double-counting output given the effective impossibility of monitoring the flow of goods across porous provincial boundaries and the lack of interprovincial trade data since the early 1980s.30 Double counting is a particular problem when intermediate goods are involved (because these are normally supposed to be excluded from GDP measures except when exported), although increasingly complex production networks within China have increased the prominence of interprovincial trade in intermediate goods, just as global production networks have done the same across countries in East and Southeast Asia. The problem of accounting for intermediate goods is not necessarily a concern in the TAR because this province is not involved in any significant way within interprovincial production networks. However, the lack of interprovincial trade data is a problem because it means that an interprovincial trade deficit does not get deducted from the provincial GDP. In this important respect, we actually have little idea of the effective level of GDP in the TAR, if GDP were properly calculated by deducting the trade deficit with the rest of China. An unknown but definitely large and probably increasing portion of the TAR GDP would effectively represent the net purchases of goods and services from other parts of China rather than value-added produced in the TAR. In this way, an increase in subsidies that is used to finance the purchase of imports from elsewhere in China is effectively reflected in GDP accounting as an increase in value-added output produced in the TAR even though it is not. The same concern would face other provinces, although if we are to judge by the international trade balances, presumably no other provinces face the extent of interprovincial trade deficit that the TAR does. While we do not know the magnitude of the interprovincial trade deficit in the TAR, we can surmise that it is very large and is probably increasing. Structural shifts in the local economy have proportionately shifted the structure of demand toward imports (from instant noodles and mobile phones to motorcycles, sedans, and beyond). These structural shifts include: the intensified integration of the TAR into the rest of China; the relative shift of economic activities out of the primary sector (which is the only sector in the TAR that is largely based on locally produced inputs) and into import-intensive activities such as construction; and the related urbanization of the TAR population, which then adopts more import-intensive consumption patterns. An important counterbalance on the export side would the booming trade in caterpillar fungus, although this too mostly goes unrecorded. The expansion of tourism in the TAR can also be seen as one strategy to right the trade imbalance through the service account, although tourist industries have an endemic tendency—particularly in poor



Instituting Economic Growth and Marginalization in Tibet 177

peripheral regions where the tourism sector is dominated by outsiders, as it is in the TAR—to intensify import demand and also to exacerbate other forms of financial outflow, thereby canceling out much (and sometimes all) of the impact this sector can have on improving a trade balance. Increasing mining output might also provide some counterbalance insofar as this output is mostly exported to other provinces in China, although as observed previously, mining remains relatively limited in comparison to the extent of the deficits involved and is largely dominated by outsiders. Beyond these obvious counterweights, it is likely that a substantial portion of subsidies are financing a large and probably increasing interprovincial trade deficit and are thereby being directly and immediately siphoned out of the local economy, if they ever entered. Once both the international trade deficit and the interprovincial trade deficit are deducted the effective weight of subsidies in the local economy would be considerably reduced, perhaps to a level not much greater than in other western provinces (albeit with an augmented consumption of imports in its wake). This helps to explain, for instance, why the deficit/subsidy per person in the TAR was seven times the provincial regional average in 2010, whereas government expenditure on education per person was only about two and a half times the provincial regional average (education spending is not import intensive). The subsidies that manage to trickle down past the major diversions and drains of funds might still amount to considerable economic stimulus for the local economy, generating the types of activity studied, by Goldstein et al. (2008; 2010) and Childs et al. (2011) around Shigatse in the TAR. However, in the highly perforated nature of such an increasingly and pervasively import-dependent economy, these activities could be become quite cloistered and underlaid with a strong tendency for monetary outflows if and when there would be any lapse in the constant institutional vigilance of directing monetary flows to various peripheral parts of the province (i.e., most areas outside the main urban centers). A Revenge of Fiscal Maoism? This situation is not new. It has been a consistent feature of the TAR since the government started to intensively subsidize the region in the late 1960s. It was one of the main critiques made by Wang and Bai (1991 [1986], 71-76), who argued that the negative multiplier of subsidies on growth in the TAR from 1957 to 1983, as discussed above, was due to the interaction of four key issues in the TAR and other western provinces, namely: financial flows; the external trade gap; the marked failure of statesponsored industrialization projects; and the “profit feedback loop.” They noted, for instance, that the trade gap (i.e., trade deficit) between Qinghai and the rest of China was equivalent to 84 percent of total subsidies to

178

Chapter 4

Qinghai in 1983, and the trade gap of the TAR was equivalent to over 58 percent of total subsidies to the TAR (or to the same proportion of total output given that subsidies were equal to total output in that year). They pointed out that 94 percent of merchandise sold in the TAR in 1983 was imported from the rest of China. Net financial outflows from these western provinces were in some cases also equivalent to the entire sum of central government subsidies flowing into the same provinces. Finally, they gave substantial evidence of the failure of local industrial development in the TAR, highlighting cases of loss-making ventures, projects that fell flat, were written off or that otherwise went to waste, or funds that disappeared without a trace (88-89). They argued that this was related to a “profit feedback loop,” whereby the financial flow of funds led to “a covert compensation mechanism . . . from the developing to the developed areas, at the same time as the central government is taking from the rich regions to assist the poorer areas” (79). These economic arguments were seminal and incisive at the time. They remain relevant, with some modifications to account for the liberalized nature of the economy in China. Nonetheless, Wang and Bai (1991 [1986]) used their observations to make the case for a deeper, culturally derogative interpretation of these perversions of development. They attributed the causes of the four highlighted issues and the apparent lack of entrepreneurship in these minority areas as stemming from the “intrinsic backwardness” of the minority populations in these western provinces. Such backwardness, they claimed, resulted in a lack of investment demand rather than a shortage of savings. As a result, subsidies remained under- or poorly utilized, local capital accumulation did not take off with opening and reform as it had among the eastern Han, and nothing inhibited funds from flowing out of the region or being spent on imported consumption goods rather than synergizing a nascent capitalist ethos (71-72). In other words, while Wang and Bai offered a variety of interesting findings and criticisms of central government policies, their ultimate diagnosis ironically deflected criticism away from the externalized patterns of ownership and control in these minority regions by locating the central source of blame on local minority modes of production and accumulation. Their criticisms of central government policies thereby reinforced a wider “Han chauvinism” toward these minority areas. This is similar, for instance, to the way that the so-called Post-Washington Consensus in international development policy did not emerge as a critique of previous “Washington Consensus” structural adjustment policies per se, but as a view that these policies were poorly implemented or misimplemented in developing countries due to a variety of governance and institutional problems inherent to these countries. Institutional weakness therefore prevented market reforms from functioning in the way they are presumed to function in



Instituting Economic Growth and Marginalization in Tibet 179

developed economies. While the policy content is different in this comparison, the logic is similar in terms of blaming the victim, or to borrow from the famous words of Carlos Diaz-Alejandro (1984, 335) in his seminal post-mortem of the Latin American debt crisis in 1982: “Blaming victims is an appealing evasion of responsibility, especially when the victims are far from virtuous.” The logic thereby tends to undermine arguments for empowering local agency through the enhancement of local ownership, but instead evokes the need for a return to vanguard state paternalism in dealing with minority regions. This is especially ironic given that the TAR rural economy had barely de-collectivized at the time of the fieldwork of Wang and Bai, after being completely collectivized a little more than a decade earlier. Or, there was little that was particularly “intrinsic” in the conditions the authors had observed during their stint in these regions. The urban economy itself remained almost entirely state controlled, more likely managed by Han rather than Tibetan cadres up to 1981, both probably somewhat disoriented by the ambiguous and often lagged reforms coming by way of directives from the east, to be implemented in a region still largely dominated by geopolitical security concerns. Various versions of such argumentation remain common among scholars, officials, and the generally erudite public in China since the Chinese publication of this book in 1986, as I often noted during my fieldwork and continue to note in my regular visits to China.31 For instance, many echoes can be perceived in the books reviewed by Yan (2000). Whether Wang and Bai were the origin of this particular narrative or else were simply putting an academic voice to commonly held views is a moot point, although the influence of a particular variant of Marxist modernization theory is notable, similar for instance to certain Marxist arguments in the West regarding the historically progressive impacts of colonialism.32 These arguments will be discussed further in chapter 7. Here it suffices to note that, despite the market discourse that is now gaining currency in the TAR and other Tibetan regions, it appears that little has changed in terms of the very visible hand of the state in structuring most aspects of the economy, including the rural economies.33 In particular, the central government has been overwhelming in establishing and reinforcing externalized patterns of ownership as the institutional basis for the TAR economy, thereby entrenching an extreme form of dependency as the economic logic and modus operandi of the local economy. This point is not in contradiction to the discussion of neoliberalism by Yeh (2013), who refers to a very selective use of neoliberal discourse introduced in Tibet to argue that Tibetans should welcome migrants and that barriers to free flows of people are not acceptable. The adoption of these certain select elements of a neoliberal governmentality or subjectivity has ironically occurred in a context whereby the state has remained omnipresent in the local economy.34

180

Chapter 4

Conclusion: the Instituting of Extreme Dependence in Tibet This chapter examined the instituting of rapid growth in the Tibetan areas of western China since the mid-1990s. This was first analyzed in terms of the sectoral structure of rapid economic growth in the TAR since the early 1990s. Sectoral transformation in the TAR economy occurred much more rapidly than in other western provinces in China and was uniquely based on alternating bouts of expansion in either government administration or large-scale construction projects, both disconnected from local productive activities. Second, growth was driven by heavily subsidized state-led development strategies of expenditure and investment, which together reached heights never observed before in the history of the TAR or elsewhere in China. In particular, direct budgetary subsidies from the central government to the TAR surpassed 100 percent of GDP for the first time ever in 2010, while investment in fixed assets reached more than 91 percent of GDP. The resulting inefficiencies are explained by the fact that western development strategies have intensified international and domestic trade deficits as well as financial outflows. Extreme examples of such strategies include the Qinghai-to-Lhasa railway completed in 2006, ongoing extensions of the railway to other locations in the TAR, and other large-scale projects (including the recently announced building of a thirty billion yuan tourist theme park outside Lhasa—see Smith 2012). The institutional character of these strategies was well entrenched since the late 1960s. Despite twenty years of intensified development efforts, the TAR remains locked in the institutional norms that have guided the subsidization of this politically sensitive autonomous region since the Maoist period. As a result, recent strategies have not altered in any significant way the long-term trend of very intense and very inefficient subsidization. Rapid economic growth since the mid-1990s has been a reflection of the intensification of subsidies. Subsidies have left more infrastructure in their wake, but at the expense of further entrenching the import intensity of the economic structure, to the extent that we do not even know the true value of the TAR GDP given the pervasion of interprovincial imports in the local economy. In contrast to every other province in China, these institutional norms appear to have been guided primarily by security and ideological rather than (political) economic concerns. In this respect, the recent phase of intensified subsidization has completed two principal tasks envisaged during the Maoist era. One is the state-led engineering of a deep integration of the TAR into China through externalized patterns of ownership, thereby entrenching an extreme form of dependency as the economic logic and modus operandi of the local economy. The second is the consolidation of



Instituting Economic Growth and Marginalization in Tibet 181

the very visible hand of the state in the structuring of most aspects of the economy, including the rural economies, albeit through a different mode of governmentality attuned to the current era of “market socialism” rather than to Maoist collectivism. As a result, the economy of the TAR can be aptly described in macro-structural terms as having become a peripheral subsidiary of the central government and related interests. Local development dynamics (and people) are increasingly captive to the discretion of these central interests, particularly in the context of rapid transition away from their traditional bases of subsistence in the rural economy, as discussed in the next chapter and in chapter 7. Indeed, the fact that these development dynamics have occurred alongside a rapid shift in people’s livelihoods out of agriculture implies that this most recent phase of subsidization has probably irremediably locked the fate of Tibetans in this province—perhaps intentionally—into a trajectory of extreme economic dependence with the central government. The perverse characteristics of this subsidized economic model can hardly be attributed to local Tibetans—as is often implied by Chinese scholars—given that local Tibetans have had little to do with the conception or path of this model, including the most recent phase that started when western development strategies were re-invigorated in the mid-1990s. Rather, the institutional characteristics of ownership within this economic model are important determinants of dependence and inefficiency in the TAR. Beyond the more dramatic cases of wasteful spending and corruption, externalized patterns of ownership are at the root of the problem of trade deficits and financial outflows, rather than lack of investment demand or attributes of “intrinsic backwardness” as argued by Wang and Bai (if indeed their hypothesis has any validity). The externalized patterns result in a low circulation of profits and wages in the local economy due to considerable leakage from the province. Out-of-province companies tend to retain and “repatriate” their profits from lucrative construction contracts, investing them in other national projects rather than in the local economy. Companies and their staff and workers usually return home or to other national jobsites upon completion of the projects, taking the benefits of the acquired skills and earnings with them, rather than investing or spending them in the local economy. In other words, money goes in and goes out, without much turnover to benefit local production or demand, besides a skimming of trade and services, which again is dominated by outsiders and based on imports from elsewhere in China. Within this context, there is little hope that growth could be sustained beyond current subsidy intensifications. The much-touted tourism and mining activities still pale in comparison to government administration and related activities that are the main targets of the subsidization strategy. Indeed, tourism, which has been proposed as one of the key pillars

182

Chapter 4

of future growth for the TAR, functions in a manner similar to the broader model, insofar as much of the sector is controlled by out-of-province businesses and employment is dominated by migrant labor. A large share of tourism revenue therefore simply leaves the region after a short circulation, perhaps not much longer than the tourists themselves, or else is saved for later repatriation. Investment in tourism is similarly limited by its enclave nature, aimed at accessing and profiting from tourist circuits while carrying few incentives for the diversified reinvestment of profits in the local economy once such access is sufficiently leveraged. Under such circumstances, tourism will most likely accentuate the externalized flows of wealth and is unlikely to substantially improve the interprovincial trade balance. The Tibetan medicine industry, which is increasingly attracting the attention of Han investors, follows similar patterns whereby ownership and head offices are often located outside Tibetan areas. Even if mining proves to be a new “pillar,” it will also tend to reinforce the already entrenched dualism of the economy given its high-tech and enclave character, together with the fact that processing will most likely take place outside the province. In sum, in the absence of protection or promotion of locally oriented forms of ownership, subsidy intensification simply reinforces these externalized patterns of ownership, which in turn provide the institutional foundation for accelerating the externalized wealth flows. Local processes of accumulation are short-circuited as a result, albeit not for the reasons argued by Wang and Bai, who confused symptoms as causes by postulating that inefficiencies are due to the intrinsic backwardness of the minorities themselves, rather than to their political disempowerment and subordination to Han-centered political and economic processes. These institutionalized dynamics of growth create the conditions for effective discrimination through a form of economic segregation based on privileged access to the state-controlled levers of almost all aspects of the economy outside the rural areas. Because flows of wealth are increasingly centered outside the local economy, the main beneficiaries of growth are those who are well positioned to access these externalized flows, be they Tibetan or Chinese, government officials, traders and businesspeople, or others. Moreover, because these flows are controlled in ways that reflect the dominance of the Han Chinese majority, they structure advantage along axes that reinforce this dominance along linguistic and cultural modes of bias, such as Chinese fluency, Chinese work cultures, and connections to government or business networks in China proper. The social, political, and economic reproduction of this minority in the provincial labor force therefore bears little relation to the indigeneity of the Tibetan areas, in terms of the socio-economic circumstances that condition the lives of most Tibetans. There are few incentives to direct such elite wealth into locally oriented forms of accumulation given that much more money can be made



Instituting Economic Growth and Marginalization in Tibet 183

with much greater ease through privileged access to state patronage, rather than through painstaking and relatively petty profit-making activities based in local productive activities. As a result, there is little dispersion or linkage of the externalized flows of wealth into the local economy, and outflows of wealth are privileged at the expense of local reinvestment. Those excluded from privileged access, whether they be the majority of Tibetans or even many Chinese migrants, are in turn progressively marginalized from these dominant processes of accumulation driving growth in the local economy, even though their material consumption might nonetheless improve in the process, as it has for the average Tibetan in the TAR and other Tibetan areas. Yet it would be surprising if these improvements had not taken place given the sheer torrent of subsidies that has been channeled into the TAR. In this sense, state-subsidized development strategies carry an important ethnically discriminatory institutional bias, whether or not they are intended as such. This understanding is an important distinction from the contention that ethnic inequalities in the TAR and other Tibetan areas are mostly the result of spatial inequalities, that is, that Tibetans are mostly rural and poor, hence ethnic inequalities reflect urban-rural inequality, as argued by Sautman and Eng (2001) or Hu and Salazar (2008). Rather, the institutional orientation of development strategies has had a strong tendency to short-circuit locally oriented processes of wealth circulation in deference to sources of wealth that are located outside the local economy. This tendency needs to be understood as a predominant factor driving structural polarization as well as effective institutional modes of ethnic bias even within Tibetan urban areas, as analyzed further in the next two chapters. As a result, most local Tibetans—both rural and urban—face increasing hurdles to integrate into the rapidly growing parts of the economy, particularly in light of their much lower schooling levels, especially at the secondary level where fluency in Chinese is mostly acquired. These challenges are exacerbated by the context of open migration of Han and Muslim Chinese who are attracted to the artificially subsidized affluence of the cities and towns of the TAR. Yet the migrants are hardly to blame, as they are merely following the same economic incentives that attract many Tibetans from the countryside to these same cities and towns, including many Tibetans from Tibetan areas outside the TAR. Notably, these processes have rewarded a small upper stratum of the population in the TAR, including a small minority of Tibetans and a large proportion of non-Tibetan migrants, concentrated mostly in urban areas and well positioned to access the flows of wealth as they pass through the region with increasing velocity. The wealth of these well-positioned people is increasingly determined by factors located outside the local economy and instituted along lines that confer advantage not only to political obedience, but more generally to cultural or linguistic attributes deriving from mem-

184

Chapter 4

bers of the dominant group that controls the externalized flows. While the assimilationist tendencies might well be intentional as a matter of policy (given the conviction of many Chinese leaders of the inevitable need for Tibetans to assimilate in order to become functional Chinese citizens), the discriminatory tendencies might not be intentional, at least not as a matter of policy (even though they might be intentional as a matter of informal practice, e.g., see ICT 2012a). Regardless, the discriminatory tendencies are more generally produced through these structural and institutional dynamics underlying development, alongside considerable improvements in the material consumption of local Tibetans.

Notes   1.  According to the 2000 population census, 63 percent the population in the TAR with their place of household registration based outside the TAR were from Sichuan, or 68,496 people out of a total of 108,669 (Tibet Census Tabulation, table 7-2). Lower shares for Sichuan were measured by Ma and Lhundup (2008, 16-17) in their survey of 1,470 migrants in Lhasa in September-October 2005, in which 30 percent were from Sichuan and 24 percent from Gansu, including both Tibetan and Han. However, it is not clear to what degree their survey was representative of the broader migrant population.  2. One challenge of future research would be to extend the types of macrostructural analyses that we can do with the TAR provincial data to the prefecture or county-level Tibetan areas outside the TAR.  3. These rates are different from those recorded at comparable prices in the yearbooks. The calculation here is based on the per capita GDP for each province reported in each annual yearbook in table 2-15, indexed by the provincial CPI in each year. The rates reported in the yearbooks must be using a different index, although this is not specified. Nonetheless, the data provide a rough idea of the growth accelerations that took place in western China from 1995 onward, particularly in comparison with previous data presented in figure 2.1 in chapter 2.   4.  Discussions of price dynamics and agricultural output can be found in Brandt et al. (2002, 97) and Lu (2000, 196). The terms of trade can be found in the CSY (2001, table 9-3), measured as an index of industrial over farm prices. Terms of trade for agriculture fell again during the post-1997 national price disequilibrium.   5.  Price inflation from 1997 onward was relatively marginal, unlike the previous period. The general CPI for the TAR increased only 23 percent between 1997 and 2010. Obviously, underlying this relatively low rate of price inflation were quite significant shifts in relative prices, particularly between agricultural and manufactured goods on one hand and services on the other (the former were deflationary for several years after 1997 whereas the latter were very inflationary). However, it is not possible to account for these relative price changes in the provincial accounting given the lack of appropriate data.   6.  The main categories of the tertiary sector since 2004 (as reported in the yearbooks) include: Transport, Storage, and Post; Information Transmission,



Instituting Economic Growth and Marginalization in Tibet 185

Computer Services, and Software; Wholesale and Retail Trades; Hotels and Catering Services; Financial Intermediation; Real Estate; Leasing and Business Services; Scientific Research, Technical Services, and Geologic Prospecting; Management of Water Conservancy, Environment, and Public Facilities; Services to Households and Other Services; Education; Health, Social Security, and Social Welfare (mostly health care); Culture, Sports, and Entertainment; and Public Management and Social Organizations (was “Government and Party Agencies, and Social Organizations,” which is essentially the state and para-state administrative apparatus). These categories are mostly the same in the previous reporting prior to 2004, except that some components have been reorganized under other categories, making a precise evaluation of trends between the older and newer categories difficult in several cases.   7.  It is not clear whether these two shares correspond precisely to the same content because there appears to have been a correction to the data in 2004, parallel to the change in the terminology, from 5.6 percent in 2003 to 9.5 percent in 2004. Moreover, the trend was falling up to 2003.  8. For further discussion of these speculations, see Fischer (2005a, 44-45). Barnett (2011) also observes that, since 2006, government spending per person on security in the Tibetan areas of Sichuan where the self-immolations occurred in 2011 and 2012 was 4.5 times greater than in neighboring non-Tibetan areas and increased at twice the rate up to 2009. He therefore suggests that a security buildup had begun in these areas prior to the major protests in 2008.   9.  As noted in the previous chapter, according to data presented by the then TAR governor Padma Choling, tourist numbers in the TAR (mostly domestic Chinese) rose from 1.9 million in 2006 to 6.82 million in 2010 (see Tibetinfonet 2011). Tourist numbers would have exceeded the total population of the TAR of about 2.8 million in 2007. 10.  For good discussions of these dynamics, see World Bank (2002b), Tsui and Wang (2004), Wong and Bird (2005), and Tao et al. (2010). 11. The shares of central and local government revenue and expenditure are taken from CSY (2011, tables 8-3 and 8-4). The measures of the regional total fiscal balance are calculated from CSY (2011, tables 8-7 and 8-8) and equivalent tables in previous yearbooks. 12. The difference was mostly compensated for by a much higher share of “compensation of employees” than in all other provinces. The share of “operating surplus” was also much lower than in most other provinces, while “depreciation of fixed assets” was slightly higher than the norm. 13.  The comparison of these two shares is not very appropriate because the latter calculation of TAR government revenues as a proportion of GDP is not calculated from the income approach GDP accounting data but from the government fiscal data, which would also include taxes on consumption and so on. However, the comparison offers a very rough indication of the magnitude of central government taxation in a province relative to provincial/local taxation. 14.  The category of “Expenditure for Post-earthquake Recovery and Reconstruction” in the Sichuan government expenditure data was 19 percent of total expenditure in 2008 (CSY 2009, table 7-8). It remained at the same level in 2009 and 2010. The category was not reported prior to the 2008 data.

186

Chapter 4

15.  Wang and Bai (1991 [1986], 73) state that subsidies were equal to the older national accounting measure of “gross value of agricultural and industrial output” from 1980 to 1983. This output measure was the national accounting measure used by the PRC until the early 1990s. It is not equivalent to GDP measures because it does not include the tertiary sector, among other considerations, and hence is lower than the equivalent GDP measure. 16.  This per capita GDP measure is calculated on the basis of the aggregate GDP data and total population reported in CSY (2011), although for some reason it does not correspond to the per capita GDP cited elsewhere in CSY (2011, table 2-15) as 17,309 yuan. The measure used here is the relevant comparator for this exercise because it corresponds to the same data used to measure the overall deficit/GDP ratio. 17. The reporting categories of government expenditure changed from CSY (2008) onward and no longer included expenditure on capital construction or on government administration as identifiable categories. These were presumably absorbed into other categories. The largest categories of expenditure in the allprovincial government total in 2010 were (ordered from the largest and with a cutoff of 400 billion yuan): education; social safety net and employment effort; general public services; agriculture, forestry, and water conservation; urban and rural community affairs; medical and health care; and public security. 18. Under this system, minority areas have both regular schools taught in Chinese-medium alongside minority schools taught largely in Chinese-medium but with some minority language content. Between the Tibetans, Mongolians, and Muslims in this province, and the fact that most of the province is made up of minority autonomous counties or prefectures in one form or another, this duplication of schools is much more prevalent than in the TAR, where most schools would be equivalent to the minority schools in Qinghai given the predominance of Tibetans in the population, or else in most other provinces in China, where Han Chinese predominate in the population. As a further note, seven out of the sixteen specialized secondary schools in Qinghai in 2001 were secondary teacher training schools, versus only one out of eleven in the TAR. 19.  Data on the number of senior school teachers for 2000 is not available in CSY (2001), and the numbers of secondary schools and teachers might not be comparable between the two years in any case given that the 2000 data do not appear to include specialized secondary schools, whereas the 2010 data appear to include specialized secondary schools. Given the lack of clarity on this matter, the relevance of these data is in the inter-provincial comparisons in one year, not intertemporal across the two years. 20.  The category shift from “staff and workers” to “employed persons in urban units” also appears to involve a slightly broader categorization of urban employed people, in which case the data from these two years are not necessarily comparable and the rate of change not necessarily accurate. See the next chapter for further discussion of this terminological shift. 21. Dreyer (2003, 421) also refers to declining economic rates of return, although she does not explain how these are measured. She appears to cite from the various debates among Chinese economists in the 1990s that she refers to in this section of her article. Presumably, these measures are based on a marginal measure, showing the value of growth that is associated with a unit increase in the value of



Instituting Economic Growth and Marginalization in Tibet 187

subsidies, which is the same as the measure used here. She also refers to measures based on total factor productivity—from which she appears to derive her conclusion of extensive versus intensive growth—although TFP is a theoretical construct from the neoclassical Solow growth model that is arguably poorly conceived and contested by many (see Fischer 2011b). 22. Five-year calculations are preferred because annual calculations make this measure erratic in the TAR given large swings in the alteration between subsidies and investment from year to year, which can give the impression of a huge role of subsidies in one year and little in the next, even though subsidies are operating through investment rather than government expenditure in the off years. The five-year period also allows for consideration of some lagged effects of subsidies on growth, at least for the beginning of the five-year period, which here corresponds to the beginning of a five-year planning cycle. 23.  See note 15 above. 24.  According to earlier methods of accounting, subsidies to loss-making enterprises were calculated as a revenue item rather than an expenditure item, and thus a negative revenue results when these exceed taxes and other fees. 25. For instance, in the case of several small and medium-scale hydroelectric projects that I visited in several Tibetan areas of Qinghai in 2004, funding was raised through private investors from Xining, the provincial capital, and smaller contributions were made by county, prefecture, and provincial governments. Government support also plays an important role in securing loans. 26. Investment in fixed assets (IFA) is usually larger than gross capital formation (GCF). For instance, IFA for China as a whole in 2010 was 27.8 trillion yuan, whereas GCF was only 19.2 trillion yuan. Similarly, IFA was 316 billion yuan in Gansu and GCF was 234 billion yuan. IFA was 1.3 trillion yuan in Sichuan and GCF was 922 billion yuan. The difference is consistent whether there is a trade surplus (China) or trade deficit (Gansu and Sichuan); this is logical given that both fixed assets and capital formation include, by definition, assets purchased domestically or from abroad in addition to assets constructed. In contrast, IFA in the TAR was 46.3 billion yuan in 2010, versus GCF of 56.6 billion yuan. As explained by Shen (2009) in the case of China, IFA differs from GCF in that IFA includes the purchase of land, old equipment, and old buildings, where GCF does not. Other possible reasons include exaggerated statistics of fixed asset investment and low efficiency of fixed asset investment, which would both result in a higher IFA than GCF, or some capital written off by the phasing out of non-efficient projects, which would result in a lower IFA than GCF. In this respect, IFA can sometimes fall below GCF in the event of intensive industrial upgrading and restructuring, as was the case in China in the 1990s and early 2000s during the intensive restructuring of state-owned enterprises during these years, as noted by Shen (IFA was less than GCF in China from at least 1990 until 2003). Alternatively, lower IFA than GCF might be the result of under- or unreported investment that nonetheless results in a capital asset, or else other reporting problems that might have been particularly prevalent as national accounting systems were being changed and refined in the early 1990s. The huge discrepancy for the TAR in 2010, when GCF was larger than IFA by about 20 percent of the TAR GDP, is difficult to explain, although such a scale of discrepancy is specific to this year only. In previous years in the TAR, GCF oscillated

188

Chapter 4

above and below IFA, and the discrepancies were smaller in magnitude and more in line with the patterns in other western provinces. Further research would be required to decipher the reasons for this discrepancy, although given that this would require interviewing the statistical authorities of the TAR, I have little hope that possibilities for such research would be forthcoming in the near future. In any case, the use of GDP data in this book implicitly acquiesces to the GCF data. 27.  In comparison, the share of state-owned units in total investment between 2004 and 2010 also fell in Qinghai, Gansu, and China as a whole, but rose slightly in Sichuan. The fall in the SOU share in the TAR between 2001 and 2004 was due to a rising share of investment by individual, shareholding, and “other” units, which probably represents the expansion of private or semi-private business interests in tourism, real estate, or various construction projects during this early phase of the OWC. The shares of individual and other forms of private investment in the TAR stabilized after 2004. 28.  For instance, it is difficult to see how the Qinghai–Tibet railway, which does generate revenue, will ever become economically profitable, even at maximum capacity, given the amount that was invested in the project. Lustgarten (2006) reported that the Germo-to-Lhasa railway capacity is estimated at just two million tons a year. Citing Lou Thompson, a former advisor at the World Bank who was involved in lending to China for other railway projects, he notes that it would take 300 years to pay off the USD 5 billion of capital invested in the project, based on the USD 13 million annual operating profit of the Alaska Railroad as a measure. This is assuming that no major complications arise due to global warming and a faster than anticipated permafrost melting. From this perspective, this railway project and ongoing extensions are probably best understood as a national industrial policy strategy of subsidizing the buildup of expertise in large and technologically complex infrastructure projects, rather than any rationale related to the economic development of the TAR. 29.  Note that these trade data do not correspond at all to the trade data reported in the Tibet Statistical Yearbook, which for some reason show a trade surplus rather than a deficit over these years, as well as a much smaller value of trade (e.g., see TSY 2010, table 13-1). The discrepancy must derive from the fact that the trade statistics in TSY (2010), which are obtained from local government customs sources, account for only a small portion of the international trade occurring in the TAR. This is probably explained by the fact that most of the trade—especially imports—has been conducted by central government procurement or by national or other provincial enterprises not bound by reporting to local TAR customs authorities. Notably, TSY (2010)—and all previous TSYs—contain very little detailed national accounting data, with no disaggregation according to expenditure or income approaches. More detailed and comprehensive national accounting data for the TAR can actually be obtained from the more general China Statistical Yearbooks. 30.  For a good discussion of this in China, see Rabinovitch (2012), although his explanation is not precisely accurate. The problem with double counting is that the value of an intermediate input produced in one province and then used in the production of a final good in another province is not deducted from the value-added of that final good. Hence, the value-added of the specific input is counted twice, once in each province (whereas it should be counted only in



Instituting Economic Growth and Marginalization in Tibet 189

the first and deducted from the value-added of the second province). The same problem underlies international trade accounting—the traded value of a good processed across several countries in an international production network can be several times the actual value-added of that good. 31.  For an excellent discussion of these narratives, see Yeh (2007). 32.  See Warren (1973; 1980) for a classic example. 33.  I do not engage here with the question of problematizing the notion of the “state” and “governance,” particularly in Tibetan areas, although suffice it to say that this is very relevant field of research, particularly in comparison to research on these issues elsewhere in China. 34.  I am indebted to numerous lengthy exchanges with Emily Yeh on these matters.

5 The Great Transformation of Tibet? Rapid Labor Transitions, Polarization, and the Emerging Fault Lines of Stratification in Urban Tibet

As analyzed in the previous chapter, the economies of the Tibetan areas in western China—epitomized by the TAR—have been growing very rapidly since the mid-1990s. Unlike in the rest of China, rapid economic growth in the Tibetan areas has been disconnected from local processes of productive accumulation. Rather, rapid growth has been the result of a massive degree of subsidization, mostly from the central government and heavily concentrated in urban services and construction. In combination with political disempowerment and outside control of most sectors of the economy besides agriculture, the economy of the TAR can be aptly described in structural terms as having become a peripheral subsidiary of the central government and related interests, or a quintessential aid economy par excellence, resulting in numerous perversions. However, while this growth experience has been to a large extent an accounting illusion, its socio-economic consequences are not. Rather, rapid subsidy-sustained growth has been associated with very real and rapid changes in the socio-economic structure of Tibetan society. These changes have been more rapid than changes occurring elsewhere in China, albeit without the relative autonomy that local people and governments in other regions of China can rely on to mediate the consequences (with the exception of Uyghurs in Xinjiang). Most fundamental has been the rapid transition of the local (mostly Tibetan) labor force out of the primary sector (mostly farming and herding). In the Tibet Autonomous Region (TAR), for instance, the share of the local labor force considered as employed in the primary sector dropped from 74 percent in 2000 (the most agrarian labor force in China at the time) to 53 percent by 2010—a reduction of 191

192

Chapter 5

over twenty percentage points in a decade. This shift out of agriculture was mostly absorbed by rapid increases in the shares of labor employed in services and to a lesser extent in construction, as would be expected by the GDP data analyzed in the previous chapter. Compared to other parts of western China, the speed and character of employment transition as represented by official data has been exceptional, to the extent that within one decade the TAR has to a considerable degree caught up with the (also rapidly changing) norm in China, albeit without the productive economic foundations to support these changes as elsewhere in China. Moreover, the speed of such transitions in Tibetan areas outside the TAR might well be even faster given the implementation of resettlement schemes in pastoral areas (which have largely bypassed the TAR to date), the earlier implementation of education and employment reforms, and the closer integration of these areas into neighboring Han Chinese urban centers. For better or for worse, the consequences of these transitions in Tibet deserve urgent attention, particularly if they prove to be irreversible. Indeed, the question of irreversibility deserves some attention for the framing of this chapter. Some of the decline in the Tibetan primary employment share probably reflects migratory workers who are still fairly well embedded in the rural economies from which they seasonally emigrate for part of the year in search of off-farm employment. Goldstein et al. (2008) refer to this as the new trend of “going for income” in the rural areas of the TAR, while Cencetti (2012) refers to it as the “new nomadism” of rural or semi-urbanized Tibetan labor migration in the Tibetan area she studied in Qinghai. These local migrants might not be registered as primary sector workers even though they continue to work in the primary sector for at least part of the year or, conversely, they might be registered as working in the primary sector even though they also engage in informally organized off-farm work. In either case, the official data probably exaggerate the degree to which the local labor force has become disembedded from the rural economy. “Disembedding” in this sense refers to a Polanyian understanding of labor commodification, meaning that rural labor is forced into wage employment through dispossession from the land or through other forms of economic or non-economic coercion. Some dispossession has undoubtedly taken place with the resettlement of pastoral communities in parts of eastern Tibet, although apparently not in the TAR, at least not yet (see Bauer and Nyima 2011). More generally, the continued embeddedness of rural labor in the Tibetan areas might be taken to imply that these labor transitions could be reversible if urban employment opportunities were to become more austere, in the sense that these migrants could easily return to farming or herding, as has sometimes happened even in the case of the more coercive resettlement policies in eastern Tibet (e.g., see Cencetti 2012). Nonetheless, such



The Great Transformation of Tibet? 193

migratory employment patterns do not necessarily lessen the sense of comparative rapidity that the official data reflect regardless of their precise accuracy given that similar migratory considerations also apply in other parts of western China. Moreover, from a global demographic perspective, we can expect that these transitions will probably continue once started, in the broad structural sense that populations rarely move back into farming or herding once they have moved out of these activities (short of some massive traumatic event).1 The migratory employment patterns discussed above are also fairly typical in early to middle-stages of urbanization. One of the most powerful mechanisms of transition in this regard is education rather than employment. For instance, my own qualitative observations among secondary students in the Tibetan areas of Qinghai, Gansu, and Sichuan suggest that once young people leave their rural areas for a few years to attend boarding schools in towns, especially at the secondary level, they rarely return to farming or herding, and their families usually consider them lost causes with respect to these occupations. Such students might return temporarily to their rural households to help out, particularly during summer holidays or spells of postgraduate unemployment, but I have rarely come across secondary students who express the desire or intention to move back into farming or herding as an occupation.2 This observation corroborates with the findings of Iselin (2011), Cencetti (2012), and Zenz (2013), and it seems to confirm the broader observation, as noted by White (2011, 1), of “mounting evidence from all over the world . . . that young people are increasingly uninterested in farming or in rural futures.” Hence, the structural shifts observed in the employment data discussed in this chapter plausibly represent the unleashing of profound social transformations that, once started, are unlikely to reverse, even considering the rural embeddedness of migratory labor or else the potential prospect of dire economic conditions in the urban areas. These transformations will obviously not spell the death of farming and herding in Tibet, but they will undoubtedly change the nature of farming and herding within the broader socio-economic system. To the extent that many of these socio-economic changes might be irreversible, they highlight a variety of concerns particular to the disempowered circumstances of Tibetan areas and to the role of government policies in mediating the pace and character of change. For instance, the dependence on subsidies to sustain the conditions on which an increasing number of rural Tibetans have come to rely presents a constant source of politicized uncertainty. The continuing, if not strengthening, dominance of Han Chinese in the urban economies of Tibet and the associated urban exclusionary pressures faced by Tibetans compounds contention, as was arguably evidenced by the outburst of large-scale protests in March 2008.

194

Chapter 5

However, because urbanizing migrants tend to be upwardly mobile, such urban exclusions are not necessarily reflected in poverty or even inequality statistics. This chapter analyzes these socio-economic transformations through a longitudinal trend analysis of aggregate employment, wage, and national accounting data, as well as household income and expenditure data from household surveys and my own insights from fieldwork (see the appendix of this chapter for a detailed explanation of the data sources used). The analysis continues the structuralist and comparative method of the previous chapter, highlighting the cases of the TAR, Qinghai, Gansu, Sichuan, and China. The structural changes in the labor force are followed by an analysis of changing income and wage inequality within and between rural and urban areas. The temptation to report poverty measures in this chapter is avoided for several reasons. First, the rural income distribution data for the TAR have not been made available since TSY (2000). While they have been made available for Qinghai, these are with reference to a majority Han Chinese experience. Second, through the course of my fieldwork and subsequent analysis, I have become very cautious in relying on the official rural household income data, particularly with reference to pastoral or semi-pastoral areas, as explained in chapter 7. Third, it is questionable the degree to which the rural household income data account for the enormous boom in the caterpillar fungus harvesting and trade that has taken place in many Tibetan areas since the 1990s (e.g., see Winkler 2008a, 2008b). Finally, the household income surveys include only households registered as permanently residing; this is less of a problem in the rural data but a huge problem in the urban data given the exclusion of migrant (or otherwise temporarily registered) households. Indeed, all these points even call into question the use of household income data to measure urban-rural inequality, as I have done in this chapter, and I have come to rely more on measures of wage inequality precisely for this reason. However, wage data are reported only for urban areas, not rural areas, and only for a portion of the urban labor force. Nonetheless, the general story is that absolute poverty measures have been falling in the context of rapid growth, as would be predicted, meaning that basic material consumption has been increasing. This is evident in the field as well, reflected by the marked increase in conspicuous consumption by Tibetans across all Tibetan areas. Instead, the main concern of this chapter is with the processes of dislocation and relocation of livelihoods in the course of fundamental structural transitions out of agriculture and the rural areas altogether, which the income measures are particularly weak at capturing, particularly when these transitions involve an increasing monetization and commoditization of the rural economy.



The Great Transformation of Tibet? 195

Various dimensions of inequality that have emerged and subsided in the course of these transitions help to clarify some—although not all—aspects of these processes of dislocation. Urban-rural inequality increased sharply in the late 1990s and early 2000s but then subsided with the invigoration of a rural focus in development policy from about 2003 onward. However, there was a sharp (and administered) increase in urban inequality in the TAR from the late 1990s up to 2007, the eve of the wave of protests that rocked the TAR and other Tibetan areas in China, suggesting that intra-urban inequality had taken over from urban-rural inequality as the dominant locus of polarization in the TAR over this period. This is particularly significant given that the sharp increase in urban inequality occurred at such a key moment of the labor transition out of agriculture, thereby having a powerful effect on establishing the foundations of future social stratification in the urban areas of the TAR. The fact that this occurred alongside a shrinking number and share of Tibetans among the privileged employees of the highly subsidized state sector (at least up to 2003) speaks to the ethnic dimensions of such rising intra-urban inequality that had nothing to do with either urban-rural differences or to the low (albeit rising) levels of schooling among Tibetans in the TAR, thereby clearly dismissing the contentions of Sautman (2008) and Hu and Salazar (2008), for instance. In other words, underlying some heavily subsidized silver linings in the rural areas of the TAR (if the rapidity of the changes in these areas is to be taken as positive), there was a broader overarching trend of heightened polarization in the overall economy, which was more or less instituted by the government through its subsidy, wage, and state-sector employment policies. This trend and the related educational, linguistic, and cultural modes of bias that severely disadvantage the majority of Tibetans within their urban labor markets provide important insights into the outburst of protests in March 2008 across the Tibetan areas. The chapter is divided into three sections. The first focuses on the changing characteristics of employment structure in the TAR and the other provinces. In the second section, these employment trends are combined with national accounting data as a means to measure sectoral imbalances across the economy, demonstrating the exceptionally heavy urban bias guiding development strategies since the mid-1990s in the TAR, particularly in the early 2000s. Despite some attempts to compensate for these imbalances, intersectoral polarization continued unabated up to 2010 even despite the huge transition of labor out of agriculture, while new forms of inequalities appear to have rapidly emerged in the urban areas. In the third section, these aspects of polarization are analyzed through household income and wage data. The conclusion reflects on the significance of such polarization at such a key moment of transition in terms of its potentially powerful impacts on future trajectories of social stratification in the urban areas of the TAR.

196

Chapter 5

Labor transitions in the context of rapid subsidized growth According to the official aggregate employment data,3 the TAR labor force (mostly Tibetan) experienced one of the latest and, once started, fastest transitions out of agriculture from the late 1990s onward relative to the rest of China. This transition is shown in figure 5.1 with reference to the share of the labor force employed in the primary sector (mostly farming and herding) from 1990 to 2010. The primary employment share of the TAR was at 81 percent in 1990, then the most agrarian labor force in China. The share remained at 76 percent in 1999 (still the most agrarian of China) or 74 percent in 2000 (which is presumably a more accurate measure given that it would have been derived from the 2000 census). The share then started to fall sharply, to 65 percent in 2003 and 53 percent in 2010. The shift of labor out of the primary sector was proportionately similar in China and Sichuan, except that the drop was spread out over the twentyyear period, whereas it was mostly concentrated in the 2000s in the TAR. In China, the primary share fell from 60 percent in 1990 to a plateau of about 50 percent in the late 1990s and early 2000s, and then fell sharply from 2003 onward, to just below 37 percent by 2010. The share in Sichuan dropped from 73 percent in 1990 to 60 percent in 2000 and then to 43 percent by 2010. In contrast, the shift started later and more suddenly in the TAR as well as in Qinghai, the province with the next highest proportion of Tibetans in Figure 5.1: Primary sector employment share, 1990-2010

Sources: CSY (2011, Table 4-4) and equivalent tables in previous yearbooks (data for 2006 was not available in CSY 2007 and was taken from provincial statistical yearbooks for that year, although where an obvious discrepancy was apparent in the case of Gansu, an average was taken between 2005 and 2007 instead. Some data from the early 1990s was also taken from provincial yearbooks, although the discrepancies in these data were so severe in the case of Gansu before 1993 that they were not used).



The Great Transformation of Tibet? 197

its population. An equivalent drop in share of about twelve percentage points occurred in all the provinces shown from 2003 onward (besides Gansu). However, the overall pace of change in the TAR since 1999 has been exceptional. About 21 percent of the local (mostly Tibetan) TAR labor force moved out of agriculture between 2000 and 2010 according to these data, more or less converging with the norm of other poor (but more densely populated) provinces such as Gansu and even falling below the share in Yunnan (not shown). Notably, the pace of such transition in all these cases besides Gansu is very rapid from a comparative international perspective.4 The declining share in the TAR appears to represent a gradual stabilizing of the absolute numbers of Tibetans working in farming and herding despite ongoing population growth. The absolute number of people working in the primary sector in the TAR reached an earlier peak in 1999 at 922,000 people, after which the number fell to 850,000 in 2003, although it then gradually increased, reaching about 930,000 people by 2010. Some of the changes between 1999 and 2003 probably reflect adjustments to estimates after the 2000 census or else reclassifications and even actual resettlements in the beginning of the OWC (e.g., see Yeh and Henderson 2008). Nonetheless, the slow increase in this number since 2003—around 0.5 percent per year—is significantly less than the rural rate of natural population increase, which was well over 1 percent over these years, or an even faster rate of growth in the working-age population.5 This indicates that even in the context of falling fertility and substantial shifts to off-farm employment, population momentum nonetheless resulted in declining per capita landholdings for those continuing to work in agriculture, thereby exacerbating other problems such as stagnant grain prices, as discussed in chapter 2 and by Goldstein et al. (2003; 2008). Moreover, these absolute numbers are significant because they reflect the fact that the rapid transition in the local labor structure out of agriculture was happening regardless of the effect that non-Tibetan (i.e., Han Chinese) out-of-province migrants might have had on the overall employment shares of the TAR given that very few of these migrants come to the TAR to work in agriculture in rural areas (besides temporary migrants working as vegetable farmers in cities such as Lhasa or Shigatse, albeit most of them were probably recorded as urban labor in these statistics). It is plausible that actual trends are both under- and overestimated by these data. For instance, on one hand some of these trends might reflect administered changes in registration status that exaggerate actual socioeconomic changes, that is, people are reclassified as non-agricultural or as urban residents even though they might continue to farm or herd. Similarly, as noted in the introduction, some rural migrants might be registered as employed in secondary or tertiary activities even though they still spend part of the year working in farming or herding. On the other hand, much labor migration might be also hidden from these data, such as when farm-

198

Chapter 5

ers migrate to urban areas for six months a year in search of temporary work but otherwise remain registered as rural residents working in the primary sector. On balance, these data are probably accurate in a rough sense of reflecting real changes in socio-economic structure, as corroborated by the field insights by other scholars and me.6 To a large extent, the shift of labor out of agriculture in Tibetan areas implies urbanization, much more so than in other regions of China. This is because off-farm rural employment opportunities in Tibetan areas are much scarcer relative to more central and coastal areas of China, where much offfarm employment remains in rural areas. The recent (and heavily subsidized) surge in rural entrepreneurship and employment (as discussed by Childs et al. 2011) has attenuated this trend in the TAR to a certain degree. Nonetheless, despite the prevalence of entrepreneurial activities in the three villages surveyed by Childs et al., labor migration still remained the most prevalent emerging livelihood strategy for households in even the most “entrepreneurial” of these villages. In their similar research reported in Goldstein et al. (2008, 522), urban labor migration to Lhasa, Shigatse, or the local county seat accounted for about half of the overall labor migration in these three villages. Rural-rural labor migration, such as on infrastructure projects or housing construction, accounted for the other half of labor migration, albeit these three villages are located relatively close to a major city (Shigatse) and hence would have been relatively privileged in terms of off-farm rural employment generation. In this light, the predominant trend in the TAR overall has likely been toward a relatively rapid urbanization of the local TAR labor force. Two measures can be used to reflect these off-farm rural trends, as shown in the figures below. One is the rural share of employment (versus the primary share of total employment), and the other is the combination of three registered categories of rural employment, also as a share of total employment: township and village enterprises (TVEs), rural private enterprises, and rural self-employed individuals.7 In the first case, there is a difference—often even in trend—between the shares of total rural employment and primary sector employment. This difference could be taken as a very rough proxy for rural off-farm employment although, as discussed above, some of this difference might represent misclassifications of people who have migrated to urban areas but have maintained their registration status in the rural areas and hence are counted as part of the rural employed (or vice versa). The second measure (the three subcategories of registered rural employment) offers a more restrictive proxy measure of off-farm rural employment although, in this case, these three categories probably do not reflect informally organized off-farm rural employment in units not formally registered in administrative records, which might be well developed in richer coastal areas but much less so in remote and poor western areas, particularly in Tibetan areas. Moreover, some of the rural TVEs, private enterprises, and self-employed individuals might be involved in



The Great Transformation of Tibet? 199

agricultural activities, such as a TVE engaged in raising chickens for urban markets, as discussed in Goldstein et al. (2010), in which case, they would actually be part of the primary sector. In effect, these three categories in the TAR are only a fraction of the first measure (the difference between rural and primary employment), perhaps reflecting the fact that much of off-farm employment in the TAR is fairly informal and not formally registered.8 In contrast, in China on average, the combination of these three categories was greater than the difference between rural and primary sector employment, perhaps indicating that some of the people employed in these categories were working in the primary sector (or else that rural employment was over-reported or agricultural employment under-reported). Regardless, the broad observation holds that there is relatively much less off-farm rural employment in the TAR than elsewhere in China and that shifts out of agriculture tend to lead to urbanization much more so than elsewhere in China, as would be expected of a sparsely populated region with “primate” towns and cities.9 These various measures are shown in the next three figures for the period from 1993 to 2010.10 The share of total labor that is classified as rural is shown in figure 5.2. The difference between rural and primary sector employment shares is shown in figure 5.3. The combined share of TVE, rural private enterprise, and rural self-employed individuals of total provincial employment is show in figure 5.4. Comparing figure 5.2 with figure 5.1 on primary labor shares, it is apparent that a much stronger shift out of rural employment took place in the TAR than in other western provinces, implying that the transition out of Figure 5.2: Share of the total employment classified as rural, 1993-2010

Sources: calculated from CSY (2011, Table 4-2) and equivalent tables in previous yearbooks (data for 2006 was not available in CSY 2007 and due to lack of equivalent data in the provincial yearbooks for that year, the average was taken between 2005 and 2007 for 2006 in the provincial cases. Also note that the data for Sichuan up to 1996 includes Chongqing.)

200

Chapter 5

agriculture has involved much faster urbanization of the local labor force than elsewhere in western China. This was not the case throughout the 1990s, when the TAR labor force, like other western provinces, was actually becoming slightly more rural up to 1999, in contrast to the declining rurality of the labor force in China overall. The rising share in these western cases is probably explained by demographic factors (faster population increase in rural areas than in urban areas) combined with more restricted rural-to-urban mobility, which was slower to take off in the west than in the east (e.g., see Iredale et al. 2001; 2003). In the case of the TAR, the high rural employment share was sustained by the high primary sector employment share, given that the rural cannot decline below the level of the primary and a decline in the rural is usually stimulated by a decline in the primary. Indeed, in 1993 almost all rural employment in the TAR (at 81 percent of the labor force) was in the primary sector (at 78.5 percent of the labor force), whereas by 2000 rural employment was at 82 percent of the labor force while primary sector employment had declined to 74 percent. However, after the late 1990s, the pattern of the TAR diverged from the other three western provinces shown here, which continued to experience a rising rural share for several years (up to the mid-2000s in the case of Gansu). The rural share of the TAR started to fall from 82 percent in 2000 to 70 percent in 2010. This was over half of the almost 21 percent drop in the primary employment share over these same years. Notably, this corroborates with the above-mentioned survey results of Goldstein et al. (2008, 522), in which about half of the respondents who were “going for income” (i.e., labor migration) were doing so by migrating to urban areas, whereas about half migrated to other rural areas. As a result, the TAR ended this period with a much smaller rural labor force than in Sichuan or Gansu, converging with Qinghai and in tandem with the national trend. In contrast, the rural share of employment in the other western provinces fell later and by much less. The rural share in Qinghai, the next most similar province to the TAR in terms of population and topography, fell by less than 5 percent between 2000 and 2010, albeit it started this period with a much lower rural employment share than most other western provinces, almost on par with the national average. The primary sector share in Qinghai fell almost 19 percent. If these data are accurate, three-quarters of the shift of labor out of the primary sector in Qinghai was absorbed by other types of rural employment. Similarly, there was only a 6 percent drop in the rural share of Sichuan despite the 17 percent drop in the primary share, resulting in a surprisingly rural province (at almost 80 percent of total employment in 2010) despite the sharp reduction in primary share to 43 percent, which was close to the national average and probably reflects strong rural off-farm employment generation over these years. Thus, while the Sichuan labor force was apparently less urbanized than that of the TAR, it was also



The Great Transformation of Tibet? 201

much less agrarian. In Gansu, the rural share fell by only 1.3 percent alongside a decline in the primary share of almost 9 percent. Nationally, trends between these two shares were broadly correspondent over this period, with the rural share falling 13.5 percent while the primary share fell 13.3 percent, implying that, by the end of the decade, almost all the labor that shifted out of agriculture shifted out of the rural areas altogether (in an aggregate sense). Among the western cases shown here, the TAR exceptionally shows the strongest shedding of primary sector employment outside of the rural areas altogether, coming closest to the national aggregate trend, albeit under very different economic conditions. Inversely, if the urban employment share can be taken as a rough proxy of the rate of urbanization,11 it also suggests that the TAR population— and Tibetans in the TAR more specifically (given that the rural areas are almost entirely Tibetan)—has been experiencing some of the most rapid urbanization over this period in comparison to other western provinces. The TAR also started from a low urbanization rate of almost 20 percent according to the 2000 census (including temporary migrants), or 15 percent for Tibetans only, as discussed in chapter 3. In other words, the relative scarcity of off-farm rural employment in the TAR (and other Tibetan areas) implies that movements out of agriculture require relatively greater movements to towns and cities, and that urban labor markets are relatively much more central to labor transitions in the Tibetan areas than in other parts of western China. This is reflected further in figures 5.3 and 5.4, which are shown together for comparison of these two proxy measures of off-farm rural employment. As discussed above, the difference between rural and primary sector employment shares (figure 5.3) can be considered as a generous measure for off-farm rural employment given that it might include significant amounts of urbanizing migrants who have maintained their rural registration status (and possibly even agricultural employment status) and an informal status in urban areas. Nonetheless, a substantial increase in the difference between these two shares was registered in the TAR in the early years of the OWC, rising from 6 percent of total TAR employment in 1998 to 14 percent in 2003, and thereafter stabilizing at around 16 percent. The OWC thereby appears to have generated a substantial increase in the share of non-agricultural employment in the rural areas, similar to other western provinces but at a consistently lower level, as would be expected of a more sparsely populated remote area. This would be the result of efforts to raise rural incomes through the provision of rural employment opportunities in the TAR by means of intensive subsidization, particularly since 2003, as discussed by Goldstein et al. (2008; 2010) and Childs et al. (2011).12 The convergence with the national trend is also interesting, insofar as other western provinces, particularly Sichuan, diverged from the national trend in terms of a widening gap

202

Chapter 5

Figure 5.3: Percentage difference between rural and primary shares, 1993-2010

Sources: calculated from CSY (2011, tables 4-2 and 4-4) and equivalent tables in previous yearbooks for provincial data. The same adjustments as specified in figure 5.2 apply to this figure.

between rural and primary shares. This presumably reflects a strong focus on rural off-farm employment in these other western provinces, perhaps as a way to stem out-migration. In contrast, the gap actually narrowed after 1996 at the national aggregate level, reflecting the declining importance of TVEs and the increasing prominence of urbanization in national employment patterns. As noted above, the fact that the TAR converged with this national trend, in terms of percentage difference between rural and primary shares, masks a fundamentally different economic setting. Figure 5.4 provides a more restrictive proxy measure of off-farm rural employment. According to this measure, off-farm rural employment generation in the TAR has been much sparser than in the other provinces shown, at least in terms of officially registered employment in TVEs, rural private enterprises, or as rural self-employed individuals. The contribution of these three categories accounted for 5 percent of total TAR employment in 2008, up from 3.7 percent in 1999, versus 16 percent in Qinghai (up from 12 percent), 19 percent in Gansu (up from 18 percent), 23 percent in Sichuan (up from 17 percent), and 26 percent nationally (up from 25 percent).13 The data on TVE employment in the TAR are lacking for 2009 and 2010 (and before 1999), although based on the previous figure and the data on the other two categories that are available, we can presume that the slightly upward trend continued. According to this measure, there is a much sharper difference between the TAR and the rest, revealing a much greater scarcity of (formally registered) employment opportunities in the rural areas of the TAR. Considering the upward and downward biases of these two proxy measures of off-farm rural employment, it is likely that actual experience in the



The Great Transformation of Tibet? 203

Figure 5.4: TVE, rural private enterprise and self-employed individual employment as a share of total employment, 1993-2010

Sources: calculated from CSY (2011, Table 4-2) and equivalent tables in previous yearbooks for provincial data. The provincial data for 2006 was averaged between 2005 and 2007, as in figures 5.2 and 5.3. The gaps in certain years, including for the TAR in 2009 and 2010, are because the data was not complete in those years (typically, the TVE employment is missing)

TAR lies somewhere between the two. A fairly substantial increase in off-farm rural employment probably occurred in largely informally organized activities, although this increase was significantly scarcer than elsewhere in China (as would be expected given the population and geographic characteristics of this region). This resulted in relatively stronger urbanization for an equivalent degree of employment transitioning out of the primary sector. Secondary Sector Employment Transition out of agriculture and, in large part, into urban areas resulted in an equally rapid transition toward tertiary employment in the TAR, largely bypassing employment in the secondary sector (especially manufacturing). Figure 5.5 below presents the changing trends in the share of secondary sector employment as a proportion of total employment for the five cases discussed, along with some highlighted data on the resulting composition of secondary employment in 2010 for Qinghai, Sichuan, and the TAR in the text boxes embedded in the figure (subsector data is not available for Gansu or China). Figure 5.6 presents the same for tertiary sector employment. The share of secondary employment in the TAR was significantly lower than in all other cases, as was historically the case (cf. Fischer 2005a) and would be expected of a sparsely populated and remote agrarian region. Nonetheless, there was a notable increase in share following the beginning of the OWC, particularly between 2002 and 2003, when the share rose

204

Chapter 5

Figure 5.5: Secondary sector employment share, 1990-2010

Sources: calculated from CSY (2011, Tables 4-4) and equivalent in previous yearbooks (secondary share for Gansu in 2006 was averaged between 2005 and 2007, as in figures 5.2 and 5.3); subsector data from QSY (2011, Table 5-3), TSY (2011, Table 3-4), SSY (2011, Table 4-4). Note that sectoral employment data is not available in the CSYs past 2002 or in the GSYs. Also note that the subsector data for Qinghai does not precisely add up to the total sectoral employment as reported in other sources, although the discrepancy is marginal, hence these data offer a rough sense of the structure of employment within the secondary sector of this province. The fourth secondary subsector (production and distribution of electricity, gas and water) is not included but is marginal in all cases, amounting to a share of a few percent.

from 6.2 percent to 9.1 percent of total employment in the TAR. This corresponds with the beginning of major railway construction in the TAR and related OWC projects. The increase was sustained and rose further to more than 10 percent in 2007 and 2008, even after the completion of the railway construction in 2006. This corresponds with the boom in rural construction activity generated by the Comfortable Housing Project (CHP) under the Eleventh Five-Year Plan, which started in 2006 (see Goldstein et al. 2010). The entirety of the increase in share was due to sharp increases in the numbers of people registered as employed in construction or in mining and quarrying. The number of people employed in manufacturing in the TAR apparently actually declined from about 27,000 people in 2000 (CSY 2001, table 5-5) to 25,000 in 2010 (and hence from 2.2 percent to 1.4 percent of total provincial employment), versus an increase in construction from 35,000 to 119,000 people and an increase in mining and quarrying from 4,000 to 35,000 people. If these data are accurate (there is a possibility that they are relatively accurate with respect to Tibetan permanent residents, although less so with respect to migrant workers), then the reduction in manufacturing employment would imply that the increase in village-based artisanal manufacturing activities described by Goldstein et al. (2010) was compensated for by a reduction of manufacturing employment elsewhere



The Great Transformation of Tibet? 205

in the economy, such as in urban areas, as described in chapter 8 for the Tibetan areas of Qinghai. Another possibility is that activities such as brick making, one of the activities described by Goldstein et al. (2010) and Childs et al. (2011), was recategorized in the 2009 and 2010 data as quarrying given that the standard method of brick making in these areas involves topsoil removal (and hence amounts to natural resource depletion). This is a definite possibility because between 2008 and 2009, as reported in TSY (2009) and TSY (2010), the sharp reduction in reported manufacturing employment from 47,695 to 24,210 people was compensated for almost precisely by a sharp increase in mining and quarrying employment from 4,785 to 29,788 people. Together, the combined total of manufacturing and mining and quarrying employment only rose moderately from 2.5 percent of provincial employment in 2000 (about 31,000 people) to 3.5 percent in 2010 (about 60,000 people), thereby accounting for just under one-fifth of the increase in secondary employment over this decade, much of which was stimulated by construction activity (as in the case of brick making). Construction accounted for 63 percent of secondary employment in the TAR by 2010 (up from 49 percent in 2000), versus 13 percent in manufacturing, 19 percent in mining and quarrying, and 5 percent in the production and distribution of electricity, gas, and water. In contrast, secondary employment was led by manufacturing in most other provinces, even despite the apparent construction booms over this period. For instance, 46 percent of secondary employment in Sichuan was in manufacturing in 2010, versus 43 percent in construction and 9 percent in mining and quarrying. Even in Qinghai, with its high levels of subsidization and investment, manufacturing accounted for 39 percent of secondary employment, versus 51 percent in construction and 6.6 percent in mining and quarrying. The structure of secondary employment in Qinghai was more similar to the TAR, although with a much higher and surging share of secondary employment in overall employment, accounting for almost 23 percent of total employment in the province in 2010. As a result, manufacturing accounted for 8 percent of total employment, rising from 170,000 people in 2000 to 260,000 people in 2010. Notably, the decline in the secondary employment share in Gansu, Sichuan, and Qinghai from the mid-1990s to the early 2000s and the subsequent strong increases in Sichuan and Qinghai reflect the restructuring of the antiquated industrial structure inherited from the Third Front era in these two provinces and in particular the shedding of labor from state-owned enterprises, which happened later than in coastal China. In terms of urban-rural distribution, secondary employment in the TAR was more rural than in Qinghai, the only other province for which these specific data are available. In the TAR, 74 percent of the construction employment was registered as rural in 2010, as was 84 percent of mining and quarrying employment. Manufacturing employment was reported as 50 percent

206

Chapter 5

rural in 2008.14 Again, this could represent the relatively large amount of activity that was generated by the CHP, from construction to a related range of relatively small-scale processing activities such as brick making. These data reflect efforts by the government to stimulate off-farm employment in rural areas, although we do not know the degree to which out-of-province (Han Chinese) migrants are included in these data—particularly in urban construction and even in some rural construction activities (such as the railway versus the CHP). Also, once rural employment is deducted from overall secondary employment, the sheer paucity of urban secondary employment is striking, despite the construction boom over these years. Again, this might be reflective of the fact that much of the urban construction activity employed out-of-province temporary migrants, who might not be recorded by these data sources, especially if they were employed informally. Tertiary Sector Employment Despite these signs of increasing secondary employment in the rural areas of the TAR, such employment nonetheless remained much more limited than elsewhere in China, and the increase in the secondary employment share by 5.3 percentage points from 2000 to 2010 only accounted for about a quarter of the decline in the primary share by 20.7 percentage points over the same period. The bulk of the declining primary share (about three-quarters) was absorbed by the tertiary sector, which rose from a share of around 20 percent of total employment in 2000 to almost 36 percent in 2010. The tertiary share rose so rapidly in the TAR over this period that it surpassed the naFigure 5.6: Tertiary sector employment share, 1990-2010

Sources: same as figure 5.5.



The Great Transformation of Tibet? 207

tional share in 2008 and was on par with that of Qinghai. Despite quite divergent patterns in the 1990s, all western provinces and the national average had converged at very similar (and rising) tertiary shares from 2006 onward. However, the composition of such tertiary employment reveals several significant differences between the TAR on one hand and Qinghai and Sichuan on the other—the three provinces for which disaggregated data is available in the provincial yearbooks. The fact that the shares of tertiary employment to total employment were more or less the same by 2010 in these three cases implies that the comparison of tertiary compositions is proportionate in terms of shares of total employment. The differences are predictable given the analysis of the previous chapter. First, the category of “public management and social organizations” accounted for a much larger share of tertiary sector employment in the TAR, at 14 percent in 2010 versus 8 percent in Qinghai and 6 percent in Sichuan. Second, despite all the hype about tourism in the TAR, hotels and catering apparently generated proportionately much less employment in the TAR than in Qinghai and Sichuan. In comparison to Qinghai, hotel and catering employment was also much less rural, which is not surprising given the overbearing restrictions that the TAR government places on foreigners and even domestic Chinese with regard to traveling outside of certain select destinations such as Lhasa and Shigatse, even in the most relaxed of times. This insight suggests that the tourism sector in the TAR is very concentrated and has relatively weak employment generation potential. For instance, total employment in hotels and catering in the TAR in 2010, when tourist numbers apparently surpassed 6.8 million people, only amounted to 60,481 people, almost three-quarters based in urban private enterprises. Moreover, these categories of employment in the TAR—public management, trade and hotel/catering—tend to be dominated by migrant (particularly Han Chinese) workers, who are probably recorded in the public sector employment data (e.g., a temporary resident cadre) but much less so in the private sector data (such as in the case of an informally employed caterer). Conversely, the employment shares in education and health, social security, and welfare were slightly lower in the TAR than in the other two cases, which corroborates with the analysis of education infrastructure in the previous chapter, that is, that the TAR is not actually privileged in terms of actual educational inputs (and certainly not in terms of outcomes). Finally, the TAR data included an undefined category not included in the others provinces— namely, “other”—which amounted to 18 percent of tertiary employment in 2010 and was mostly rural. Presumably, this was counting forms of rural public employment that are not typically included in the more formal categories of staff and workers, such as village leaders, village health care workers, and village teachers. Indeed, some of the discrepancies in the data of the other two provinces are probably due to similar accounting gaps.

208

Chapter 5

Intersectoral Polarization The rapid increase in the tertiary employment share in the TAR over the 2000s is a predictable outcome of the rapid growth of the tertiary sector in the TAR economy, which came to account for almost 54 percent of GDP in 2010, up from 45 percent in 1999. Thus, the rapid labor transition has, to some extent, balanced the imbalance in the late 1990s and early 2000s between a very large tertiary GDP share and a much smaller tertiary employment share. Nonetheless, this balancing within the tertiary sector has been accompanied—remarkably considering the speed of labor transition—by continuing intersectoral polarization between the primary and secondary/tertiary sectors of the TAR given the very imbalanced nature of growth focused on construction and tertiary services within a largely agrarian workforce (as of 2000). “Intersectoral polarization” refers to a divergence in labor productivity (unit of output per unit of labor) across different sectors, with the strong qualification that, as discussed in the introduction, “productivity” in this sense is proxied (and can only be proxied in aggregate terms) by monetary GDP value-added data, which include wages and prices, and thus constitutes a poor meter of actual output, effort, or effectiveness.15 The monetary value-added data are nonetheless useful to examine because they reflect the monetary remunerative and profit potentials of various sectors of an economy (recalling that the income approach to value-added GDP accounting is decomposed into compensation or remuneration to employees/workers, net taxes on production, depreciation of fixed assets, and operating surplus). In other words, employees and employers cannot earn more than the total value of their output, however produced and however valued. Intersectoral polarization need not occur if labor moves proportionately into more rapidly growing sectors, thereby equalizing out value-added productivities across the economy, as happened with labor transfers out of agriculture in Europe. However, this had not (yet) happened in the TAR as of 2010. It appears to have started to happen in China overall during the 2000s. Nationally, intersectoral imbalances were led by industry (especially manufacturing) while the construction and tertiary sectors played compensating or equalizing roles, in the sense that expansions of employment in construction and services occurred at close to average levels of value-added per employed person in the economy (i.e., GDP per employed person). In contrast, polarization in the TAR was led by construction and tertiary services, while industry was closer to the economy norm than was typical in other provinces. This difference is important because an imbalance would tend to focus profit-seeking investment and other economic activities into sectors with higher value-added potential, whether the imbalance is created by state



The Great Transformation of Tibet? 209

policy, market prices, or other economic processes. The implication is that value-added incentives in China create a bias toward investment and employment in industry (especially manufacturing), whereas in the TAR they create a bias toward investment and employment in construction or tertiary services (especially in the state sector). Moreover, the imbalance in the TAR increased over the decade despite the enormous expansion in construction and tertiary activities, and even though such expansions would tend to eventually lead to diminishing returns under most normal circumstances, especially at the scale and speed of expansion observed in the TAR where construction came to account for almost one-quarter of GDP by 2010. This outcome in the TAR (both the expansion and that fact that it occurred with an increasing imbalance) was entirely guided by state policy and was heavily subsidized, as discussed in the previous chapter. It is meaningless to refer to it as a “market” outcome, even though discourses of market incentives and discipline might have been used in certain cases to justify certain elements such as Chinese migration to the TAR, as discussed by Yeh (2013). The setting in the TAR was compounded by the fact that nominal valueadded per employed person in the inflated construction and tertiary sectors was significantly higher than the norms prevailing in western and central China (albeit an underreporting of migrant labor in these sectors, particularly in construction, might exaggerate the disparity). In other words, the value-added per employed person in construction in the TAR was several times the norm in China, reaching 104,728 yuan per employed person in 2010, versus 38,490 yuan in Qinghai and 24,168 yuan in Sichuan. Similarly, the value-added per employed person in the tertiary sector in the TAR reached 79,577 yuan in 2010, versus 45,147 yuan in Qinghai, 35,433 yuan in Sichuan, and 65,732 yuan in China on average. These interregional imbalances set the incentives for attracting out-of-province entrepreneurs and migrants into these sectors given their high profit and/or remunerative potentials, particularly in construction, where the imbalance is exceptional. In the case of intersectoral imbalances within a province, such dynamics can be represented by relative GDP/labor ratios. The term “relative GDP/labor ratio” is used to indicate the value-added per employed person in a sector relative to the provincial economy average (i.e., GDP/total employment, which is usually used as a proxy for the “productivity” of an economy, as noted above). A ratio of one implies that a unit-share of labor in that sector (e.g., 1 percent of total employment) contributes a unit-share of valueadded to GDP (e.g., 1 percent of GDP). More than one means that a unitshare of labor contributes more than its share of value-added (and thus is more “productive” in the sense of creating more monetary value or being more valued). Less than one means the opposite. The ratio of these ratios (e.g., the tertiary or secondary GDP/labor ratio over the primary GDP/labor ratio) can be taken as a measure of the relative value-added per employed

210

Chapter 5

person in one sector versus another (e.g., the tertiary or secondary sector vis-à-vis the primary sector) as opposed to the first ratio, which measures the contribution of each sector relative to the average in the economy as a whole. These measures show the relative profit and/or remunerative potentials offered by each sector in the local economy, thereby offering structural insights into the local labor transitions. They are presented in table 5.1 together with the nominal measures mentioned above. Table 5.1: Relative and nominal GDP/labor ratios, 2000 and 2010 TAR GDP/Labor ratios Primary

2000 0.42 2010 0.25 Secondary 2000 3.99 2010 2.92 … Industry 2000 2.89 2010 1.94 … Construction 2000 5.15 2010 3.58 Tertiary 2000 2.25 2010 1.51 Ratio of GDP/Labor ratios Secondary/Primary 2000 9.53 2010 11.45 … Industry/Primary 2000 6.88 2010 7.74 … Constr./Primary 2000 12.25 2010 14.33 Tertiary/Primary 2000 5.37 2010 5.93 Nominal value per employed person (rmb) Primary 2000 3,991 2010 7,438 Secondary 2000 37,792 2010 71,167 ... Industry 2000 27,378 2010 56,839 … Construction 2000 48,800 2010 104,728 Tertiary 2000 21,401 79,577 2010

Qinghai

Gansu

Sichuan

China

0.24 0.24 3.23 2.44 3.58 4.63 2.63 0.94 1.64 0.98

0.33 0.28 3.24 3.19 3.70 n/a 2.36 n/a 1.34 1.10

0.40 0.34 2.92 2.19 4.11 3.05 1.27 0.66 1.31 1.03

0.30 0.28 2.04 1.63 3.22 n/a 1.12 n/a 1.42 1.25

13.45 10.25 14.93 19.30 10.97 3.92 6.83 4.13

9.84 11.20 11.22 n/a 7.15 n/a 4.08 3.87

7.39 6.49 10.27 8.98 3.17 1.97 3.32 3.06

6.77 5.92 10.74 n/a 3.74 n/a 4.71 4.53

2,652 10,934 35,737 112,143 39,680 189,398 29,087 38,490 18,088 45,147

2,738 8,192 26,937 91,769 30,808 n/a 19,694 n/a 11,197 31,739

3,576 11,591 26,492 75,175 37,109 110,022 11,520 24,168 11,857 35,433

4,146 14,512 28,088 85,881 44,860 n/a 15,547 n/a 19,530 65,732

Sources: calculated from the same sources as figures 5.5 and 5.6 and the GDP figures in chapter 4.



The Great Transformation of Tibet? 211

According to the first measure, at the beginning of the rapid labor transition in the TAR in 2000, 20 percent of the TAR labor force was employed in the tertiary sector, accounting for 46 percent of the GDP of the TAR, and resulting in a relative GDP/labor ratio of 2.3. By 2010, 36 percent of the labor force was employed in the tertiary sector, accounting for 54 percent of GDP and resulting in a ratio of 1.5. Similarly, the ratio in the secondary sector fell from 4.0 in 2000 to 2.9 in 2010. The reduction in these two ratios indicates a balancing between the GDP and labor shares of the tertiary and secondary sectors and equalization between these sectors and the average of the economy over these years. Out-of-province non-Tibetan migrants probably accounted for a much larger share of tertiary and secondary sector employment and value-added in 2010 than in 2000 due to significant net in-migration to urban areas over this period and the fact that Han Chinese migrants have tended to increasingly dominate the most lucrative sectors of urban construction and tertiary activities, in partnership with a small strata of Tibetan elites. However, we do not have access to data that would allow for a proper evaluation of this likely scenario. In contrast, 74 percent of the TAR labor force was employed in the primary sector in 2000, accounting for 31 percent of GDP and resulting in a relative GDP/labor ratio of 0.42. The primary sector ratio then fell to 0.25 by 2010 (53 percent of employment accounting for 14 percent of economic activity). The fall in this ratio indicates marginalization of the primary sector from the norm of the economy even despite growth in this sector and the rapid transition of labor out of this sector. In other words, even more transfer of labor out of the primary sector or else faster economic growth of the primary sector would have been required to match the speed of growth and structural change in the rest of the TAR economy. In nominal terms, the value-added per employed person in the primary sector in the TAR grew much less between 2000 and 2010 than in the other provinces shown in table 5.1, from 3,991 yuan to 7,438 yuan, versus from 2,652 yuan to 10,934 yuan in Qinghai, and 3,576 yuan to 11,591 yuan in Sichuan. The faster increase in the other provinces was due to faster growth in the value-added of the primary sector, not due to a faster transition of labor out of this sector. An even more nuanced differentiation of these trends in the TAR can be gained by combining the subsector data presented in the two previous figures with the equivalent subsector GDP data where available. For instance, whereas the overall GDP/labor ratio for the secondary sector of the TAR in 2000 as 4.0, the GDP/labor ratio for the subsector of industry was 2.9 (“industry” is a combination of mining, manufacturing, and production and distribution of electricity, gas, and water in the GDP data), whereas it was 5.1 for construction. Both subsector ratios fell by 2010, to 1.9 for industry and 3.6 for construction (versus 2.9 for the secondary sector overall). These falling ratios also imply equalization within these sectors, although the ra-

212

Chapter 5

tio for construction in 2010 was nonetheless particularly high considering that it came to account for one-quarter of GDP, for only about 6.8 percent of the workforce (according to official statistics, again, probably not including much migrant contract labor). Temporal comparisons are trickier with the tertiary data given changing tertiary GDP categories over the 2000s. Nonetheless, it is notable that the GDP/labor ratios of trade and hotels and catering in the TAR in 2010 were less than one (0.65 and 0.89 respectively), implying that average valueadded per employed person in these two subsectors of the tertiary sector were closer to the norm of the primary sector than to other parts of the tertiary sector or construction. This probably reflects a strong degree of saturation and competition in these subsectors, which accounted for 37 percent of total tertiary sector employment, or about 13 percent of overall provincial employment, much of which might have been employed (or self-employed) informally. The ratio for the category of transportation, storage, post, and telecommunications was 1.3, also lower than the overall tertiary sector, although higher than the economy average. Again, the fact that migrant labor was probably underreported in these three subsectors further emphasizes the fact that these activities were not particularly lucrative in the TAR, at least not in comparison to the subsidized affluence in other parts of the tertiary sector or in construction. Given that these three subsectors accounted for 47 percent of tertiary employment in 2010, the GDP/labor ratio for the remaining 53 percent of tertiary sector employment would have been significantly higher than the overall tertiary ratio of 1.5 in order to compensate, reflecting a significant polarization within the tertiary sector itself between the more “vernacular” parts and the more formal, state-employed parts. Specific GDP data is no longer published for the more lucrative public sector components of the tertiary sector. However, the tertiary GDP category of “other” in the TAR (which would have been dominated by public management and education) had a GDP/labor ratio of 1.8 in 2010. The subcategory of “government agencies, party agencies, and social organizations” had a ratio of 2.6 in 2000 (slightly higher than the overall tertiary sector), when it employed 57,301 people, or 4.6 percent of the TAR labor force for about 12 percent of GDP. It is likely that this ratio was sustained throughout the decade given that the subsector (renamed as “public management and social organizations”) employed 88,263 people in 2010, or 5.1 percent of total employment for probably a similar, if not slightly higher, share of GDP (as analyzed in the previous chapter). If so, public management would have been one of the only categories that did not become more balanced over this period. This would not be surprising given that value-added in this sector is mostly determined by public sector wages, which became among the highest in the country over these years, as analyzed in the next section.



The Great Transformation of Tibet? 213

Despite the balancing that was happening within the overall secondary and tertiary sectors in the TAR (relative to the economy average GDP/labor ratio), the intersectoral ratios between these two sectors and the primary sector continued to increase due to the marginalization of the primary sector. This ratio of the ratios is shown in the second set of rows in table 5.1. For instance, the tertiary/primary ratio in the TAR rose from 5.4 in 2000 to 5.9 in 2010, meaning that the average employed person in the tertiary sector in 2010 accounted for 5.9 times more value-added in the economy (measured in money terms) than the average employed person in the primary sector. Similarly, the secondary/primary ratio rose from 9.5 to 11.4, and the more specific construction/primary ratio rose from 12.3 to 14.3. These increasing intersectoral ratios give an indication of the extent of intersectoral imbalance and polarization occurring in the TAR economy since 2000—despite growth in all sectors—and the degree to which such polarization has served as an underlying economic driver of rapid labor transitions. Similar albeit less pronounced intersectoral imbalances exist across China. However, as noted in the beginning of this section, imbalances in most cases have been led by the secondary sector, particularly industry, whereas the GDP/labor ratio of the tertiary sector has typically been much closer to the norm of the overall economy, if not below in some cases (as with trade and hotels and catering in the TAR). Moreover, with some exceptions, intersectoral imbalances were moderately attenuated over the 2000s, opposite to the situation in the TAR. In other words, there was a degree of intersectoral depolarization in China as a whole. For instance, the GDP/ labor ratio of the primary sector in China fell from 0.30 in 2000 to 0.28 in 2010, versus a fall in the secondary ratio from 2.04 to 1.63 and in the tertiary ratio from 1.42 to 1.25. As a result, the secondary/primary ratio fell from 6.8 to 5.9 and the tertiary/primary ratio fell from 4.7 to 4.5. Similar trends were observed in Sichuan. There was a rise in the secondary/primary ratio of Gansu, similar in magnitude to that of the TAR, although this was compensated for by a falling tertiary/primary ratio. Qinghai stands out as one of the most remarkable cases of industry-led imbalances, with strong compensation in both construction and tertiary sectors, in contrast to the case of the TAR. Over the 2000s, the primary GDP/labor ratio in Qinghai remained at 0.24, while the ratio of industry rose from 3.58 to 4.63, the ratio of construction fell from 2.63 to 0.94, and the tertiary ratio fell from 1.64 to 0.98. As a result, the industry/primary ratio increased from 14.9 to 19.3, while the construction/primary ratio fell from 11.0 to 3.9 and the tertiary/primary ratio fell from 6.8 to 4.1. These were the strongest increases and decreases of these ratios observed across all the provinces shown in table 5.1. The strong increase in industry probably reflects the restructuring and upgrading toward higher value-added and more capital-intensive forms of industry, such as high-grade aluminum

214

Chapter 5

and steel production, and more capital-intensive forms of mining, aided by the completion of one of the largest hydroelectric dams in China in the late 1990s.16 The capital intensity of these industries would have limited their potential for employment generation (and also would have required higher skilled labor) even while generating higher value-added. As a result, the nominal value-added per employed person in industry in Qinghai reached 189,398 yuan in 2010, versus 110,022 yuan in Sichuan, and 56,839 yuan in the TAR. In contrast, construction and tertiary activities in Qinghai had a strong equalizing rather than polarizing intersectoral influence over this decade, in striking opposition to the trend in the TAR despite a similar intensity of subsidization. The contrast between the TAR and these other provinces is reflective of the particular nature of unbalanced rapid growth in the TAR since the mid-1990s, driven by extremely intense subsidization concentrated in construction and public management and resulting in unabated intersectoral polarization within these sectors despite the rapid shift of local labor out of farming and herding. In particular, the polarizing influence of both the tertiary sector in the TAR, which accounted for over half of GDP since 2002, and construction, which came to account for one-quarter of GDP by 2010, is especially pertinent given that these two sectors have also been the main targets of interprovincial migration to the TAR. Indeed, the nominal GDP value of construction in the TAR, which amounted to 104,728 yuan per employed person in 2010, was far greater than the equivalent measure for all sectors in Sichuan besides industry. Given that these measures represent both the remunerative and profit potential of such activities, the attraction for both workers and entrepreneurs from other parts of China is obvious, especially from Sichuan, the largest single source of interprovincial migrants to the TAR. However, construction and urban service activities such as trade and catering are also particularly important as sources of employment for local (Tibetan) labor transitions out of farming and herding given that these subsectors are, in principle, the easiest and most accessible for rural migrants to enter in terms of the skills required and the formalized institutional barriers to overcome. In this respect, the sectors where rural Tibetans appear to have made significant inroads, as studied by Goldstein et al. (2010) and Childs et al. (2011), have been in low value-added construction, artisanal manufacturing, and local transport and trade. These sectors are much closer to the norms of the primary sector or the economy average—in terms of their GDP contribution per employed person—than the bulk of construction concentrated in high-value (and highly subsidized) ventures or else the higher value-added half of tertiary sector employment concentrated in the public sector. The equalizing roles of the tertiary sector in China and other provinces, and of construction in Qinghai, are notable in this respect,



The Great Transformation of Tibet? 215

whereas the polarizing role of the same two sectors in the TAR—and the polarization between subcomponents within these two sectors—is suggestive of increasing rather than lessening difficulties for local Tibetans to access and enter the employment opportunities generated by rapid economic growth within these two sectors, as discussed in the next chapter.

Household Income and Wage Inequality Whether or not intersectoral (or intra-sectoral) polarization results in increasing inequality between households is difficult to judge without more detailed panel data on income distribution given that a rural household (or a family that shares resources across multiple households) might include a farmer, a construction worker, a trader, and even a public employee among its members. However, tabulated panel data on rural household income distribution in the TAR were last made available in TSY (2000), for data up to 1999, making it impossible to evaluate both income poverty rates and income inequality in the rural areas of the TAR in the 2000s on the basis of publicly available data. We can take for granted that “absolute” income poverty rates (meaning insufficient income to meet minimum food and non-food basic needs) have been falling in the rural areas of the TAR given rapidly increasing real per capita rural household incomes since the early 2000s (see figure 5.7). Real per capita rural incomes in the TAR had fallen (relatively) to the lowest in China by 1997 and had stagnated for about a decade up to 2002, as discussed in chapter 2. They then rose rapidly after 2002 to a mid-western position, most likely due to a variety of rural development initiatives to increase rural incomes from 2003 onward, such as those discussed by Goldstein et al. (2008; 2010) and Childs et al. (2011). It would have taken very large and sustained increases in inequality in order to counteract the absolute poverty reduction implied by the speed of this increase in average incomes.17 While substantial increases in rural household inequality in rural TAR appear to have taken place since the late 1990s, as reported to me by Melvyn Goldstein based on his work in western Tibet (see chapter 7), these increases do not appear to have been large enough to have canceled out the poverty reduction of increasing average incomes (and wealth), particularly once government welfare transfer payments targeted to the poorest households are taken into consideration. Hence, it is likely that there has been substantial reduction in absolute poverty since the late 1990s, when, according to official income distribution data and poverty lines, about one-third of the rural populations of both the TAR and Qinghai were living below or around the national absolute poverty line.18 Nonetheless, absolute rural poverty or intra-rural inequality measures are not the most pertinent for examining marginalization within the regional

Sources: calculated from CSY (2011, tables 9-5 and 10-21) and equivalent in previous yearbooks.

Figure 5.7: Real per capita rural household net income, 1990-2010



The Great Transformation of Tibet? 217

economy. Urban-rural and intra-urban household inequality measures are more pertinent given that these offer much better indications of the challenges faced by rural Tibetans to integrate into the economies of local towns and cities, where the most lucrative economic opportunities have been concentrated. For this purpose, some tabulated data are available in various provincial yearbooks on urban income distribution in the TAR and other provinces, although household income surveys in China are generally sampled only from populations registered as permanent residents, thereby excluding both local and interprovincial migrants. This is not a problem for the representativity of rural surveys given that few migrants end up in rural areas, but it is a huge problem for the representativity of urban surveys. Therefore, several roundabout ways must be used to gauge inequality dynamics in China, especially in a context of urbanization. Three measures can be used to decipher these dynamics from the patchwork of publicly available data. One is urban-rural household income inequality, which typically accounts for a large proportion of overall interhousehold income inequality in China.19 This can be easily measured as a ratio of average per capita urban over rural household incomes. Again, these data have the disadvantage that they exclude temporary residents, which is not a concern for rural areas but a major concern for urban areas. The two other measures rely on urban household income survey data. One is based on inequality measured by grouped data, such as ratios of richest and poorest deciles and quintiles. The second compares urban household income data with urban wage data as a means to gauge urban inequalities in a way that takes advantage of deficiencies in the household income survey data and thereby reflects indirectly at least some categories of migrant labor. Two trends in the TAR can be observed in the combination of these measures. One is a sharp polarization in urban-rural inequality from the mid-1990s up to 2001, followed by equalization in urban-rural inequality after 2001, albeit converging with the norm of rising urban-rural inequality in other western provinces. The second trend is a sharp increase in various measures of urban income and wage inequality from the early 2000s onward. The first trend appears to be related to the sudden onset of labor transition out of agriculture around 2000, which would have offset rising urban-rural inequality. The second trend suggests that the subsequent decline in urban-rural inequality represents not an end of polarization, but rather a transfer of urban-rural inequality to the urban areas through urbanization. Rapid urbanization would tend to balance out urban-rural inequality, if only by reducing the absolute number of people living and working in the rural areas and increasing the number of relatively poor people living and working in urban areas. As a result, intra-urban inequality quickly emerged in the 2000s as the dominant new schism driving polarization and stratification in the TAR.

218

Chapter 5

Urban-Rural Household Income Inequality The trends in urban-rural inequality are shown in figure 5.8, which measures the ratio of per capita urban disposable household income (of households registered as permanently residing) over per capita rural household net income, both deflated by their respective urban and rural consumer price indices. Urban-rural inequality in the TAR reached the dizzying height of 5.5 in 2001, that is, the average urban per capita household income was 5.5 times higher than the average rural per capita household income—a level never before seen at the provincial level in the PRC.20 Urban-rural inequality rose in all western provinces from 1997 or 1998 onward, more or less following the national pattern multiplied by a factor to account for the heightened disparity in the west.21 In other words, the success of various western development strategies from the mid-1990s onward in reversing the trend of worsening interregional inequalities was achieved alongside a worsening of this dimension of inequality within provinces. However, such polarization was excessive in the TAR and was notably out of synch with the other provinces given that urban-rural inequality in the TAR started to increase sharply after 1995, precisely at the same time as it was (briefly) falling nationally and in every other western province due to a combination of pro-rural poverty reduction strategies and agricultural recovery from 1994 to 1997. This reflects the fact that the takeoff of the TAR in the mid-1990s was primarily urban and excessively de-linked from the stagnant local rural economy. Urban-rural inequality in the TAR was just as sharply rectified from 2001 to 2006, at least back down to the level observed in the TAR in the mid1990s and converging with the norm of increasing urban-rural inequality in most other western provinces up to around 2006 (besides in Sichuan and Xinjiang, where urban-rural inequality did not increase after 2002 and even fell below the national level). According to these data, urban-rural inequality was falling in all western provinces from about 2007 or 2008 onwards, the TAR included. The sharp correction in the TAR from 2001 to 2006 partly reflects strong growth in per capita rural incomes after 2002, as noted above. It also partly reflects the fact that real per capita urban incomes in the TAR (officially) stagnated in 2004 and 2005, and even fell sharply in 2006 (see figure 5.9). This was possibly due to a respite during these three years in the otherwise rapidly increasing money wages of urban state-sector staff and workers in the TAR (see figure 5.11), which in turn accounted for a large part of the dynamics observed in average urban incomes of the TAR. It was also possibly due to the retrenchment of locals registered as urban permanent residents (who are largely Tibetan) from state-sector employment in the TAR over these years, thereby lowering the weight of these privileged workers in the calculation of household incomes. Both likelihoods are discussed further below.

Sources: calculated from CSY (2011, tables 9-5, 10-15 and 10-21) and equivalent in previous yearbooks.

Figure 5.8: Urban-rural household income inequality, 1990-2010

Sources: calculated from CSY (2011, Tables 9-5 and 10-15) and equivalent in previous yearbooks.

Figure 5.9: Real per capita urban household disposable income, 1990-2010



The Great Transformation of Tibet? 221

These relative dynamics in urban household incomes mark a radical departure from previous norms. Average urban household incomes in the TAR were previously well above the national average and the highest in western China up until the early 2000s due to the high salaries and wages of state-sector staff and workers, who accounted for almost 73 percent of urban employment in 2000 (not including estimates for informal employment). From the early 1980s up until the early 2000s, around 70 percent of these state-sector staff and workers were also local Tibetans, as often noted by the government (e.g., PRC 2001b) and by scholars such as Sautman and Eng (2001) and Hu and Salazar (2008). However, average urban incomes in the TAR converged with the national average in 2002 and then fell below it in 2004. The sharp drop in 2006 (if accurate) brought these average urban household incomes in the TAR to the lowest in China. They quickly reverted to a mid-western position in 2007, albeit well below the national average. The fact that the average wages of state-sector staff and workers remained among the highest in China throughout these years, neck and neck with those of Beijing and Shanghai up to 2007 (see figure 5.11), suggests that the weight of these wages declined sharply in the composition of urban (permanent resident) household incomes, that is, the share of local (largely Tibetan) urban permanent residents accessing these wages fell sharply. Evidence of this point is discussed further below. Intra-urban Inequality It is difficult to derive an exact measurement of intra-urban inequality in the TAR given the prevalence of migrant workers in the urban workforce (both Tibetan and non-Tibetan), who are not usually included in any of the household surveys (surveys only include permanent residents). Intra-urban inequality in the TAR among those registered as permanently residing was already high in the late 1990s in comparison to the rest of China, which can be explained by the fact that over 70 percent of the registered urban workforce was employed as relatively privileged state-sector staff and workers and benefited from some of the highest salaries in the country, while the less privileged 30 percent experienced some of the highest urban poverty rates in China.22 From this baseline in the late 1990s, there are strong indications that intraurban inequality further increased more sharply in the TAR than elsewhere in China. This suggests that the wages of those without access to state-sector employment increasingly lagged behind the sharp wage increases of those with state-sector employment and/or that those with state-sector employment constituted a shrinking share of urban employment. In either case, urban wealth became increasingly concentrated among those with state-sector employment. Essentially, the development pattern established by the OWC appears to have reinforced an already polarized urban economy in the TAR,

222

Chapter 5

confirming the suggestion that the recent decline in urban-rural inequality is not so much an end of polarization, but rather a reconstitution of such polarization in the urban areas through urbanization. Two measures can be used to decipher these trends. The first is the standard measure of inequality as a ratio of the average per capita incomes of the richest and poorest shares of the urban permanent-resident population, based on the grouped data from household income surveys that are generally provided in provincial statistical yearbooks (including the TSY). However, as noted above, this measure would exclude migrants (both those from local rural areas and from other provinces). The second measure is one that I previously innovated (Fischer 2007) as an attempt to overcome this deficiency in the household survey data by comparing average wages of urban staff and workers (or simply those “employed in urban units” from 2006 onward) to urban household incomes. Average Incomes of the Richest and Poorest Urban Resident Population Shares The first measure of urban inequality can be calculated as the ratio of the average per capita incomes of the richest and poorest shares of the urban population, in this case including only households registered as permanent resident given that the data are taken from household surveys, which generally do not include migrant households.23 Here, four ratios are calculated, providing different reflections of inequality: the average income of the top two quintiles over the bottom two quintiles (40/40), which provides for a broad indication of the extent of polarization between top and bottom halves of the population; the top quintile over the bottom two quintiles (20/40), which gives a more restricted picture of the richer share; the top decile over the bottom two quintiles (10/40), which indicates the extent to which extreme wealth at the top is driving the previous ratios; and the top decile over bottom decile (10/10), which gives the most extreme measure of wealth inequality between richest and poorest urban permanent residents. In the case of China, these measures aggregate a huge regional range of urban households, from very rich urban centers such as Shanghai to very poor county towns in western China. Those from individual provinces aggregate a more narrow range, from provincial capitals down to county towns. These ratios reveal increasing urban inequality across all measures between 2000 and 2010. The TAR does not stand out as any particular exception. Qinghai is the most notable, moving from the least to the most unequal of these five cases, particularly within the measure of richest and poorest deciles (10/10). Although data is lacking for the narrower measures of Sichuan in 2010, the broader measures seem to indicate no increase in urban inequality, even though the province was the most broadly unequal of the five cases in 2000.



The Great Transformation of Tibet? 223

The TAR had a notable discrepancy in 2000 whereby its broader three ratios were within the same range as those of China and Sichuan, whereas its narrowest ratio—the ratio of richest and poorest deciles—was much higher at 7.9 (meaning that the average income of the richest 10 percent of permanent resident households was 7.9 times higher than the average income of the poorest 10 percent). This would seem to confirm the previous observation that the urban setting in the TAR around 2000 was one of the most polarized in China in the sense that relatively high average urban incomes were paralleled by one of the highest rates of urban poverty in China, most likely concentrated among the bottom 30 percent of the urban employed not working in state-owned units. Moreover, despite significant increases in the broader ratios of the TAR— similar to China as a whole—the narrow ratio (10/10) barely increased in the TAR by 2010, reaching 8.1 in contrast to the sharp jumps in the 20/40 and 10/40 ratios. Again, this is possibly reflective of the situation discussed above with respect to the stagnation of urban incomes in the mid-2000s. The retrenchment of urban permanent residents from state-sector employment in the TAR over these years, in deference to non-locals not necessarily registered as permanent residents and thus not included in the household surveys, might have tempered the upward surge in the average incomes of the richest decile of urban permanent residents relative to what the situation might have been if this subgroup of permanent residents had actually been the richest of all residents, as elsewhere in China. Indeed, the problem with these income measures taken from household surveys is that they do not include migrants and temporary residents. Hence, increases or reductions in inequality could reflect changes in relative Table 5.2: Ratios of average per capita urban household disposable income of the top and bottom shares of the urban permanent resident population ordered by income, 2000 and 2010 TAR 2000 40/40 20/40 10/40 10/10 2010 40/40 20/40 10/40 10/10

Qinghai

Gansu

Sichuan

China

2.7 3.0 3.3 7.9

2.3 2.7 3.1 4.7

n/a n/a n/a n/a

2.8 3.4 4.0 6.4

2.4 2.9 3.4 5.0

3.3 4.0 4.5 8.1

3.8 4.8 5.7 11.3

3.0 3.6 4.3 6.7

2.8 3.4 n/a n/a

3.2 4.1 5.1 8.6

Sources: calculated from CSY (2011, table 10-7) and equivalent in CSY (2001) and in various provincial yearbooks for provincial data.

224

Chapter 5

wages or other forms of remuneration (given that most urban household incomes are from wages), or else they could also reflect shifting compositions of employment toward migrants. In other words, if there were a shift in privileged high-wage employment toward migrants and away from even privileged locals, and if such a compositional shift largely impacted the richest decile or quintile of (permanent resident) households, rising wage inequality would actually disappear from these income data and could even appear as a stabilization of the narrowest measure of inequality (10/10) among permanent households. It would certainly result in an underestimation of overall inequality among all urban households. Urban Wage Data and Employment Composition A way around this problem would be to examine wage data, which appear to include migrants, at least those working formally in registered units of employment. The only wage data available in the statistical sources is with reference to “staff and workers” (up to 2008) and to an apparently slightly wider although still formalized category of urban employment from 2006 onward, referred to as employment in “urban units.” In this regard, “staff and workers” refers to a relatively privileged sub-category of urban employment in China, including persons working (permanently or on contract) in state-owned and what might be called “corporate” units, that is, units of state ownership, collective ownership, joint ownership, shareholding ownership, and foreign ownership (including Hong Kong, Macao, and Taiwan).24 The new reporting category (“employed persons in urban units”) started in CSY (2007) and fully replaced the “staff and worker” category from CSY (2010) onward, presumably as part of the gradual reform of statistical reporting systems in China and in conformity with changing employment norms. This new category appears to have a slightly wider definition of employment than “staff and workers,” although the precise difference is not clear from the explanatory notes of the CSYs.25 The new category of “urban units” still does not include (formally registered) urban private enterprises or urban self-employment.26 The actual difference between the two measures of employment was also negligible in CSY (2007), when the two data series overlapped: the number of employed persons in urban units was about 5.5 million greater than staff and worker employment for China overall in 2006 (out of total employment in urban units of just over 117 million), and the difference was only about 19 thousand in the TAR (out of total employment in urban units of 189 thousand). Similarly, the difference in average wages between the two categories in the three years that they overlapped (from 2006 to 2008) was negligible. The average wages of employed persons in urban units was slightly lower than those of staff and workers in these years, suggesting a wider inclusion of some workers



The Great Transformation of Tibet? 225

employed in formal urban units but not previously categorized as staff and workers and who were probably at the lower end of the wage scale within the respective urban units (such as part-time and/or contract workers). The difference is greater in the TAR than in the other provinces (see figure 5.10 below), although it is not clear why.27 In any case, the two overlapping series of data are presented in figure 5.10. In sum, data on average wages reported in the China Statistical Yearbooks refer to staff and workers (up to 2008) and then to employed persons in urban units (from 2009 onward).28 In principle, these data would cover many of the privileged temporary migrants working in the state sector of the TAR and other Tibetan areas given that the employment data are not defined according to household registration status (except in the case of the registered urban self-employed, who are expected to hold urban residency certificates or to have resided in urban areas for a long time). There are no publicly available data for average wage rates other than for these categories of formal employment, except in the case of the average wages of employed persons in urban private units (those that are formally registered). Wages in urban private units began to be reported in CSY (2010, table 4-17) for data starting in 2009, although these data were not reported for the TAR. In other words, no wage data have been available in any of these official statistical sources for the poorer informalized categories of urban employment in China, such as construction workers not working under contract, or even for formally registered private units in the TAR. It is useful to frame the discussion further in terms of urban employment composition in 2010, so that the context to which these data refer is clear. Table 5.3 shows a broad disaggregation of urban employment in terms of units of registration in order to highlight several points regarding the relative weight of urban units within total registered urban employment. “Registered” in this sense refers to employment in either urban units or in registered urban private units and registered urban self-employment. The percentage of total registered urban employment employed in urban units is shown in the third row of the table. The fourth and fifth rows show the proportion of total registered urban employment in state-owned units and non-state-owned units. The second-to-last row shows the proportion of registered employment in private enterprises or as self-employed individuals. The last row shows the residual, or the difference between the sum of all these categories of registered employment and total urban employment, as reported nationally in CSY (2011, table 4-2) or in various provincial yearbooks, calculated as a proportion of total urban employment.29 This residual can be taken as a rough estimate of informal employment, per Chinese regulatory standards. For instance, in China overall, 55 percent of total registered urban employment in 2010 was based in urban units, and in turn was almost exactly half in state-owned units and half in non-state-owned units. The other 45

226

Chapter 5

percent of registered employment was employed in private enterprises or was self-employed. The residual between the registered categories of employment and total employment amounted to 32 percent of total employment, meaning that an estimated 32 percent of the almost 347 million people employed in urban China in 2010, or about 111 million people, had no formal employment registration status and were “informal” in this specific sense. Notably, the scale of informality, as according to this residual measure, had not fallen significantly over the 2000s.30 The estimate of informal employment at the provincial level relies on the total urban employment data provided in various provincial yearbooks, which reveal a residual over and above the numbers of registered employed reported in CSY (2011, table 4-2). This is not possible in the case of the TAR because TSY (2011) reports only registered categories of employment. This might mean that the statistical authorities in the TAR have not bothered to estimate (or to report their estimate of) informal employment in the TAR, perhaps because of the seasonal nature of much of this employment, or else it might mean that the heightened security environment requires that all employed people in the region register, at the very least as self-employed. Keeping this point in mind, most employment in urban units in the TAR in 2010 was in state-owned units, at 93 percent of total urban unit employment. In contrast, much more of urban unit employment was in corporate non-state units in other provinces, up to half in China overall. Moreover, a much lower proportion of registered urban employment in the TAR was in urban units than in the three other western provinces shown and the national average. Conversely, a much higher proportion of registered urban employment was employed in private enterprises or was self-employed than in the four other cases shown. In other words, circa 2010, almost 40 percent of registered urban employment in the TAR was in state-owned units (which would include many migrants), whereas almost 60 percent Table 5.3: Composition of urban employment, 2010

urban N (1000 people) registered uN (1000 people) % registered in units ... % reg. in state-owned units ... % reg. in non-s-o units % reg. in private + self-emp. % residual/informal (if any)

TAR

Qinghai

Gansu

Sichuan

China

n/a 531 41.8% 39.0% 2.8% 58.2% n/a

1,280 958 54.9% 39.6% 15.3% 45.1% 25.2%

4,336 3,179 61.1% 46.4% 14.7% 38.9% 26.7%

13,819 10,295 55.4% 32.6% 22.8% 44.6% 10.8%

346,870 235,900 55.3% 27.6% 27.7% 44.7% 32.0%

Sources: calculated from CSY (2011, tables 4-2 and 4-6) and equivalent from provincial yearbooks for estimates of total provincial employment (first row), used for calculating the residual (last row).



The Great Transformation of Tibet? 227

was in either private enterprises or was self-employed. Other formal nonstate unit employment in the TAR provided for very few options of employment, less than 3 percent of registered employment in 2010. This missing middle is significant because, like the state-sector, such non-state “corporate” enterprise units typically provide higher-wage and more secure employment options, including various social security provisions that private enterprises are generally exempt from providing. Preferential employment policies could also be more easily mandated in these units than in purely private enterprises. In light of the predominance of private enterprise and self-employment, it would be very important to have data on average wages in urban private enterprises in the TAR, as provided in CSY (2011, table 4-17), but unfortunately, these are not reported for the TAR. The share of state-owned unit employment as a proportion of total registered urban employment also fell dramatically over the decade in the TAR, from 73 percent in 2000 to 56 percent in 2004 and then to 39 percent in 2010. The fall in share of state-owned unit employment was less dramatic in Qinghai, from 63 percent in 2000 to 44 percent in 2004 and then 40 percent in 2010 (or at least most of the reduction occurred in the early 2000s). In Gansu the share fell from 68 percent in 2000 to 60 percent in 2004, and then to 46 percent in 2010. In Sichuan it fell from 59 percent, to 44 percent, and then 33 percent respectively. Nationally, the share fell from 55 percent in 2000 to 40 percent in 2004 and then to 28 percent in 2010.31 The declining national share also implied an absolute decline in the number of people employed in state-owned units, from 81 million people in 2000 to 65 million in 2010. If we compare the share of staff and worker employment to employment in urban units (assuming that the two are roughly the same), similar patterns of collapsing share are observed in staff and worker employment as a share of total registered urban employment over the 2000s. The share of staff and workers (or employment in urban units) to total urban registered employment in China as a whole fell from 77 percent in 2000 to 67 percent in 2004 and then 55 percent by 2010. This falling share was due completely to the drop in the numbers employed in state-owned and collective-owned units, which was compensated for by increases in the other non-stateowned “corporate” categories of staff and worker employment. Similar trends were reported in Gansu, where the share fell from 81 percent to 71 percent and then 61 percent, and in Sichuan, where it fell from 80 percent to 66 percent and then 55 percent. Whereas the drop in state-owned unit employment in China, Gansu, and Sichuan was partly compensated for by increased staff and worker employment in other “urban unit” categories, staff and worker employment in the TAR was merely a reflection of state-owned unit employment. In other words, the share of staff and worker employment (or employment in urban

228

Chapter 5

units) in the TAR fell from 72 percent of total registered urban employment in 2000 to 50 percent in 2004, and then to 42 percent by 2010, closely matching the shares of state-owned unit employment discussed above.32 In contrast, the share in Qinghai also fell from 70 percent in 2000 to 52 percent in 2004, but then it subsequently rose to 55 percent by 2010, reflecting an expansion of staff and worker employment in non-state-owned units that more than compensated for the decline in state-owned unit share. The category of public management (i.e., government administration) also accounted for a much larger share of this urban unit employment in the TAR than in other provinces and was the largest category of urban unit employment in the TAR overall. Public management (and social organizations, i.e., para-state agencies) in the TAR accounted for almost 40 percent of urban unit employment in 2010, versus 17 percent in Qinghai, 19 percent in Gansu, 14 percent in Sichuan, and 11 percent nationally. In other words, about 17 percent of formally registered urban employment (including private and self-employment) in the TAR in 2010 was in public management, versus about 9 percent in Qinghai, 12 percent in Gansu, 8 percent in Sichuan, and 6 percent nationally. In contrast, urban unit employment in education (also largely state sector) in the TAR in 2010 was less than half that in public management (i.e., 39,553 people versus 88,263 people), whereas employment in education was greater than in public management in all the other cases besides Qinghai, where it was slightly smaller. From this perspective, the TAR exhibited a significantly lower share of urban unit (or staff and worker) employment in total registered urban employment by 2010 than in any of these other cases despite the almost complete dominance of state-owned units in urban unit employment and the fact that the state sector was by far the most important driver of economic growth in the TAR. Or, at a time when the state sector in the TAR was growing at a very rapid pace in GDP terms, to a much greater extent than in other western provinces, state-sector employment was relatively more austere than in these other provinces and was certainly not expanding at a pace commensurate with its role in GDP growth. Yet even despite this relative austerity, the TAR government was employing a much larger share of urban employment in public management than in other provinces, meaning that other categories of state-sector employment (such as in education) were even more austere than the aggregate measures indicate. Urban Wage Inequality With this background in mind, average wage data are compared to per capita urban household incomes in figure 5.10 in order to capture a sense of the disparity between the privileged upper strata of urban employment (i.e., the 42 percent of formally registered employed people who were working in ur-



The Great Transformation of Tibet? 229

ban units in the TAR in 2010) versus the average levels of affluence of urban households registered as permanent resident. Per household survey definitions, the urban household income data are based on samples of households registered as permanently residing (i.e., the regular surveys do not include temporary migrants). According to these surveys, urban household disposable incomes in China are derived primarily from salaries and wages (almost entirely in the TAR).33 Many of the permanent urban households would earn these salaries and wages—or income more generally—from outside the formal “urban units,” such as from registered private enterprise employment, registered self-employment (together representing 58 percent of formally registered urban employment in the TAR), or from informal employment. Indeed, there is no data available on the extent to which employment in urban units is composed of permanent residents versus temporary residents, although given various considerations discussed below, we can presume that Tibetan permanent residents are underrepresented in the urban units. This might or might not mean that the proportion of permanent household employment that is based in urban units is proportionate with the overall share of urban unit employment in total urban employment (i.e., 42 percent in 2010) because temporary residents might equally dominate private and self-employment (and informal employment). These unknowns notwithstanding, the per capita urban household disposable income data reflect an average of all forms of remuneration earned by all urban residents registered as permanent residents (who were about three-quarters Tibetan in the TAR, according to the 2000 census). Therefore, comparing these incomes with the average wages of staff and workers (or of employed people in urban units) can give an indirect indication of wage inequality between the privileged upper strata of urban employees (including some temporary residents) and the average of all employed urban permanent residents (as reflected by average household incomes, which includes unemployed members and dependents). Temporary residents not employed in urban units are unfortunately not included in this calculation, although we can assume that their wage conditions are similar to the permanent urban households not employed in urban units. This assumption seems reasonable given that the GDP/labor ratios in trade and hotels and catering were below the economy norm, as discussed in the previous section, and these are the sectors where migrant employment in private enterprises and migrant self-employment tend to be predominant in the TAR. In this comparison, it is to be expected that average money wages will be marginally higher than per capita urban household incomes even in a relatively egalitarian setting given that per capita incomes include both working and dependent household members. Accordingly, rising inequality would be indicated by a rising ratio. Figure 5.10 shows this ratio of average wages to average income as an innovated indirect measure of wage inequality.

230

Chapter 5

Figure 5.10 reveals a sharp polarization of urban wage inequality in the TAR from 2000 to 2007. The ratio based on the average wages of staff and workers more than doubled from 1.9 in 1999 to 4.1 in 2007, a level far above the next most unequal province, Qinghai (according to this measure). The ratio then fell slightly to 3.8 in 2008. The ratio measured according to wages of people employed in urban units stabilized at 3.3 in 2010, in contrast to 2.6 in Qinghai, 2.2 in Gansu, 2.1 in Sichuan, and 1.9 for China as a whole. In light of the dynamics of urban-rural inequality and urbanization discussed earlier, these findings suggest that there was a sharp upward adjustment of intra-urban inequality in the TAR over the 2000s, precisely at the same time as urban-rural inequality was being tempered and rapid labor transitions out of agriculture and toward urban areas were taking place. There are two principal reasons that could explain the sharp rise in this measure of urban inequality. A rising wage/income ratio could represent the fact that the wages of staff and workers (or people employed in urban units) are rising relative to the average of all urban wages. Or, it could represent a falling share of staff and worker (or urban unit) employment in total urban employment among households registered as permanently residing, thereby reducing the weight of staff and worker (or urban unit) wages in the calculation of average urban incomes. Both cases appear to apply to the TAR. First, average wages of staff and workers in the TAR, which were always above the national average due to “hardship” considerations,34 suddenly rose even further, as shown in figure 5.11. Such wages rose from 1.5 times the national average wage of staff and workers in 1999, when the TAR Figure 5.10: Urban wage inequality, 1998-2010

Sources: calculated from CSY (2011, tables 4-13 and 10-15) and equivalent in previous yearbooks (e.g. CSY 2009, tables 4-17, 4-23 and 9-15).

Sources: calculated from CSY (2011, tables 4-13) and equivalent in previous yearbooks (e.g. CSY 2009, tables 4-17, 4-23).

Figure 5.11: Average wages of staff and workers, 1998-2010

232

Chapter 5

average wage was 12,962 yuan, to two times the national average in 2002, after almost doubling to 24,766 yuan.35 Notably, they were the highest in China in 2002 and in 2004. While the central government was generally taking an approach at that time of rapidly increasing money wages as a means to stimulate consumption in China, there are varied opinions as to why the already-privileged “hardship” wages in the TAR were increased so much faster than the national norm at the beginning of the OWC. Some argue that this was meant to garner the loyalty of local Tibetan cadres and the so-called emerging Tibetan middle class. Others argue that it was to make the TAR more attractive for Chinese staff and workers considering a working sojourn in the region, particularly given the increased demand for skilled (and Chinese fluent) labor in various OWC projects. Both considerations probably motivated these wage policies in the TAR. Average staff and worker wages in the TAR were subsequently quite haphazard and abrupt, unlike the more gradually increasing wages elsewhere in China. They stagnated in 2005 and 2006, then increased very sharply in 2007, again to equality with average wages in Beijing, and then stagnated again in 2008 at 47,280 yuan (versus over 56,000 yuan for both Beijing and Shanghai and 25,038 yuan for Sichuan). According to the data on wages of people employed in urban units, the slowdown in 2008 continued in 2009 and 2010, and the TAR returned to its relative starting point, with average wages at about 1.4 times the national average and lagging behind the average wages of Beijing and Shanghai after 2007. In nominal terms, average wages of urban unit employees in the TAR reached 49,898 yuan in 2010, versus 36,121 yuan in Qinghai, 29,096 yuan in Gansu, 32,567 yuan in Sichuan, 36,539 yuan in China, 65,158 yuan in Beijing, and 66,115 yuan in Shanghai. These relative differences were more or less the same in the more specific subcategory of state-owned urban units, given that average wages in state-owned units were marginally higher than overall average wages in all cases. Much of the income dynamics observed in the TAR are closely associated with the dynamics in these wages. For instance, the sharp decline in urbanrural inequality in 2005 and 2006 (figure 5.8) is in part related to urban incomes, which slowed in 2005 and then decreased in 2006 (figure 5.9), which in turn was related to the stagnation in staff and worker average wages in 2005 and 2006. Similarly, the decline in intra-urban wage inequality in the TAR in 2005, the subsequent peak in 2007, and then the decline after 2007 (per figure 5.10), is closely related to the same trends in urban wages. In other words, the dynamics of inequality in the TAR have been almost entirely administered by staff and worker (or urban unit employee) wage policy, which in the TAR is almost entirely located in the state sector. Second, these sharp wage increases took place parallel to a sharp reduction in the share of staff and workers (or urban unit employees) in total



The Great Transformation of Tibet? 233

registered urban employment, from 72 percent in 2000 to 42 percent in 2010, as analyzed above. This was a larger share reduction than in the other provinces discussed. It was also the lowest resulting share across all provinces in China overall in 2010, despite the fact that the state sector had been the main locus of economic growth over the decade and that very few formal employment opportunities were available outside the state sector besides in the more informalized forms of private or self-employment (or else in fully informal unregistered employment). The high wages earned by this 42 percent of registered urban employees in the TAR thereby resulted in a high degree of wage and income concentration. For instance, the equivalent wage levels in Beijing were earned by almost 67 percent of formally registered urban employment in 2010 (i.e., 67 percent of such employment in Beijing was in urban units), and by over 53 percent of formally registered urban employment in Shanghai. In other western provinces, lower wages (more in line with urban income norms) were also earned by a larger share of urban employment than in the TAR. In the TAR, some of the highest wages in China were captured by the smallest share of employment earning such wages in China, which explains the heightened wage inequality of the TAR in figure 5.10, even in 2010 when wages started to lag behind those of Beijing or Shanghai. There is no access to data that would allow for an assessment of the degree to which the shrinking share of urban unit employment has been disproportionately borne by permanent urban residents or by (local) Tibetans more generally (although this data would be available, in principle, if the government cared to divulge it). However, data do exist at the beginning of the decade that definitely show that the upward revaluation of staff and worker wages in the TAR between 2000 and 2003 (relative to the national average) took place simultaneously with a sharp reduction in the number and share of Tibetan staff and workers in state-owned units between 2000 and 2003, while the number and share of non-Tibetans rose. Calculating from TSY (2004, table 4-5), the share of Tibetans in total staff and worker employment in state-owned units fell from 71 to 65 percent, while that of non-Tibetans rose from 29 to 35 percent. At the cadre level, which accounted for two-thirds of permanent state-sector employment in 2003, the change was sharper: overall cadre employment increased from 69,927 cadres in 2000 to 88,734 in 2003, while the number of Tibetan cadres fell from 50,039 to 44,069, or from 72 percent of total cadre employment to just below 50 percent. It was only at the lowest and most subordinate levels of state-sector employment, that is, non-permanent workers on contract, where Tibetan representation increased, from 71 percent in 2000 to 82 percent in 2003, although the total number employed at this level fell dramatically, from 23,453 workers in 2000 to 6,912 workers in 2003, probably in line with government efforts to rationalize irregular forms of employment

234

Chapter 5

in state-owned units over these years. However, such rationalization fell disproportionately on Tibetans, resulting in a reduction of 11,028 Tibetan contract workers over these three years, versus a reduction of 5,513 nonTibetan contract workers. This was in addition to a reduction of 19,619 Tibetan permanent workers versus a reduction of only 1,937 non-Tibetan permanent workers, and a reduction of 5,970 Tibetan permanent cadres versus an increase of 24,777 non-Tibetan cadres. The general policy of reducing employment in the state sector was undoubtedly a national directive given that urban employment in stateowned units in China overall dropped from 81 million in 2000 to just under 69 million in 2003 (and to 65 million by 2010). However, as these data (perhaps inadvertently) reveal, Tibetans in the TAR bore a grossly disproportionate share of the burden, especially considering that their overall share of the TAR population was in excess of 90 percent as of the 2000 census, even if the underestimated head count of Han Chinese migrants is accounted for, as discussed in chapter 3. Tibetans also accounted for about three-quarters of permanent urban residents. The shift in 2003 revealed a sudden move away from Tibetan representation in urban state-sector employment, that is, from the most privileged and formalized forms of employment in the TAR, and non-Tibetan cadres outnumbered Tibetan cadres for the first time since 1980. Government assertions that Tibetans were the dominant beneficiaries of increasing state-sector wages, thereby contributing to an emerging “middle class” of Tibetans,36 became much more tenuous by 2003, after which the government stopped publishing this particular disaggregation of employment data for the TAR.37 Rather, Tibetan employment was shrinking during these early years of the OWC in precisely the parts of the economy that were growing fastest, that is, the urban state sector. Up to that point, Tibetans working as staff and workers would have mostly been permanent urban residents, implying that the burden of such employment reduction was more specifically borne by these relatively privileged urban residents. As a consequence, the sharp wage increases were increasingly and disproportionately captured by non-Tibetans and by a shrinking share of permanently registered urban households, which helps to explain the growing divergence between average wages of staff and workers and urban per capita household incomes up to 2003. Unfortunately, we have no idea of these trends since 2003 because this particular minzu disaggregation of the staff and worker data was discontinued after TSY (2004) and replaced by a sex disaggregation. However, in contrast to the situation of continuously falling numbers of state-sector staff and workers in China overall up to 2010, staff and worker employment in state-owned units in the TAR underwent a substantial increase, from 136,646 people in 2003 to 170,748 in 2008. By 2010 there were 207,267 employees in state-owned urban units (out of a total of 222,116 employees



The Great Transformation of Tibet? 235

in all urban units). We do not have any information of the degree to which the increase was filled by local Tibetans versus non-locals, that is, (mostly non-Tibetan) out-of-province migrants. We do know that the state-owned unit share of urban unit employment did not decline, unlike elsewhere in China where reductions in state-sector employment were partially compensated for by increasing employment in non-state urban units, as discussed above. Rather, the state-owned share of total staff and worker employment in the TAR actually rose from 92.2 percent in 2000 to 94.5 percent in 2008,38 and by 2010, state-owned urban unit employment accounted for 93.3 percent of all urban unit employment in the TAR. Given these trends and the fact that the sharp urban wage polarization from 2000 to 2002, as observed in figure 5.10, was resumed after a three-year respite from 2005 to 2007 suggests that similar exclusionary dynamics were at work later in the decade. In other words, if an increasing proportion of (Tibetan) urban permanent residents had been employed in the state sector and had thereby benefited from the high wages in that sector, we would expect that there would have been less lagging of average urban incomes behind average staff and worker wages. If similar exclusionary dynamics were at work later in the decade, they would definitely provide much insight into the outburst of protests that took place in Lhasa and elsewhere in March 2008. Moreover, as analyzed previously, per capita urban household disposable incomes in the TAR—which had been consistently above the national average throughout the reform period—fell below the national average for the first time in 2004 and even stagnated in current value in 2005 and 2006 (i.e., they declined in real value, after accounting for inflation). Thereafter, average urban incomes in the TAR joined the ranks of other poor western provinces such as Gansu and Sichuan. This lagging was in stark divergence from the dynamics in the average wages of staff and workers. The divergence implies either a compositional effect that continued after 2003 (i.e., a shrinking share of urban permanent households employed in state-sector employment, as discussed above), or else that the incomes of the urban permanent households without state-sector employment were increasingly lagging behind, if not falling in real terms, thereby canceling out the sharp increases in average money wages of staff and workers in the calculation of average incomes. This could have been the case if, for instance, laid-off staff and workers experienced long bouts of unemployment or were forced to find employment at lower wages in private enterprises or through selfemployment. Obviously, those Tibetans who did manage to retain statesector employment did very well (in income terms). Outside the state-sector, the whole array of so-called spontaneous migrants (as they are referred to in the scholarship on China, i.e., migration not organized by the state) are not included in the household surveys and most of them would not be included in the staff and worker (or urban

236

Chapter 5

unit employment) data. As discussed above, it is not even clear if migrants working informally (i.e., not formally registered) are included in the TAR urban employment data, as seems to be the case in other provinces. Based on my own qualitative field insights and some secondary sources such as Hu (2003) or the survey work by Ma and Lhundup (2008) on temporary migrants in Lhasa in 2005, these migrants include Han Chinese, Chinese Muslims, or even Tibetans from other parts of Tibet, who largely come on their own initiative to ply their trades independently in the urban areas, such as traders, travel agents, tour guides, construction workers, shoe menders, restaurant owners, cooks, tailors, rickshaw or taxi drivers, sex workers, or even beggars. Such migrants are mostly not competing for state-sector employment, other than in cases of state-owned construction companies hiring out-of-province migrants (albeit in these cases hiring is often arranged outside the province altogether). Other than in such state-sector construction work where wages (and entire project funding) are heavily subsidized, direct monetary incentives are not necessarily being offered by the state to these migrants.39 However, the general conditions of affluence induced by the high level of subsidization relative to the average conditions found in most other areas of western or central (or, in some cases, even coastal) China obviously create strong incentives for these migrants to tap into the urban economies of the TAR (and other Tibetan areas where similar conditions prevail). Potential gains are much higher than in the highly competitive conditions of the densely populated areas of inner China. For instance, when I was doing much of my fieldwork in 2004 and 2005, a construction worker in Gansu, Anhui, or Henan might have earned about 10 to 15 yuan a day whereas he or she could have easily earned 40 yuan a day on construction sites in Lhasa or even 100 yuan a day for certain jobs on the railway. Similar conditions motivated entrepreneurs, particularly given that price levels in urban TAR are generally higher than elsewhere in western and central China, providing significant opportunities for price arbitrage. Thus, spontaneous interregional migration has been encouraged by subsidization policies even if it has not necessarily been directly subsidized. This being said, the service sectors typically occupied by these spontaneous migrants (as well as by urbanizing rural Tibetans), such as trade and hotels and catering, actually exhibited lower-than-parity GDP/labor ratios, as discussed in the second section. The ratio for transportation, storage post, and telecommunications was slightly above parity, although lower than the average for the tertiary sector. This implies that wage and profit norms in these subsectors of the tertiary sector were closer to the norm of the primary sector than to construction or to other parts of the tertiary sector (such as government administration). Indeed, there were some signs of saturation in these subsectors during my fieldwork in the TAR, resulting in a



The Great Transformation of Tibet? 237

strong degree of competition among migrants, particularly given that much of the employment in these sectors would have been employed (or selfemployed) informally. This might offer an additional explanation for the high degree of urban wage inequality in the TAR in the event that saturation in these subsectors would have depressed the earnings of the urban permanent residents not working in (mostly state-owned) urban units. Again, the fact that migrant labor is probably underreported in these subsectors further emphasizes the fact that these subsectors were not particularly lucrative in the TAR, at least not in comparison to the subsidized affluence in other parts of the tertiary sector or in construction (although they were still relatively lucrative compared to these subsectors elsewhere in western and central China). Hence, Chinese migrants working in these subsectors might well have been working at norms similar to (or not much higher than) local urbanizing Tibetan migrants or urban permanent residents not working in urban units. As a result, the hypothetical inclusion of these non-local migrants in the above calculations of inequality might not change the results significantly, although this insight also indicates the degree to which such migrants were competing directly with local Tibetans for similar types of employment at similar wage norms. While it is difficult to deduce the impact of migrants on inequality given their exclusion from most of the data, more importantly, it is precisely the confluence of these different streams of migrants in the Tibetan urban areas, together with local urbanizing rural Tibetans and permanent resident urban Tibetans, that sets the playing field for intense competition over urban employment opportunities. Given that these opportunities are overwhelmingly determined by the centrally directed subsidization policies that have driven almost the entirety of rapid urban-centered economic growth in the TAR, they are characterized by strong linguistic, cultural, and political modes of bias deriving from the dominant Han Chinese group in control of most power and most financial flows from outside the province. In combination with much worse levels of schooling and education outcomes in these Tibetan areas, Tibetans competing for urban employment in the TAR and other Tibetan areas are faced with strong disadvantages vis-à-vis out-ofprovince migrants, as discussed in the next chapter.

Conclusion This chapter focused on rapid labor transitions in the context of rapid growth and economic polarization, with a particular focus on the TAR as the representative case for the Tibetan areas. Section 1 analyzed the rapid labor transitions that occurred alongside economic growth up to 2010, namely, a rapid structural shift out of agriculture. Part of this shift was absorbed by off-

238

Chapter 5

farm employment within rural areas, particularly in construction activities. However, about three-quarters of the shift in the TAR was absorbed by the tertiary sector, and a substantial share—perhaps more than half—transferred to urban areas. The speed of these transitions was so fast that the share of tertiary sector employment in the TAR, which was only 20 percent in 2000, surpassed the share of China in 2008 and reached almost 36 percent of total provincial employment in 2010 (the primary sector accounted for 53 percent of employment, down from 74 percent in 2000). The second section then examined aspects of intersectoral polarization in the TAR. Despite the rapid transfer of labor from the primary to the tertiary sectors, the valueadded per employed person continued to diverge between these two sectors, reflecting the intensity of focus on construction and urban services in recent subsidization strategies since the late 1990s, which came to account for 54 percent and 25 percent of GDP by 2010. These trends contributed a crucial pull factor for both local urbanization and inter-provincial migration. Moreover, polarization driven by construction and the tertiary sector in the TAR stood in contrast to elsewhere in China, where these sectors tended to compensate for the polarization led by manufacturing. The third section then analyzed various aspects of inequality emerging and subsiding in the course of these transitions. Urban-rural inequality increased sharply (against the national trend) from the mid-1990s until the early 2000s. The invigoration of a rural focus in development policy since the beginning of the OWC in the TAR and especially since 2006 under the Eleventh Five-Year Plan was partly responsible for reversing the trend of rising urban-rural household income inequality by providing a significant boost to rural off-farm employment in construction and small-scale production, although other factors were also located in the dynamics of urban incomes and intra-urban inequality. Namely, there was a sharp (and administered) increase in urban inequality in the TAR from the late 1990s up to 2007, the eve of the wave of protests that rocked the TAR and other Tibetan areas in China. The increase was far above the levels observed in other provinces of China according to a comparison of average staff and worker wages with average per capita urban household incomes. This suggests that intra-urban inequality had taken over from urban-rural inequality as the dominant locus of polarization in the TAR from the early 2000s onward. The sharp increase in urban inequality was not a reflection of rural-urban migration because rural migrants are mostly excluded from these urban statistics. Rather, it appears to have been related to a sharp reduction of Tibetan representation in state-sector employment and an increasing divergence between average state-sector wages and the average incomes of permanent urban residents in the TAR. As such, rising urban inequality would have had a powerful effect on establishing the foundations of future social



The Great Transformation of Tibet? 239

stratification in the urban areas of the TAR, which is especially significant because it was occurring at such a key moment in the transition of labor out of agriculture and into urban areas. In effect, the rapid increase in subsidized urban wealth that was driving intersectoral polarization was increasingly unequally distributed between, on one hand, state-sector employees and others well connected to state-subsidized networks of wealth circulation in the TAR—including a privileged cohort of Tibetan cadres who were shrinking in number and share at least up until 2003—and, on the other hand, the less-privileged majority of urban residents, including urbanizing migrants and many out-of-province non-Tibetan migrants. There was thus a broader overarching trend of heightened polarization in the overall economy of the TAR underlying some heavily subsidized silver linings in the rural areas of the TAR (if the rapidity of the changes in these areas is to be taken as positive). The extent of polarization was more or less instituted by the government through its subsidy, wage, and statesector employment policies. This trend and the related educational, linguistic, and cultural modes of bias that severely disadvantaged the majority of Tibetans within their urban labor markets provide important insights into the outburst of protests in March 2008 in the TAR and other Tibetan areas. Indeed, the attenuation of urban wage inequality in 2008 might well reflect government responses to the protests. It is in this sense that the government strategy of attempting to mollify Tibetans through various heavily subsidized development strategies—as discussed in Goldstein et al. (2010)—essentially backfired. In particular, the relative retrenchment of state-sector employment as a share of total urban employment—much more so than elsewhere in China—limited the government’s ability to use such employment as a means to alleviate some of the pressures and potential dislocations of this key phase of social transition. For instance, such employment could have been used preferentially for employing an increasingly schooled young adult Tibetan population that, despite their improvements in schooling, still faced stark disadvantages within the pressurized context of intensifying competition over urban employment opportunities, as discussed in the next chapter. This is not to say that there were no Tibetans who benefited from the situation. Many obviously did. From a poverty perspective, most elites probably survived quite well through the various dislocations wrought by these rapid transitions, particularly if they had prepared their children well and with foresight as to the needs and demands of the “new society.” Tibetans in the middle of the social hierarchy, including some illiterates, might have also adapted relatively well in small businesses or petty trade, and some might have even established successful large businesses or engaged in profitable investments. Lesser-skilled Tibetans who have found some construction work have been able to contribute significant new sources of monetized

240

Chapter 5

income to their families. The majority of Tibetans who remained in agriculture also appear to have performed more positively since the early 2000s. However, the polarization that underlies these marginal improvements in wealth or poverty reduction also simultaneously exacerbates dislocation and insecurity across the social hierarchy. While the average Tibetan standard of living improved throughout these rapid transitions, a focus on marginal improvements distracts attention from larger dynamics in the regional economy, within which those who had marginally improved were being progressively marginalized from the more lucrative parts of the economy and from levers of decision making, even while becoming more dependent on the employment generated by the subsidies producing the affluence. Moreover, the rapid labor transitions that were induced by the development strategies have been very real, in terms of the radical transformation of people’s lives and sources of livelihoods. Indeed, the speed of transition itself calls into the question the subsidization strategy; slower change might render people more capable of self-determined adaptation, whereas the dependence of the emerging employment structure on subsidies is so great that the prospect of such subsidies one day drying up is very worrisome. In light of such predicaments and to the extent that many of these structural socio-economic changes prove to be irreversible, the prioritization of preferential employment generation in the Tibetan areas for local Tibetan people is urgently needed as a means to avoid rapidly emerging pockets of urban marginalization within these rapid labor transitions.

Appendix: Interpreting Official Employment Data for Tibetan areas in China The official aggregate employment data discussed in this chapter are collected and compiled by the National Bureau of Statistics from a variety of sources. These sources are exceptionally detailed in the explanatory notes of chapter 5 in CSY (2001) and are worth elaborating here, even though there might have been minor modifications to the reporting system since 2001. The “Comprehensive Labour Statistics Reporting System” comprises a complete enumeration and reporting from lower levels of government to higher levels of all “independent accounting units.” The resulting data is then adjusted based on the 1990 and 2000 population censuses and by the annual sample surveys on population changes. The “Sample Survey on the Population Changes” covers the population of the whole country through a multi-stage stratified cluster sampling scheme. The “Rural Social and Economic Survey” covers all rural areas below the township level. Data on the number of persons employed in urban and rural private enterprises and self-employed persons in industry and commerce are also collected



The Great Transformation of Tibet? 241

through “Statistical Reports on Basic Conditions of Urban and Rural Private Enterprises” and “Statistical Reports on Basic Conditions of Urban and Rural Individual Industrial and Commercial Business,” both provided by the State Administration for Industry and Commerce, covering the whole country and collected on the basis of administrative records. It is not clear in these sources to what extent the various employment definitions are equivalent to the definitions used for measuring populations, that is, with respect to residency and registration status. The sectoral definitions of employment (primary, secondary, and tertiary) are probably recorded according to dominant socio-economic characteristics (where these can be easily identified) rather than by agricultural versus non-agricultural registration status—a designation that has become quite meaningless in China for determining actual employment status in the context of rapid socio-economic changes over the last thirty years. Nonetheless, we can speculate that agricultural registration status might, to a certain degree, influence the reporting of employment in the more administrative Labour Statistics Reporting System, in lieu of the more social-scientific methods of the population surveys or censuses. Hence, we can speculate that some of the trends recorded in the share of the workforce in primary sector employment might reflect administered changes in registration status over this period. Similarly, the urban/rural classification of employment tabulated in table 4-2 in CSY (2001) and CSY (2011)—which report data on rural employment in township and village enterprises, rural private enterprises, and rural self-employed individuals—is based on data provided by the State Administration for Industry and Commerce and collected on the basis of administrative records. Nonetheless, the urban/rural classification probably reflects socio-economic characteristics (i.e., those actually working in urban or rural areas) rather than registration status (i.e., urban or rural hukou), particularly from the 2000 census adjustment onward, given that the 2000 census went to a great extent to improve on the enumeration of urban residency according to socio-economic characteristics rather than registration status. See Yixing and Ma (2003) for an excellent discussion of this point. There is likely to be some influence of urban/rural registration status in the reporting of the urban/rural employment data in the Labour Statistics Reporting System or in local administrative records (especially in poorer western areas where the administrative records might have been poorly adjusted to census results). Accordingly, changes in urban/rural registration status might have influenced some of the trends observed in these urban and rural employment data. Finally, because the Labour Statistics Reporting System, population surveys, and administrative records are generally based on households registered as permanently and/or formally residing in a location, these data would reflect only a portion of migrant contributions to local employment,

242

Chapter 5

that is, those migrants who register themselves or their businesses in their places of immigration (which might be a large proportion of migrants in places like Lhasa given the pervasion of the security system). As discussed in Fischer (2008d), considerable improvements were made in the 2000 census to measure migrant populations. Hence, the adjusted data as of 2000 probably reflect this improved appraisal of migrant contributions to local employment. Presumably, the same applies to the results of the 2010 census. Intercensal years would have been adjusted on the basis of the census baseline according to insights derived from population surveys. However, these annual population surveys cannot offer indications of the changing prevalence of migrant labor relative to local labor in an aggregate sense over these subsequent years (because they are sampled from the permanent resident population). Nonetheless, the consistency of the trends up until the 2010 census seems to imply a fairly consistent reporting across various systems. Further research is required to determine exactly how these employment data are collected, particularly in Tibetan areas, although suffice it to say that the trends observed in these data corroborate with the field insights by scholars including me, and they can probably be taken as roughly accurate reflections of real trends.

Notes   1.  In the modern period—that is, since the onset of demographic transitions and urbanization alongside related economic transformations—we have almost never observed situations where a labor force has become more agrarian in a structural sense, except under short-lived episodes of trauma, crisis, or extreme social engineering, such as under Pol Pot in Cambodia, certain periods under Maoism in China, or the collapse of the Soviet Union in the 1990s. However, even in all of these cases, once the proximate factor is removed, the structural trend in the population to move out of agriculture reasserts itself, often with a vengeance. For further discussion on demographic perspectives of urbanization, see Dyson (2011).  2. These observations are based on detailed interviews with twenty-five Tibetan secondary school students and focus group discussions with six classes, each with about twenty-five to thirty Tibetan high school students, conducted in Qinghai in 2004.   3.  See the appendix at the end of this chapter for a detailed explanation of the data sources used.   4.  For instance, see data in UNRISD (2010, chapter 1). To take a few examples, the share of employment in agriculture in the Republic of Korea fell from around 30 percent in 1980 to around 9 percent in 2006; in Brazil, from about 29 to 20 percent over the same period. The primary share barely fell in India between 1960 and 2005 (from just above to just under 70 percent). In Malaysia, the primary labor share fell from 55 percent in the 1960s to 16 percent in 2000.   5.  The TAR has the highest rate of population increase in China, although fertility started to fall sharply in the 1990s, meaning that successive age cohorts became



The Great Transformation of Tibet? 243

smaller in number (see chapter 3). Hence, the “youth bulge” in the population structure (see Childs 2008, 266) started reaching the working age in the 2000s (considered as age sixteen years and older in the employment statistics).   6.  Again, see appendix.   7.  Both categories of data are reported for each year in table 4-2 of CSY (2011) and equivalent tables in previous yearbooks.   8.  Previously, I used these three categories as a proxy for off-farm rural employment in the TAR (e.g., Fischer 2005a; 2009a), which seems to have underestimated the amount off-farm rural employment in the TAR and exaggerated the difference between the TAR and other regions.   9.  For a detailed discussion of Lhasa as primate city, see Yeh and Henderson (2008). 10.  Note that the provincial data for each year in these figures is derived from the subsequent annual provincial yearbooks, that is, 1993 data reported in CSY (1994). Hence, 1993 is chosen as the starting point because there was a change in the reporting categories for these particular data starting in CSY (1994); these category are not comparable with previous reporting categories. 11.  As noted in chapter 3, the measurement of urbanization is very problematic in China given that urban definitions are quite different in each of the five censuses (see Yixing and Ma 2003). Also, annual surveys on population change are based only on people registered as permanently residing and thus provide no basis for evaluating changes due to migration. With reference specifically to Lhasa on this point, see Yeh and Henderson (2008). 12.  In an interview, a senior scholar/official from the Tibetan Academy of Agriculture and Animal Husbandry (TAAAS) in the TAR in November 2004, indicated to me that a policy shift to emphasize rural incomes was taken in the TAR in 2003; following national policy trends. 13.  The volatility in this measure for Sichuan, Gansu, and Qinghai from 1993 to 1998 probably represents data problems, although it might also be partly reflective of the retrenchment of state support for TVEs and other local state-owned ventures in western China during these years (see chapter 8 for further discussion of such local industrial restructuring in Qinghai). Indeed, there was a consistent trend in these three provinces of substantial declines in TVE employment over these years, compensated for by rising private enterprise and individual employment. 14. Rural manufacturing employment was not reported for 2009 or 2010, whereas rural employment in mining was not reported for 2008 and the amounts in each case are similar, furthering the supposition that these rural activities were simply lumped together in each year, reported as manufacturing in TSY (2009) and then recategorized as quarrying in TSY (2010). 15.  Productivity is, in effect, almost impossible to measure across heterogeneous goods and, in particular, across non-tangible tertiary services. Mainstream (neoclassical) economists almost always use monetary value-added (i.e., GDP) as a proxy for measuring productivity, although this approach is severely flawed given that value-added represents a combination of output and prices/wages. See my previous article (Fischer 2011b) for further discussion of this point. As a result, I use the term “productivity” with caution in this context of intersectoral comparisons. 16.  In an interview in June 2004, an official in Qinghai pointed out to me that the industrial policy of the province was to specialize in higher-value forms of

244

Chapter 5

steel and aluminium production, due to the abundant supply of electricity in the area (and not because of the supply of raw material inputs, which were actually imported from other provinces), given that steel and aluminium production are both extremely energy intensive. The abundant electricity supply was generated by two of the largest dams in China that were completed in the 1990s, both in Tibetan counties, that is, the Lijiaxia dam in Chentsa County, Malho Prefecture, and the Longyangxia dam in Chabcha County, Tsolho Prefecture. The official noted that electricity from these two dams was prioritized for these strategic industrial activities. See chapter 8 for further discussion. 17. All else held constant, rising average income reduces absolute poverty, whereas rising inequality increases absolute poverty. Together, it generally takes a lot of inequality increase to counteract the (absolute) poverty-reducing effect of a moderate degree of growth. Obviously, it is impossible to hold all else constant, particularly in a context of rapid structural changes in rural livelihoods. 18.  See a detailed analysis of these earlier household income and poverty data, Chinese poverty lines, and poverty measurements more generally in Fischer (2005a, 88-126). 19.  Li (2012, 48) estimates that, on the basis of decomposition analysis, urbanrural income inequality in China accounted for about 48 percent of national income inequality in 2007, up from 38 percent in 1995 and 43 percent in 2002. 20.  This ratio cannot be taken as a precise representation of income differences given that the methods of measuring rural and urban household incomes are different. However, our main interest is in the relative change of the ratio over time, not its precise level. 21.  Urban-rural inequality has generally been higher in the western than in the eastern provinces due to the fact that state-sector wages (and thereby a large component of urban incomes) have historically been equalized to a national standard, give or take the cost of living in each location, while rural incomes have not. The poverty of the western rural areas therefore results in a greater divide between the western urban and rural than is the case nationally, particularly in comparison to coastal China, where many rural areas have profited from rural industrialization. As a result, it is generally the case that the poorer the rural area, the greater the urban–rural divide. For a more detailed discussion of this point, see Fischer (2005a, chapter 4). 22.  See Fischer (2005a, 121-23) for further detail. 23.  It is more common to measure the ratio of top and bottom shares of total income rather than average incomes within each decile or quartile share of the population (ordered by income). I have chosen the latter because it gives a clear indication of the extent of average wealth difference between the population shares. 24.  “Staff and workers” do not include persons employed in township or private enterprises, urban self-employed persons, retirees, re-employed retirees, teachers in the schools run by local people, foreigners, persons from Hong Kong, Macao, and Taiwan, and other persons not included by “relevant regulations” (CSY 2005, explanatory notes for chapter 5). 25.  In the explanatory notes of chapter 4 of CSY (2011), “persons employed in various units” is defined as: “all the persons working in government agencies of various levels, political and party organizations, social organizations, enterprises and institutions, and receiving wages or other forms of payment. They include fully-



The Great Transformation of Tibet? 245

employed staff and workers, re-employed retirees, teachers in the schools run by the local people, foreigners and Chinese compatriots from Hong Kong, Macao, and Taiwan working in various units, part-time employees, employees of other units working temporarily at current posts, and employees holding the second job, but do not include persons who have left their working units while keeping their labour contract (employment relation) unchanged. This indicator reflects the total number of laborers actually engaged in production or other operations in various units.” 26. The explanatory notes in CSY (2007), in which the new terminology was introduced, are explicit about this although the explanatory notes in later CSYs are not as explicit. 27.  This greater discrepancy might be related to a huge disparity revealed by the new data between the average wages of the unit registration category of “other” compared to other unit registration statuses, as reported in CSY (2011, table 4-13). However, it is not clear what “other” refers to in this case. It is not explained and this registration category does not appear in the urban employment data by registration status (i.e., CSY 2011, table 4-2). Moreover, the share of employment in this category of “other”—as implied by subtracting the rest of the categories from the total employment in urban units reported in table 4-6—is negligible, amounting to less than 1 percent of total urban employment in all the cases shown. 28.  In the three overlapping years (2006-2008) when both categories were reported (in CSY 2007 to CSY 2009) the wages of employed persons in urban units was referred to as “earnings.” 29.  The provincial total urban employment reported in CSY (2011, table 4-2) is simply the sum of the registered categories and is thus not equivalent to the national tally. Larger totals can be obtained from most of the provincial yearbooks (besides TSY 2011), even while the registered categories in these provincial yearbooks are mostly consistent with the registered categories in CSY (2011). We can therefore assume that the urban totals reported in the provincial yearbooks, where available, are equivalent to the national urban total in CSY (2011, table 4-2) in that they include estimates of non-registered urban employment. 30.  The equivalent measure of informal urban employment was 35 percent in 2000 and 39 percent in 2003, and then it gradually fell to its level in 2010. The fall was due to a substantial rise in the proportion of urban employment in registered private enterprises or registered self-employed, whereas the share of employment in urban units declined. This could have been due partly to a registration of unregistered private enterprises and self-employment, as well as a proportionate transfer of employment from state or corporate units into private or self-employment. 31.  See Fischer (2007, 181-83) for further analysis of the data in 2004. Data for 2000 are calculated from CSY (2001, table 5-4). 32.  In the earlier data, employment in state-owned units was higher than staff and worker employment in TAR, although not in any other province. This was because almost all staff and workers were working in state-owned units (unlike in other provinces), whereas not everyone working in a state-owned unit was classified as a staff or worker (see Fischer 2007, 181-83). This discrepancy has disappeared in the new definition of employment in urban units. 33.  Nationally, wages and salaries accounted for about 72 percent of the average disposable incomes of urban households (registered as permanently resident) in

246

Chapter 5

2010, and most provinces exhibited similar shares, whereas the share in the TAR was 98 percent. Net business income and income from properties apparently played a very minor role in the average incomes of urban permanent households in the TAR, and income from transfers was much lower than in other provinces. 34.  The TAR ranks at the highest of eleven levels in a ranking of so-called hardship posts in public sector employment in China (“hardship” defined according to a lowland Han Chinese perspective). 35. Calculated from CSY (2009, table 4-23) and equivalent in previous yearbooks. 36.  See PRC (2001b). For an academic version this argument, see Sautman and Eng (2001). 37.  Coincidentally, I published a report on these data in early 2005 (see TIN/ Fischer 2005). The subsequent TSY (2005) no longer reported these data. 38.  Calculated from CSY (2009, table 4-8) and equivalent in previous yearbooks. 39.  The idea that “spontaneous” migrants are directly subsidized by the state is a common Tibetan belief (exile and local). However, during fieldwork in the TAR in 2004, I could find no indications of this except in state-sector construction projects such as the railway, where wages on offer were higher than normal.

6 The Education-Employment Nexus of Urban Exclusion in Tibet

The experience of marginalization or exclusion in Tibet needs to be understood in terms of the interactions between, on one hand, the transitions and polarizations analyzed in the previous three chapters and, on the other hand, underlying educational inequalities, cultures of work, and networks of political and economic power. The synthesis of these interactions in this chapter represents the culmination of the more macro-structural analysis of this book, combined with an institutionalist1 analysis of changing employment norms and an entry into the qualitative fieldwork findings presented in the final chapters. An important point emphasized in the chapter is that the particular character of these interactions in Tibet has conditioned the ethnically exclusionary and discriminatory patterns of development at both the lower and, importantly, the middle and upper strata of the urban labor hierarchy in the Tibetan areas. Notably, exclusionary processes occurring in middle and upper social strata will not necessarily appear as rising poverty or even inequality, even though inequality measures might offer some hints or partial insights about the presence of such processes. It is precisely for this reason that, as noted in the introduction of the book, absolute or even relative economic and human development indicators (such as falling poverty or improving levels of literacy and schooling) offer little insight into the reasons why Tibetans might have come to be so aggrieved in the course of development and rapid economic growth. Rather, attention must be placed on structural disjunctures and asymmetries, and institutional modes of integration occurring across social hierarchies, not just at the bottom, as essential elements for understanding exclusionary processes. Educational inequality, for instance, might offer 247

248

Chapter 6

clues of these processes by revealing the spaces within which we might expect to find exclusions operating, but it does not actually provide a means of identifying exclusions. In other words, a disparity in itself does not necessarily imply exclusion. If farmers choose to remain in farming, they do not necessarily face exclusion despite poorer incomes or schooling outcomes. Similarly, rural people migrating to cities and towns might face much stronger exclusions than those remaining behind even though they also tend to be upwardly mobile, with better incomes and schooling on average. Following this example, the study of exclusion is not about identifying whether urbanizing migrants are excluded in an absolute sense. Indeed, they are definitely more “included” into the urban economy than their rural compatriots, even while simultaneously facing much stronger exclusionary pressures within certain dimensions of their integration, particularly within the categories of urban employment that they target in accordance with their wage and various other social expectations. In this sense, exclusion and inclusion work together dialectically, in that exclusion plays a role in determining the subordinate integration of migrants into the urban economy, whether or not this results in greater income or other “development” outcomes. Indeed, migration might lead to an increase in exploitative or degrading work performed by migrants. Parallels can be made with workers in sweatshops who, despite the exploitation involved, might nonetheless earn more wages than in other less exploitative forms of employment. The fact that this would be reflected as rising incomes (and hence falling poverty rates) tells us little about the exploitation or degradation involved in such work. This approach to the study of exclusion differs from dominant approaches in the development studies or social policy literatures, which tend to focus on the identification of exclusion by way of deprivations or disparities, implying that exclusion is more or less synonymous with poverty, whether poverty is defined in absolute, relative, or capability terms (see Fischer 2008a; 2011a). As a way to step beyond this static dichotomous view, an interdisciplinary approach to identifying institutional modes of integration and/or segregation operating across social hierarchies arguably needs to be taken in order to properly ascertain the significance of exclusion. With respect to employment, attention should be focused on biases and disadvantages operating among comparable educational cohorts sharing similar levels of schooling attainment and similar types of employment expectations, such as between Tibetan and Han high school and university graduates, most of whom are not reflected in any deprivation or disparity measures. Particular attention needs to be focused on qualitative educational inequalities, such as types of schooling, quality of schooling outcomes, and especially linguistic ability in Chinese. Such an approach brings to light the importance of



The Education-Employment Nexus of Urban Exclusion in Tibet 249

differentiating exclusion from poverty (even relative poverty) given that exclusionary processes can occur vertically throughout social hierarchies, among the poor and the non-poor, and in many cases they might intensify with movements out of poverty, such as in a context of urbanization or rising schooling levels. Indeed, the most politically contentious exclusions are often those that occur among relatively elite or upwardly mobile sections of a population. Therefore, the methodological challenge that faces studies of exclusion, as with the horizontal inequality approach (i.e., Stewart 2002), lies in finding ways to measure structural disjunctures and asymmetries and institutional modes of integration that move beyond either absolute measures, per mainstream approaches to human development, or even relative (i.e., inequality) measures, given that both are capable only of identifying potential exclusions occurring at the bottom of a social hierarchy.2 Critical in this regard is the interaction between schooling and urban employment systems, here examined through the cases of Qinghai Province and the Tibet Autonomous Region (TAR). Notably, the economic development strategies described in the two previous chapters had been implemented alongside ongoing efforts to raise school enrollments, in line with the national goal set in 1986 of achieving nine-year compulsory education for all by 2000, and subsequent efforts to achieve the same after 2000. Nonetheless, the implementation of competitive labor market reforms, in a situation of underlying educational inequalities and political and economic subordination, became a crucial mechanism by which rapid growth accentuated linguistic, cultural, and political modes of bias deriving from characteristics of the dominant cultural and political group, such as Chinese fluency, Chinese work cultures, and connections to government or business networks in China proper. While labor market reforms have also generated problems elsewhere in China, such as high rates of unemployment among university graduates,3 the Tibetan areas manifest a unique structural asymmetry whereby the most educated category of local residents (urban Tibetan men) is much less educated on average than even the least educated category of interprovincial migrants competing in local urban labor markets (i.e., rural women from Sichuan). Even high-schoolor university-educated Tibetans face comparable disadvantages in this context, which are simply not reflected in the quantitative comparisons. In other words, on top of facing an institutional environment where opportunities created in the local economy are dominated by state-centered and externally oriented networks of power and wealth, urban Tibetans competing at the middle and elite strata of employment also have to contend with the fact that they do not profit from natural educational advantages over the average Han migrant coming from elsewhere in China, contrary to all other regions of China.

250

Chapter 6

The resultant playing field, particularly in upper labor strata, offers important insights into recent tensions, even before considering the more blatant proximate causes such as prejudice, discrimination, or political repression. Indeed, exclusions experienced at the upper end of the labor hierarchy (such as with regard to accessing public employment) are important from the perspective of conflict given that they are very politically sensitive, even if not necessarily reflected as increasing poverty, especially given that nationalist sentiments tend to be more pronounced (or at least more clearly articulated) among the elite and educated. Moreover, the fact that these exclusionary experiences operate along educational, linguistic, or cultural modes of disadvantage provides the basis for strong cross-class perceptions and expressions of grievance. This being said, the focus in this chapter on the structural and institutional dimensions conditioning experiences or perceptions of exclusion is not meant to imply a deterministic understanding of structure. As stated in the introduction, it is important to differentiate structuralism as an approach in (early) development economics from the typical understanding of structuralism in sociology and anthropology regarding debates about structure versus agency. As will be evident in the final chapters, the approach taken here accepts a wide range of agentive responses to these structural and institutional factors. The focus on structural constraints and opportunities or else institutional compulsions and incentives that condition individual choice and/or agency should not be misunderstood as implying determinism. This assessment of the conditions underlying the outbreak of widespread and large-scale protests in 2008 and subsequent ongoing protests differs somewhat from Barnett (2009), who argues that the 2008 unrest was unique. In comparison to earlier protests, he argues that the recent protests were not primarily urban but involved new cross-sections of the population, including substantial participation from rural areas, which he claims had not been involved in previous protests during the reform period (such as the protests in and around Lhasa in the late 1980s, which were predominantly urban). My own reading is that the recent protests have been more specifically related to urbanizing rural areas (as well as urban centers) or to populations that have been increasingly in spatial flux in the midst of a crucial moment within rapid development transitions. While the social basis of these recent and ongoing protests has been definitely wider than in the protests of the 1980s—and hence more rural in a sense—they have not been more rural in comparison to widespread resistance in the 1950s, the memory of which is still quite vivid among the older generation of Tibetans today. In this respect, despite important lineages and continuities from the past, there is something very new about the protests occurring now, although this is not necessarily their rurality.



The Education-Employment Nexus of Urban Exclusion in Tibet 251

These observations are important because they imply that two critical dimensions of current Tibet policy need to be seriously addressed in order to lessen both economic polarization and social instability. On one hand, urban employment for locals requires protection and promotion at both lower and upper strata. On the other hand, the linguistic and cultural disadvantages faced by Tibetans in urban employment need to be lessened, albeit in ways that do not undermine Tibetan language and culture. The importance (and realities) of Tibetanization have been amply explored by other scholars,4 such as with respect to bilingual education or else in response to the arguments of Ma (2007) and other scholars who argue that the eradication of the preferential minzu system is necessary in order to deal with the politicization of minorities in China, as noted in the introduction.5 To a certain degree, the existing national minority laws in China already provide ways to resolve both dimensions of current Tibet policy if their full implications are put into practice, as exemplified by the recommendation by the Tenth Panchen Lama that public sector employees working in Tibetan areas should have at least some working knowledge of Tibetan language, supported by a strong promotion of bilingual education (i.e., Tibetan medium in all subjects) in the minority areas. Nonetheless, some would argue—and have argued—that this suggestion is unrealistic, particularly in the current context and given the assimilationist agenda of the central government. Rather, a common argument is that the best way to address the educational disadvantages faced by Tibetans and their consequent marginalization from urban employment opportunities is through improving their fluency in Chinese and their competency in Chinese-medium education, so as to improve their ability to function and compete in a regional economy and society where Chinese is effectively the hegemonic lingua franca. Admittedly, the evidence of this chapter could be used to support either argument, as I have come to increasingly realize over the years. While I have my own opinions on this matter, I do not delve into them in this chapter, but reserve them for the final concluding chapter. The exploration of this chapter is in two parts. The first section analyzes exclusionary processes at the lower end of the labor hierarchy in Tibetan areas through a comparison of literacy and schooling outcomes between 2000 and 2010 in the TAR, Qinghai, Gansu, Sichuan, and China, which sheds light on the exceptional educational disadvantages faced by local Tibetans in these areas, particularly in a context of in-migration from regions with much higher schooling outcomes. The second section then examines exclusion at the upper end of the labor hierarchy due to labor market reforms and the increasing institutional emphasis on rationalization and standardization in public employment, which has resulted in a notable disadvantage of elite Tibetans in these coveted areas of employment. The conclusion delves into some of the Tibetan nationalist responses, which will be taken up further in subsequent chapters.

252

Chapter 6

Exclusion at the Lower End of the Labor Hierarchy As noted in the previous chapter, heavily subsidized urban-centered rapid economic growth has effectively caused a confluence of two migration flows into the urban centers of Tibet: the interregional migration of nonlocal Han Chinese, Chinese Muslims, and even Tibetans from other Tibetan areas, and the urbanization of local rural Tibetans. The bulk of “spontaneous” or “floating” interregional migrants, that is, those not working in urban units, compete directly in activities where the average local Tibetan is more likely to find employment in the urban areas, such as, in relatively low-skilled and low value-added construction or service sector activities such as trade or catering. In this context, non-locals—especially Han Chinese non-locals—easily out-compete locals due to their higher skills (relative to the skills demanded in such work), more competitive aptitudes, knowledge of Chinese language, acculturation into Chinese work cultures, and better connections to the sources of economic and political power inside and outside the TAR. In addition to these advantages, Han Chinese and Chinese Muslim migrants from locations such as Sichuan and Gansu possibly even have lower wage and related work expectations, as discussed in the next chapter. This competitive disadvantage of locals is most simply represented by the severe lags in educational outcomes observed among Tibetans relative to most other regions and nationalities in China, such as in literacy and levels of schooling attained. Such educational inequality, particularly in comparison to the western and central provinces that are the main sources of emigration to the Tibetan areas such as Sichuan, Gansu, Henan, or Anhui, have profound implications in terms of blocking the effective adaptation and integration of rural Tibetan migrants into the urban labor force, which in turn complicates the role given to local urbanization in regional development policy. In this manner, educational inequality can provide an indirect indication of ethnically exclusionary processes at the lower end of the urban labor hierarchy by reflecting the mismatch between the skills (including linguistic and cultural skills) demanded by polarized growth and the skills present in the majority of the local Tibetan labor force. This emphasis on educational inequality is in contradistinction to the urban-rural divide emphasized by many authors, such as Sautman and Eng (2001) and Hu and Salazar (2008), as discussed in earlier chapters. By the early 2000s, both the urban-rural divide and the proportion of Tibetans with some form of secondary schooling and above (including vocational) sliced a similar 15:85 ratio across the Tibetan population. Nonetheless the two did not necessarily overlap given that the proportion of permanent urban residents in the TAR who were illiterate was exceptionally high. Thus, the educational divide rather than the spatial divide



The Education-Employment Nexus of Urban Exclusion in Tibet 253

is much more relevant in determining the exclusionary nature of development in the case of the TAR.6 Accordingly, ethnic discrimination would be best identified through differences in the provisioning of education across majority and minority ethnic groups. For instance, as analyzed for the TAR in the two previous chapters, despite the emphasis on education and training in the official rhetoric for western development, alongside the priority given to urbanization as means to confront rural poverty, the actual inputs, outputs, and outcomes of education and training in the Tibetan areas vary widely from the stated objectives. This is particularly the case at the level of vocational education—presumably the most important for forming practical skills at lower levels of the urban labor hierarchy—particularly if compared to the priority given to public management, which is generally restricted to those with tertiary levels of schooling. Zenz (2013) provides similar evidence of such discrepancies in the Tibetan areas of Qinghai, suggesting systemic biases in education provisioning across the Tibetan areas of China. Illiteracy, Schooling Rates, and Educational Disadvantage in Tibet The schooling and skills imbalance between locals and inter-provincial migrants in the Tibetan areas is best portrayed by inter-provincial comparisons of illiteracy and schooling rates, the latter expressed as the cumulative proportion of the population with up to certain levels of schooling (not necessarily completed). These rates should be seen as probabilities, that is, the chance that a sampled person is literate or has such and such a level of schooling or above. Illiteracy or no schooling would generally imply few skills beyond agriculture, manual labor, basic trades, or localized commerce (“generally” because idiosyncrasies obviously exist, e.g., the richest man in Qinghai in 2004 was apparently illiterate). Because literacy in any official language qualifies someone as literate in the Chinese surveys, this measure does not necessarily offer a good evaluation of Chinese literacy: the government itself estimated sometime in the early 2000s that about 80 percent of the rural population of the TAR did not speak or understand Chinese, meaning that a significant proportion of literate rural Tibetans were illiterate in Chinese.7 A primary level of education would generally imply, at best, low or semi-skilled labor (outside agriculture) and would not necessarily guarantee literacy. Moreover, the medium of most rural primary schooling in the TAR (and other Tibetan areas) has been in Tibetan language and thus a primary level of education in these areas probably precludes literacy in Chinese or even functional proficiency in spoken Chinese. Significant skills formation starts to take place at the secondary and vocational levels, as does Chinese fluency in the Tibetan areas given that most secondary education switches over to a predominant use of

254

Chapter 6

Chinese-language medium, although the precise timing and extent of this switch varies from region to region and even school to school. At least completed senior secondary school and ideally tertiary education would be a necessary qualification for most urban unit employment, especially in the state sector.8 As in the previous chapters, four provinces suffice for this comparison, keeping in mind that these quantitative indicators do not reveal any information about the quality of schooling, which is often very poor in Tibetan areas. Sichuan is the most important source of emigration to the TAR and to the Tibetan areas of Sichuan and Qinghai, while Gansu is the second most important source.9 Education indicators in these two provinces and the national average therefore give a rough indication of the average schooling levels in the sources of emigration to the Tibetan areas. As pointed out in much migration research, migrants often have higher levels of education than the average in their sources of emigration. In other words, those who migrate are not necessarily the most destitute or deprived in their communities, but tend to be among the more educated, entrepreneurial, and even wealthy (given that it generally requires resources to finance migration). This insight is confirmed by several surveys in Lhasa among both Tibetan and Han temporary migrants.10 Nonetheless, levels of education in the sources of emigration reflect the general culture of schooling and skills that the migrants have been influenced by, are leaving from, and maintain networks with throughout the course of their migration. The illiteracy rates among the population aged fifteen and older from 1990 to 2010 are shown in figure 6.1. In all provinces shown, illiteracy rates declined over time. Nonetheless, variability between censuses and surveys makes exact evaluation difficult. Illiteracy apparently dropped by about 15 percentage points in the TAR, between the 2000 and 2010 censuses, presumably the two most accurate measurements available, from 47 to 32 percent. However, the sharp drop between the 2009 survey, at almost 40 percent, and the 2010 census might partly reflect the inclusion of more literate out-of-province migrants in the census (or else measurement problems with the surveys). Excluding such migrants—particularly non-Tibetan migrants—from the census would presumably result in a significantly higher illiteracy rate, although the tabulated census data were not available at the time of writing in order to verify this likelihood. Similar reductions in overall illiteracy rates were observed in other provinces, from 25 to 13 percent in Qinghai, from 20 to 11 percent in Gansu, from 10 to 7 percent in Sichuan, and from 9 to 5 percent nationally. In all cases including the TAR, there was little improvement between the 2000 census and the 2005 one-percent survey (presumably the next most accurate measure), which might have been due to either measurement problems with the 2005 survey or else was reflective of the fact that children impacted by education cam-



The Education-Employment Nexus of Urban Exclusion in Tibet 255

paigns implemented at the primary school level since the late 1990s had not yet reached the age of fifteen by the mid-2000s. Notably, the illiteracy rate for all Tibetans in China (more than double the population of Tibetans in the TAR) was identical to the TAR in the 2000 census (see CPSY 2003, table 2-2). This implies that illiteracy rates among Tibetans inside and outside the TAR were almost identical at that time. Ethnically disaggregated data from the 2010 census was not yet available for Qinghai or the other provinces containing Tibetan areas at the time of writing, although we can presume that the reductions in illiteracy among Tibetans in these provinces were similar to those in the TAR. Indeed, my own field observations, as well as those of Bangsbo (2008) and Zenz (2013), also in Qinghai, Iselin (2011) in Sichuan, and Goldstein et al. (2008) and Postiglione et al. (2012) in the TAR, all clearly indicate that there was a strong push toward schooling from within Tibetan communities since the late 1990s.11 Hence, if the educational drives that started in the late 1990s or early 2000s targeted enrollments among primary-school-aged children, these initiatives would have taken five to ten years before appearing in the population aged fifteen and older. Age-specific education data would offer a much better evaluation of changes in this respect, but these are not available in the official data.12 Indeed, the largest cohort in the 2000 population census was in the ten-to-fourteen-year-old age group. Much of this cohort might have missed out on earlier educational drives in the middle to late 1990s but nonetheless would have had a disproportionate impact on illiteracy rates in surveys for several years after 2000 as they entered the age of fifteen and above but before younger and more educated cohorts would have started to Figure 6.1: Illiteracy rates, two-period moving averages, 1990-2010

Source: CSY (2011, tables 3-9 and 3-11) and equivalent tables in earlier and later yearbooks.

256

Chapter 6

reach the same age threshold, which possibly explains the more significant reductions in the illiteracy rate for the TAR in the later 2000s. One reason why illiteracy rates might not reflect improvements in enrollments—at least not by as much—is that the poor quality of many rural schools resulted in a situation of poor educational outcomes despite improved schooling. This possibility is supported by numerous field observations. For instance, I visited one village primary school in Yushu where three grades of students were taught more or less the same lessons from year to year by a teacher who was regularly absent and often inebriated when present. In the context of a survey, such students would be recorded as enrolled and having achieved some level of primary education, although they would probably fail the basic literacy test of the same survey. Similarly, Tashi Rabgey noted cases from her fieldwork in the Tibetan areas of Sichuan where students who completed five years of primary education were still unable to read or write their names.13 Under these situations it is likely that improving levels of primary schooling would be reflected by much slower improvements in literacy, even despite the lag effect described above. In other words, increasing levels of primary school enrollment might not result in improved literacy outcomes if such enrollments took place in already poorly funded, low-quality, and often overcrowded rural schools.14 This is precisely one of the issues that has motivated national government policies of consolidating primary schools in county towns, as discussed in Postiglione et al. (2012) with respect to the TAR or else by Cencetti (2012) with respect to the implementation of this policy in a prefecture of Qinghai from 2009 onward, which resulted in the closure of many rural primary schools. Nonetheless, this policy response to poor-quality rural schooling in the Tibetan areas has likely created new problems in an attempt to solve older ones, particularly given the reliance on boarding. Regardless of the precise accuracy of these literacy measures in both Tibet and elsewhere in China, the interprovincial comparisons nonetheless reveal several exceptional asymmetries faced by Tibetan areas, as represented by the TAR. These asymmetries are shown in table 6.1, which disaggregates illiteracy rates by rural, town, and city dwellers from the 2009 population survey. The equivalent from the 2010 census was not yet available at the time of writing, although the advantage of the survey data in any case is that they would be more representative of a (permanent resident) Tibetan urban experience, versus the aggregated census data that would include a much stronger representation of non-Tibetan migrants.15 The category of “city” in the TAR survey refers to the two cities of Lhasa and Shigatse (TSY 2005, table 3-7). The town category would include both county towns as well as smaller townships that were reclassified as urban towns in 2000 (e.g., see Yeh and Henderson 2008). Most of such smaller townships in the TAR and other Tibetan areas are effectively rural (in socio-economic terms), whereas



The Education-Employment Nexus of Urban Exclusion in Tibet 257

Table 6.1: Illiteracy rates among the population aged 15 and older, 2009 survey Total

Rural

Town

City

Rural/City

TAR

39.6%

40.9%

34.3%

37.2%

1.1

Qinghai

14.7%

20.0%

11.7%

3.2 %

6.3

Gansu

15.9%

19.6%

12.1%

5.6 %

3.5

Sichuan

9.2%

11.9%

6.4%

2.8 %

4.3

China

7.1%

9.8%

6.2%

2.6 %

3.8

Source: CPEY (2010, tables 2-28, 2-29, 2-30, and 2-31).

they would be more urban in the more densely populated areas of China. Accordingly, many of the towns in the TAR would be mostly Tibetan, as with the rural areas in the TAR. All the Tibetan urban centers in Qinghai, Gansu, and Sichuan would be classified as “towns” although the majority of the population associated with the town category would nonetheless represent Han Chinese, as discussed in chapter 3. The high level of city illiteracy in the TAR in 2009 is exceptional, at over 37 percent, or almost parity with the rural rates. This occurrence is unique among all the Chinese provinces. The next highest rate of city illiteracy was observed in Gansu, at 5.6 percent. Obviously, the small city sample of the TAR (n = 197) would tend to lead to more substantial measurement errors from year to year in comparison to the larger provinces, although these results have been consistent with earlier surveys. For instance, the 2001 TAR survey recorded illiteracy rates at 44 percent in the cities and 46 percent in the rural areas; the 2002 TAR survey recorded 36 percent in the cities and 49 percent in the rural areas; the 2003 TAR survey recorded 53 percent in the cities and 42 percent in rural areas; and the 2004 survey recorded 47 percent in the cities (n = 586) and 43 percent in rural areas (see Fischer 2009a, 18). In the 2005 one-percent survey, which was presumably the most accurate measure since the 2000 census and with a city sample of 3,944 people, the city illiteracy rate was 32.4 percent and the rural rate was 47.9 percent (CPSY 2005, tables 2-28 and 2-30). The 2000 census recorded a lower urban (not city) rate at 23.7 percent (versus a rural rate of 53.9 percent), although the census included temporary out-of-province migrants, who in turn were concentrated mostly in the urban areas of the TAR, thereby bringing down the urban rates in comparison to the surveys, which are sampled only from the permanent resident population.16 Some of the persistently high levels of city illiteracy might be due to the administrative expansions of Lhasa and Shigatse, whereby semi-rural peri-urban areas and their less-educated populations are incorporated into the expanding boundaries of these two cities, although the fact that illiteracy would be

258

Chapter 6

so high in such peri-urban areas located so close to major cities is in itself striking. The fact remains that the working- age permanent-resident city population of the TAR has been burdened with high levels of illiteracy not observed anywhere else in China, even within rural areas. Indeed, strong improvements in reducing rural illiteracy had been made over the decade in all provinces besides the TAR (such as from 32 to 20 percent in Qinghai from 2004 to 2009, versus from 43 to 41 percent in the TAR). City rates in every other province were only a small fraction of rural rates and were within a close range of the national average, and this range had been narrowing throughout the decade due to consistent improvements every year since 2001. For instance, the city rate in Qinghai was 9.6 percent in 2002, almost four percentage points above the national average of 5.7 percent, whereas it then fell to 5.9 percent in 2004, and to 3.2 percent by 2009 or less than one percent above the national average. Gansu was an exception in this regard because its rate of city illiteracy stayed the same between 2004 and 2009, at 5.6 percent in both years, and the province replaced Qinghai as the second most illiterate in China after the TAR, albeit at rates that were nonetheless much lower than those of the TAR. These results reveal the urban bias in current education strategies across China given that rural illiteracy rates were consistently three or more times higher than city rates in all cases besides the TAR. The TAR was the only exception, in a perverse sense that high rates of urban illiteracy were sustained despite the heavy urban bias in economic strategy. Against this setting, it is possible to make a rough assessment of the extent of literacy differences between (Tibetan) locals and migrants in the TAR. Out-of-province Han Chinese migrants in the TAR are best described by the rates of Sichuan, which were close to the national average and to other core western provinces such as Shaanxi. The most extreme comparison is between the city illiteracy rates of Sichuan and the rural illiteracy rates of the TAR, which is appropriate given that many of the Sichuanese migrants in the TAR emigrate from urban conditions whereas local urbanization in the TAR involves migration from Tibetan rural areas. In this case, the rural TAR rate of 40.9 percent in 2009 contrasted appallingly with the city Sichuan rate of 2.8 percent. Interestingly, Iredale et al. (2001, 156) measured a similar spread between Tibetan and Han migrants in Lhasa almost fifteen years earlier, in the mid-1990s. The spread and even the rates also showed little change from 2004 onward, as noted above. Even in the least extreme comparison, rural Sichuanese were more than three times less likely to be illiterate in 2009 than city residents in the TAR, at 11.9 and 37.2 percent respectively. This strong educational advantage of (Han) rural migrants over (Tibetan) city residents is an anomaly that is simply not observed anywhere else in China. In Qinghai, there was a sharp difference between the rural rate and the city rate. The rural rate was the second highest in the country in 2009 (al-



The Education-Employment Nexus of Urban Exclusion in Tibet 259

though closely followed by that of Gansu) and similar to the TAR rural rate up to the early 2000s. The city rate was very close to the national norm, as noted above. The town rate was in between, at almost half the rural Qinghai rate although almost double the national town rate. This city-rural disparity within Qinghai is plausibly equivalent to the inter-provincial comparison between Sichuan and the TAR given that it partially reflects the contrast between the urbanized Han northeast core of the province (Xining) and the rest of the province. As noted in chapter 3, Qinghai Tibetans were over 91 percent rural in the 2000 census and thus were mostly captured by the rural surveys, although these also equally capture the Han rural areas of the northeast core. Drawing from some county or prefecture level census data, literacy rates and schooling levels in most Tibetan areas of Qinghai were similar to those in the TAR in 2000, albeit with higher levels in the farming areas closer to the core region around Xining. Given that many of the Han (and Muslim and Tibetan) migrants in the Tibetan areas of Qinghai emigrate from this northeast core of the province itself, the disparity within Qinghai offers some reflection of the literacy differences between locals and migrants in these areas. The literacy data decomposed by sex, as shown in table 6.2, offer an even more evident portrayal of the severity of the educational asymmetries between local Tibetans and non-Tibetan migrants. A Tibetan woman migrating from the rural areas of the TAR in 2009 was almost 37 times more likely to be illiterate than a male city resident migrating from Sichuan. Inversely, a permanent-resident man in Lhasa or Shigatse was about one-third times more likely to be illiterate than a woman in rural Sichuan, despite the fact that she was almost three times as likely to be illiterate than her male counterpart in rural Sichuan. These odds had improved somewhat for the average male city resident of the TAR between 2004 and 2009 (from 1.7 to 1.3), although they slightly worsened for the average rural Tibetan woman of the TAR (cf. Fischer 2009a, 19). Again, the exceptional anomaly of the Table 6.2: Illiteracy rates among the population aged 15 and older, by sex, 2009 survey Total

Rural

City

Male

Female

Male

Female

Male

Female

TAR

31.6%

47.2%

33.6%

48.0%

24.3%

47.2%

Qinghai

7.8%

21.6%

11.1%

28.8%

1.2%

5.2%

Gansu

10.0%

21.9%

12.7%

26.6%

2.7%

8.6%

Sichuan

4.9%

13.4%

6.4%

17.5%

1.3%

4.2%

China

3.8%

10.5%

5.4%

14.2%

1.1%

4.2%

Source: CPEY (2010, tables 2-28, 2-29, 2-30, and 2-31).

260

Chapter 6

Tibetan areas is to be noted: the most educated local sex cohort (male city permanent residents) is much less literate on average than even the least educated sex cohort of interprovincial migrants coming to Tibet and competing in local urban labor markets (female rural residents from Sichuan). This asymmetry is simply not observed in any other province in China, nor had its severity improved significantly over the decade. Similar asymmetries can be examined through the data on schooling levels attained among the population aged six and older, shown in table 6.3. The no-schooling rate (not shown) would be the inverse of the proportion of the population with a primary level of schooling or above. Noschooling and illiteracy are not necessarily synonymous in these data given that, for the reasons discussed above, we cannot assume that a child with a primary level of schooling is necessarily literate. Nonetheless, the rates of no-schooling generally corroborate with the above illiteracy data. Based on the 2000 census tabulations, the primary level includes literacy classes; the senior secondary level includes senior middle schools (i.e., high schools), specialized secondary schools, and vocational schools; and tertiary education includes junior college, university, and post-graduates. Table 6.3 shows data from three years: 2000c (from the 2000 census), 2009s (from the 2009 population survey), and 2010c (from the 2010 census). In all five cases, the 2009 data is roughly correspondent with the trends since the 2000 census (as with the data on illiteracy shown in figure 6.1). As mentioned above, the detailed 2010 census tabulation was not yet available at the time of writing, so the ethnic decomposition of these schooling levels could not be calculated. Moreover, the 2010 census data reported in this table are not precisely comparable with the previous data because they are calculated on the basis of the whole population, not that aged six and older, and hence the denominator is larger than it should be (this was the only data from the 2010 census that was available at the time of writing, reported in CSY 2011, table 3-10). However, comparing the 2009 survey results with the 2010 census results nonetheless offers some insights (if the 2009 survey data can be taken as roughly representative) given that the surveys are sampled from the permanent-resident population whereas the 2010 census includes temporary residents (e.g., migrants) and thus would reflect a much stronger weight of Han Chinese and non-locals, who would tend to have much higher levels of schooling than locals. The effects of this become very apparent at the tertiary level. In terms of the comparison between the 2000 census and 2009 survey results (which are comparable in terms of being measured by the population aged six and older), several points are worth highlighting. First, in the TAR and Qinghai—the two most lagging provinces in terms of literacy and schooling rates—there was strong improvement at the primary level, from 47 to 63 percent in the TAR and from 73 to 86 percent in Qinghai (meaning



The Education-Employment Nexus of Urban Exclusion in Tibet 261

Table 6.3: Levels of schooling among the population aged six and older in the 2000 census, 2009 survey, and 2010 census (2010 census results with respect to the entire population) 6+ population with a level of schooling of at least (percent of total) Primary

Junior Secondary

Senior Secondary

Tertiary

2000c 2009s 2010c 2000c 2009s 2010c 2000c 2009s 2010c 2000c 2009s 2010c TAR

47.1

62.7

59.3

12.7

18.4

22.7

5.5

5.6

9.9

1.5

1.7

5.5

Qinghai 72.8

86.2

79.7

38.6

45.1

44.4

14.9

19.1

19.0

3.6

8.8

8.6

Gansu

80.2

85.0

83.9

39.9

50.0

51.4

13.7

16.6

20.2

2.9

4.8

7.5

Sichuan 89.1

90.8

87.4

42.7

51.2

52.8

10.9

16.1

17.9

2.7

5.6

6.7

China

92.9

88.5

52.3

62.8

61.8

15.8

21.1

23.0

3.8

7.3

8.9

90.5

Source: calculated from CSY (2002, table 4-12), CSY (2010, table 3-12), and CSY (2011, table 3-10).

that 86 percent of the population aged six and older had at least some level of primary schooling). These two provinces (and Gansu) also closed the gap with the rates in China or in Sichuan, where there was little improvement at this level of schooling given that the rates were already around 90 percent in 2000. There were also strong improvements at the junior secondary level (albeit from a very low level in the case of the TAR), in accordance with the drive to enforce nine-year compulsory education from the mid-1990s onward. However, at this level the improvements were progressively stronger in the provinces with higher rates in 2000, implying a widening of absolute disparities at this level. For instance, the proportion of the population with at least some level of junior secondary schooling increased 5.7 percentage points in the TAR (from 13 to 18 percent), versus an increase of 6.5 percentage points in Qinghai, 10.1 percentage points in Gansu, 8.5 percentage points in Sichuan, and 10.5 percentage points in China overall. Moreover, the very low rates at the junior secondary level in the TAR were relatively much lower than the primary rates in comparison to other provinces: the TAR primary rate was about three-quarters that of Qinghai in 2009 and about two-thirds that of China, whereas the TAR junior secondary rate was only two-fifths that of Qinghai and less than one-third that of China. There was also a much sharper drop-off between the primary and junior secondary levels in the TAR than in other provinces: the junior secondary rates in the TAR were less than one-third the primary rates in 2009, whereas they were just over half in Qinghai and just over two-thirds in China overall. The slower improvement, large lag, and sharper drop-off at the secondary level in the TAR might have been partly due to a combination of a later start in the education drives together with very low starting rates of junior secondary schooling in the TAR in the late 1990s.

262

Chapter 6

At the senior secondary and tertiary levels there were no discernible improvements in the rates for TAR up to 2009 (according to this comparison), whereas improvements were strong in all the other cases. In contrast to the TAR, the senior secondary and tertiary rates in Qinghai surpassed the rates in Gansu and Sichuan in 2000 and 2009, despite having lower rates of junior secondary schooling in both years, possibly implying a stronger retention of students between the junior and senior secondary levels. The tertiary rate in Qinghai surpassed that of China according to the 2009 survey (although not according to the 2010 census). This would tend to indicate a fairly unique education strategy in Qinghai, with strong education drives to expand senior secondary and tertiary enrollment despite having the second-highest level of illiteracy of all provinces in China (as of the 2010 census), in comparison to the relatively stronger outcomes at lower levels in more populated western provinces such as Gansu and Sichuan. The fact that such expansions are reflected in the 2009 survey results as well as the 2010 census—to the extent that the rates are basically the same in both— indicates that they were occurring proportionately among the local (permanent resident) population as well as among migrants from other parts of China. Hence, despite the influx of students from other parts of China into both regular and minority universities in Qinghai, as documented by Zenz (2013), this was arguably matched by a rapid increase in the intake of locals into these same universities. This reflects the intense drive—as elsewhere in China but apparently even more intense—toward the expansion of tertiary education in Qinghai by governments and universities, and from the ground up by people’s own initiatives, including Tibetans’. The uniqueness concurs with the assertion of Zenz (2013) that the educational situation in Qinghai differs considerably from other Tibetan regions and that Tibetan enrollment in both Tibetan and Chinese medium secondary and tertiary schooling has gone through considerable expansion since the mid-1990s. This expansion has happened even despite the intensified exclusionary pressures experienced by Tibetan graduates in labor markets (as analyzed later in this chapter). The situation appears very different in the TAR according to these data, arguably pointing to a polarization in education outcomes at the senior secondary and tertiary levels between locals and migrants, and a heavy reliance on migrants for tertiary-educated labor rather than educating locals for the supply of such labor. This becomes evident in the comparison of the 2009 survey and 2010 census data. As noted above, the only 2010 census data that were available at the time of writing (from CSY 2011, table 3-10) appear to have been calculated on the basis at the total population, rather than the population aged six years and older (as per convention). All other factors constant, this should result in lower rates than in the 2009 survey data (assuming that the latter are roughly representative) given that the



The Education-Employment Nexus of Urban Exclusion in Tibet 263

denominator used in CSY (2011, table 3-10) is larger. This would be counterbalanced by the fact that there might have been continued improvement in these rates between 2009 and 2010. This deduction holds at the primary level for all the cases shown in table 6.3, that is, the primary rate is lower in these 2010 census data than in the 2009 survey data for all five cases. The deduction holds at the junior secondary level for Qinghai and China, whereas the 2010 census rates are slightly higher for Gansu and Sichuan, and even more so at the senior secondary and tertiary levels. However, the 2010 census data for the TAR shows the greatest discrepancy: the junior secondary rate according to this 2010 census data (as reported) is 4.3 percentage points higher than the 2009 survey. Similarly, the senior secondary rate measured by the census is 4.3 percentage points higher and the tertiary rate is 3.8 percentage points higher. Whereas only 1.7 percent of the survey sample in the TAR in 2009 had a tertiary level of schooling, 5.5 percent of TAR residents in the 2010 census had a tertiary level of schooling. The difference is most likely explained by the inclusion of temporary out-of-province migrants in the census, versus their exclusion in the surveys. Moreover, the fact that the discrepancy is almost identical at the junior and senior secondary and tertiary levels suggests that the source of most of the discrepancy is coming from tertiary-educated temporary migrants. Unlike in Qinghai, the jump in senior secondary and tertiary levels of schooling recorded in the TAR by the 2010 census would have been mostly, if not entirely, due to the in-migration of out-of-province migrants, who in general had much higher levels of education than the local (mostly Tibetan) population (note that the schooling rates of local Tibetan migrants from within the TAR would have been reflected by the rural rates in the 2009 survey). The 2005 survey results of Ma and Lhundrup (2008, 24) show that 53.7 percent of Han temporary migrants had a junior middle school level of schooling, although it is also clear from their data that public employees were severely underrepresented in their sample (e.g., see page 26). Out-ofprovince temporary residents occupying state-sector employment would mostly tend to have tertiary levels of education, as required for such forms of employment. This extrapolation is further supported by another discrepancy revealed by the 2010 census data: if the levels of schooling are measured in a noncumulative manner, that is, the proportion of the population that has obtained up to a particular level of schooling not including those that have obtained higher levels, then in almost all cases we observe a tapering off at each successively higher level of schooling, that is, there is a smaller proportion of the population with a tertiary level of schooling than with a senior secondary level of schooling. This applies to all provinces in the 2010 census besides Beijing, Shanghai, and the TAR. In these three, the

264

Chapter 6

proportions of the population with a tertiary level of schooling exceeded the proportion with a senior secondary level of education, for example, 4.4 percent of the population of the TAR had a senior secondary level of education whereas 5.5 percent had a tertiary level. While this result is understandable in Beijing and Shanghai, which have by far the most educated populations in China (e.g., 32 percent of the population of Beijing, and 22 percent of Shanghai, had a tertiary level of education in the 2010 census), the result is problematic in the TAR given that it has by far the least educated population in China (in terms of formal schooling rates). In every other western province including Qinghai, the non-cumulative tertiary rate is substantially lower than the senior secondary rate (except in Xinjiang, where it is almost on par, probably for the same reasons as the discrepancy in the TAR). The discrepancy in the TAR could be due to a remarkably high degree of retention of students from the senior secondary to the tertiary level within the TAR, leaving relatively fewer with a senior secondary level as their final qualification. However, this is highly unlikely given the very poor retention between the primary and junior secondary levels, as discussed above, as well as the substantial lag in the per capita provisioning of schools and teachers in the TAR relative to other provinces, as discussed with respect to table 4.1 in chapter 4 (and in strong contrast with Qinghai). Rather, the much more likely explanation is that the census data is skewed at the upper end by the inclusion of out-of-province, mostly Han migrants whose levels of schooling would be disproportionately concentrated at the tertiary level, especially among those working in the state-sector such as cadres. The implications of these observations for the TAR are quite serious. They appear to lend support to the argument made in the previous chapter that, since 2003 (after which the TSY stopped reporting the minzu decomposition of state-sector employment in the TAR), the bulk of the coveted and privileged state-sector employment had been filled by out-of-province temporary residents (often on short tours of duty), particularly given that such employment generally requires a tertiary level of qualification. The data also appear to suggest that there has been relatively little effort made by the government to educate and train the permanent-resident (and mostly Tibetan) local labor force to fill the expansion in state-sector employment since 2003, at least not in proportional population terms and certainly not in comparison to the very strong drives at the senior secondary and tertiary levels in other western provinces. Indeed, as noted above, this situation in the TAR is probably quite different from the Tibetan areas in Qinghai in terms of the increase in numbers of Tibetans accessing tertiary education, although the schooling rate of all Tibetans in China (more than double the Tibetan population of the TAR) was almost identical to, if not slightly worse than, those of the TAR in the 2000 census.17 In other words, even if



The Education-Employment Nexus of Urban Exclusion in Tibet 265

the dynamics of change might have been quite different among Tibetans in Qinghai from those in the TAR during the 2000s, the starting point as of the year 2000 was more or less the same, and comparable dynamics of employment exclusion also occurred in these other areas, as analyzed in more detail below. In addition to this analysis of changes over time, several other points are worth emphasizing in table 6.3 with regard to static interprovincial comparisons, with further elaboration in table 6.4, which decomposes the provincial averages by city and rural differences (the town category is left out). As with the illiteracy data, the proportions of the population aged six and older with a primary level of schooling or above (or inversely, with no schooling) were quite similar across rural and city residents, much more so than in the other provinces, and were quite low in absolute terms (explaining, for instance, the high level of city illiteracy in the TAR). Regarding the secondary level, a rural Sichuanese person in 2009 was more likely to have a junior secondary level of schooling than a TAR city resident. Interestingly, the junior secondary rate in rural Qinghai was also much lower than the average and city rates of Qinghai, more in line with the TAR than with the other western provinces shown, suggesting a similarly stark polarization within Qinghai as between the TAR and Sichuan. The fact that Tibetans in Qinghai were even more rural than in the TAR in the 2000 census implies that they would have been mostly reflected by these rural rates and probably also pulled down the rural averages for the province. In the TAR, it was only at the senior secondary and tertiary levels where TAR city residents outperformed their rural counterparts in Sichuan, but only because these levels of education are heavily concentrated in urban centers in China (and thereby also act as important conduits for urbanization, removing people with tertiary schooling from rural areas). Nonetheless, a city resident from Table 6.4: Levels of schooling among the population aged six and older in the 2009 survey by rural and city residency (town residency not included) 6+ population with a level of schooling of at least: Rural

City

Prim.

J. Sec.

S. Sec.

Tert.

Prim.

J. Sec.

S. Sec.

Tert.

TAR

61.6%

16.8%

3.9%

1.1%

65.6%

34.4%

19.5%

7.2%

Qinghai

81.7%

29.8%

5.4%

0.7%

96.1%

77.2%

47.7%

25.2%

Gansu

81.7%

41.0%

8.6%

0.9%

94.3%

73.4%

42.8%

19.0%

Sichuan

88.3%

41.8%

7.7%

1.0%

96.7%

74.3%

42.4%

21.8%

China

90.5%

53.2%

9.7%

1.5%

97.0%

80.1%

44.5%

20.3%

Source: calculated from CPEY (2010, tables 2-25 and 2-27).

266

Chapter 6

Sichuan was more than three times as likely to have tertiary education than his or her city Tibetan counterpart in the TAR. Again, these types of asymmetry are simply not observed across any other provinces in China. As a final note, the sharp drop-off at the secondary level in the TAR is particularly important—especially at the senior secondary level—because it is precisely at these levels that the schooling system in most Tibetan areas (including those outside the TAR) switch from mostly Tibetan-medium to mostly Chinese-medium. The timing or degree of the switch (i.e., whether at junior or senior secondary levels, and/or the extent of Chinese-medium used in teaching) varies from region to region and even from school to school, although fluency in spoken Chinese, let alone written literacy, would generally be obtained only at the senior secondary level (unless students are streamed through regular Chinese-medium schools, as is often the case in many Tibetan areas outside the TAR). Hence, referring to the 2009 survey data, the fact that only 5.6 percent of the (permanent resident) population of the TAR had some level of senior secondary schooling could be taken as a rough indication of the percentage of the population with a functional degree of fluency in spoken and written Chinese. The percentage might be slightly higher than this given that some junior secondary or even primary schooling might take place in Chinese (such as in Lhasa), although the effect of this is likely to be marginal. This stands in contrast to a Han Chinese migrant, who would be fluent in Chinese even if illiterate and would probably be functionally literate with at least a junior secondary level of schooling. In other words, not only was the availability of secondary school-educated local labor in extreme short supply relative to other provinces, but the availability of functionally Chinese-fluent labor was even in shorter supply. This point is crucial to understanding the dynamics of urban employment exclusions. Exclusionary Implications of Educational Inequality These educational inequalities highlight a variety of structural disjunctures underlying rapid economic growth, which in turn are exacerbated by the institutional processes through which the bulk of economic activity and employment is mediated in the TAR. In terms of structural disjunctures, educational inequalities between locals and non-local migrants in the Tibetan areas obviously disadvantage locals in competing for urban employment opportunities. This disadvantage is exacerbated by a striking disjunction between the employment demands of growth, which are strongly driven by urban services and large-scale construction projects, and the actual skills among the large majority of Tibetans, most of whom have either no education or only a primary level. For instance, according to the 2009 population survey for the TAR, 37.3 percent of the population aged



The Education-Employment Nexus of Urban Exclusion in Tibet 267

six and older had no schooling and 81.6 percent had either no schooling or primary schooling (slightly down from 2004, when the proportion was 84 percent). The proportion increases to 94.4 percent if the junior secondary level is included, which is perhaps the more appropriate measure given that not much proficiency in Chinese language is acquired at the junior secondary level, at least not in the TAR. Out-of-province in-migration reinforces this disjuncture by saturating the demands for skilled and/or educated labor within the structurally polarized growth model, thereby enabling its extension even while allowing for equivalent skills formation at the local level to remain under-prioritized. Indeed, inter-provincial migration to the TAR during the 2000s took place in the context of a widening, not narrowing, secondary and tertiary education gap with the rest of China. Improvements in local schooling rates at the primary and junior secondary levels would do little to overcome this disjuncture in the short to medium term given that they would take five to ten years or more before impacting the supply of skilled labor (if sufficient numbers of students even manage to pass through the senior secondary and tertiary levels). These structural disjunctures are compounded (or, to a large extent, caused) by the institutionalized patterns of subsidized integration into China, as analyzed in chapter 4. For instance, the sectors that are growing rapidly in the TAR and that demand low skill labor, such as tourism, commerce and other urban services, or construction run by out-of-province companies, are mostly dominated by out-of-province (Han) Chinese firms and thus provide a natural advantage to those who are literate and fluent in Chinese, and/or who are connected to networks outside the province. The dependence on subsidies further enhances this character of growth given that the subsidies are directed through state channels centered outside the region. Consequently, the channels demand labor according to nationally standardized criteria, thereby conferring strong advantage to workers and entrepreneurs with Chinese fluency, Chinese work cultures, and connections to government or business networks in China proper. Conversely, employment creation has been the sparsest in rural-based off-farm industries and services oriented toward indigenous demand, in which Chinese fluency would not necessarily be a competitive factor. The resultant linguisticcultural mode of bias constitutes a crucial way in which the dominance of one nationality (the Han Chinese) structures and conditions disadvantage in such a peripheral setting. Rural employment did increase with the boom in the 2000s. As discussed in the previous chapter with reference to the work of Goldstein et al. (2008; 2010) and Childs et al. (2011), a variety of rural initiatives in the TAR such as the Comfortable Housing Program appear to have stimulated substantial amounts of rural employment and entrepreneurship, particularly since

268

Chapter 6

2006. Such rural (or small town) employment generation has helped to buck the trend of these broader structural disjunctures. Nonetheless, as also argued in the previous chapter, such increases in rural employment in the TAR have accounted for only a minor proportion—or at best half—of the labor transition out of agriculture. The rest would have transitioned to urban areas on a seasonal or permanent basis. Moreover, such rural nonagrarian employment remained much more limited relative to elsewhere in western China, and the employment generated by growth was also much more focused in urban areas, precisely where the linguistic and cultural biases are most dominant. Hence, despite the rural development initiatives, the dilemma remains that skills formation in the local Tibetan labor force must contend with a striking disadvantage in competing with net inflows of migrants from other parts of China who come to take advantage of the heavily statesubsidized economic boom. This disadvantage includes linguistic and what might be termed “cultural skills,” as in the ability to act “like a Han.”18 The 5.6 percent of local (mostly city) residents with a level of senior secondary schooling or above (per the 2009 survey for the TAR) might perform relatively well in such an environment due to their more advanced fluency in Chinese, but the remaining 94.4 percent with poor Chinese language skills are left ill equipped to face the competitive conditions of urban employment, not simply in terms of skills but more importantly in terms of linguistic ability, which serves as a precondition for skills evaluation. The majority of those who remain in rural areas may, however, never have to confront such competitive disadvantages. Rather, it is the upwardly mobile cohorts of rural labor, together with the 81 percent of TAR city residents with a junior secondary level of schooling or less, who are most confronted by these competitive pressures. Thus, exclusionary processes in this context can be seen to operate more intensely with respect to upward movements from lower to middle strata in the regional labor hierarchy, rather than at the bottom of the hierarchy (if that is how farming and herding is to be considered, which is addressed in the next chapter). Expansion in the secondary and tertiary sectors in the Tibetan areas has therefore, to a significant degree, been absorbed by inter-regional migrants, both skilled and unskilled, rather than by local rural migrants or by the less-privileged majority of urban residents, thereby ironically shortcircuiting the whole logic behind urbanization as a means to “modernize” Tibetan labor. The government typically justifies inter-regional migration by arguing that Tibetans will learn by seeing and imitating migrants, and thereby progressively appropriate the new economic spaces opened up by enterprising out-of-province migrants. The classic example recounted to me on many occasions is the Chinese bicycle rickshaw driver in Lhasa who



The Education-Employment Nexus of Urban Exclusion in Tibet 269

in time comes to be emulated by Tibetans, with more and more Tibetans plying the rickshaw service trade, while the rickshaw driver moves up to purchase a taxi, and yet again, within a few years, Tibetans start to be seen driving taxis, emulating the good example set by the migrant.19 However, other scholars personally noted to me in 2005 that Tibetans, who only a few years previously were much more present in the Lhasa taxi business, appeared to have been squeezed out of this service trade, particularly since the beginning of the OWC. As of 2005, only four out of the nineteen taxi or truck companies operating in the Lhasa municipal area were employing Tibetan drivers.20 Along similar lines, Yeh (2003; 2013) discusses at length this narrative of Han technology transfer to Tibetan areas and why it has not worked out as predicted. This situation is difficult, if not impossible, to portray explicitly with the publicly available employment statistics given that these are not broken down by ethnicity and, in any case, they probably do not capture much of informal employment in any case. Nonetheless, these aspects of disadvantage in the lower employment strata are well inferred by the various indicators of educational inequality analyzed in this section, particularly with the understanding that functional proficiency of Chinese language would be acquired mostly at the senior secondary level in most Tibetan areas (versus the situation for a Han Chinese person, who would be fluent in his or her language even if illiterate). This does not imply that such Tibetans are entirely excluded from urban employment, but that their integration into urban employment is conditioned by strong forces of subordination, such as in the lowest skill levels of construction work on Chinese-run job sites. Some Tibetans leverage protected “ethnic” niches, such as servicing Tibetan clientele or supplying Tibetan demand. Otherwise, even lesser-educated Han Chinese migrants (relative to lowland standards) would easily out-compete such Tibetans in comparable categories of employment. We must nonetheless be careful with suggestions that such Tibetans face high levels of unemployment because, as in most developing countries, unemployment is more appropriately understood as a privilege of middle- and upper-class status given that poor people are usually driven by necessity to accept any work, whatever the work and whatever the wage or conditions.21

Exclusion at the Upper End of the Labor Hierarchy While the data portraying unequal literacy and schooling outcomes serve well to portray the disadvantages faced by the large majority of Tibetans, they are ineffective in representing the disadvantages faced by the upper strata of Tibetan labor in the Tibetan urban areas precisely because these

270

Chapter 6

upper strata are already represented as the privileged in the inequality data. In other words, Tibetans with high school or university education were among the 5.6 percent of TAR permanent residents in 2009 with a senior secondary level of schooling or above, or the 1.7 percent with a tertiary level. Employment at this level is essentially dominated by the state sector in Tibetan areas (as local government employees, teachers, or related professional occupations) to the extent that Tibetans with tertiary levels of education more or less expect to be employed in such types of occupations. Zenz (2013) points out that very similar conditions and expectations operate in the Tibetan areas of Qinghai. Hence, exclusionary processes experienced by Tibetans in these upper strata need to be understood relative to the institutional modalities operating within and controlling access to such forms of state-sector employment, among cohorts with comparable levels of schooling attainment and employment expectations. Since the early 2000s, these institutional modalities have played out through the implementation of state-sector employment reforms, that is, the end of guaranteed public employment for graduates (often known as the “job assignment system”)22 and the introduction of competitive systems of recruitment. The implementation of these reforms interacted with continuing political and economic subordination and the resultant linguistic and cultural biases implicit within the wider development strategies, as well as the persistent structural disparities in qualitative educational outcomes at similar levels of educational attainment (which the data do not reveal).23 The synergy of these institutional and structural factors, some changing and others not, effectively reinforced the linguistic and cultural modes of competition and bias within state-sector employment. This has led to disadvantages for even those Tibetans with the highest levels of schooling attainment in the region and who have thus been the best suited to overcome the disadvantages described in the previous section, in that they are the most likely to meet the educational and linguistic attributes of advantage relative to the average Han migrant. Therein lies the fallacy of simple intergroup comparisons, because effectively these educated Tibetans are not competing with the average Han migrant for the same types of employment, but with comparable cohorts of equally, if not more educated, Han migrants. The apparent advantages of educated Tibetans with respect to lower-skilled employment, as indicated by the comparisons of schooling attainment, are irrelevant because these Tibetans do not target those lower employment strata. Rather, exclusion operates in the employment strata that they do target (i.e., state-sector employment) through similar modes of bias as those operating in lower strata (i.e., quantity and/or quality of educational outcomes and linguisticcultural biases), albeit through different institutional mechanisms.



The Education-Employment Nexus of Urban Exclusion in Tibet 271

The Erosion of Tibetan Representation in Public Employment in the TAR One example of such processes has already been discussed near the end of the previous chapter on employment transitions regarding the sudden and sharp shift away from Tibetan representation in state-sector employment in the TAR between 2001 and 2003. For the sake of recollecting this important point, the number and share of Tibetan staff and workers in state-owned units fell between 2001 and 2003, while the number and share of nonTibetans rose. Calculating from TSY (2004, table 4-5), the share of Tibetans in total staff and worker employment in state-owned units fell from 71.3 to 64.6 percent, while that of non-Tibetans rose from 28.7 to 35.4 percent. At the cadre level, which accounted for two-thirds of permanent state-sector employment in 2003, the change was even sharper: overall cadre employment increased from 69,927 cadres in 2000 to 88,734 in 2003, while the number of Tibetan cadres fell from 50,039 to 44,069, or from 72 percent of total cadre employment to just below 50 percent. In sum, these changes effectively resulted in a substitution of permanent Tibetan employees with non-Tibetans (mostly Han). The shift in 2003 revealed a sudden move away from Tibetan representation in urban public employment, that is, from the most privileged and formalized forms of employment in the TAR, and non-Tibetan cadres outnumbered Tibetan cadres for the first time since 1980. Government assertions that Tibetans were the dominant beneficiaries of increasing statesector wages, thereby contributing to an emerging “middle class” of Tibetans,24 became much more tenuous at this time, after which the government stopped publishing this particular disaggregation of employment data for the TAR.25 Rather, Tibetan employment was shrinking during these early years of the OWC in precisely the parts of the economy that were growing fastest, that is, the urban state sector. The processes underlying this sudden shift in the ethnic balance of statesector employment do not necessarily imply explicit discrimination; they might be as much circumstantial as intentional. For instance, the reductions in the TAR may have been targeted at employment categories with higher representations of Tibetans, such as loss-making state-owned enterprises, without explicitly intending to discriminate against Tibetans per se. Similarly, the increases in employment may have been targeted at categories where Chinese representation is the highest, such as in departments related to the construction of the Qinghai–TAR railway or in security-related government posts. Nonetheless, the result was disproportionately borne by Tibetans in the TAR, as these data (perhaps inadvertently) reveal. As such, the outcome was effectively discriminatory in the sense that it resulted in discriminatory outcomes regardless of whether discrimination was specifically or primarily intended. This is all the more relevant

272

Chapter 6

because the restructuring of state-sector employment was likely the result of generalized national initiatives by the central government to restructure, rationalize, and standardize state-sector employment.26 “Rationalization” refers to the implementation of means-based entry requirements for public employment, or else ending unnecessary duplication in administrative or even productive processes. Standardization refers to bringing employment, management, accounting, contracting, production, or other standards in line with the rest of the country in order to facilitate regional and national operations. As with most major policy decisions stemming from Beijing, these initiatives were implemented with reference to the core regions of China, according to criteria determined to be external to the Tibetan areas. This tendency is reinforced by the intensified reliance in the Tibetan areas on external flows of capital and labor for the upper layers of the local economy, as analyzed in the two previous chapters. Externalized dependence in turn feeds back into demands for increased rationalization and standardization within the bureaucratized modes of accumulation due to concerns from the center regarding accountability and in order to facilitate the development strategies implemented by the center and/or non-local firms.27 Indeed, principles of standardization might be applied even more stringently in the TAR than in other parts of China, where local governments are more likely and able to subvert such principles to their own needs and agendas, given the massive relative power of the central government and non-local state-owned enterprises in the local economy of the TAR. Similar to the exclusionary dynamics at the lower end of the labor hierarchy, these policies effectively discriminate by way of educational disadvantages. The group with less educational advantage (i.e., in terms of quality of education, status of degree and schooling, or linguistic ability) loses out in the shift toward principles of greater procedural equality. Furthermore, the implicit dominance of one group (Han Chinese) within rationalization and standardization accentuates this imbalance through linguistic biases that are imposed in selection criteria (e.g., use of Chinese in examinations) or cultural networks that informally underlie the practice of power in political and economic units (Ch. guanxi). For example, exams for permanent public-sector employment, which were introduced in the TAR and other Tibetan areas in the early to mid-2000s, are conducted mostly (or in many cases entirely) in Chinese, per the national standard.28 Many local Tibetans and foreign observers in the TAR whom I interviewed between 2003 and 2007 noted that this placed Tibetans competing for such positions at a distinct disadvantage, particularly those who might have excellent Tibetan language skills (and are thus ideally suited for governing Tibetans), but only average Chinese language skills.29 Employment reforms have similarly intensified competition in public employment throughout China.30 However, as analyzed in the first section,



The Education-Employment Nexus of Urban Exclusion in Tibet 273

the situation in the Tibetan areas is unique due to the educational asymmetry between locals and migrants, whereby urban Tibetan locals have much worse education levels on average than incoming Han migrants, even Han migrants emigrating from rural areas. Thus, in the context of increasing migration from higher- to lower-educated regions, large educational lags combined with inferior cultural or social connections to regional centers result in severe disadvantages for locals competing for public employment, including the most educated urbanites. This is the inverse of the norm in most areas of China, where urban residents benefit from a natural form of protection in coveted upper-strata employment given their higher education levels in comparison to rural or western migrants (on average). As a result, reforms that promote more merit-based selection procedures for public employment elsewhere in China tend to privilege local urban residents. In these situations, migrants mainly intensify competition at the lower levels of the labor hierarchy. This is a problem for the erstwhile privileged urban working classes facing layoffs (Ch. xiagang) from more secure forms of urban-unit employment, especially state-owned unit employment, but it is much less of a problem for the new educated elites.31 In addition, urban governments elsewhere in China have had more autonomy to protect or promote local skilled labor.32 Tibetan areas have had very little autonomy in this regard given their particular situation of political disempowerment. The exceptionality of these educational and political asymmetries reflects the observation made by Emily Yeh that Han migrants in Lhasa consider themselves to be of higher “quality” (Ch. suzhi) than local urban residents, in contrast to everywhere else in China where migrants are generally considered to be of lower quality than local urban residents.33 Quantitative literacy or schooling rates in this sense underestimate the qualitative disadvantages faced by Tibetans in their increasingly sinicized urban economies, given that they include those who are literate only in Tibetan. In other words, one can find very literate Tibetan elites with limited Chinese skills (more so in the TAR, less so in Qinghai or other areas of eastern Tibet). Indeed, if knowledge of Tibetan were a prerequisite to public employment in the TAR, Tibetans would have an obvious advantage over Chinese in finding employment in the public sector. This argument influenced the policies of the 1980s, when the CCP leadership emphasized ethnic competence, based on the idea that local people knew best how to implement party policy. However, under current policies, non-hegemonic groups (Tibetans and Muslims) are left with the choice to assimilate or be excluded from privilege, yet their capability to assimilate is inhibited by their severe lags in quantity and quality of education. Ultimately, such policies further encourage the external orientation of the human resource system given the strong demand for Chinese-fluent labor and the short supply

274

Chapter 6

of such labor among the local population, particularly at levels of fluency that would conform to the expectations of out-of-province employers. Educational inequality need not have produced such results if the government remained committed to policies of proportional ethnic representation in public employment, as had been the earlier reform consensus. Tibetan representation in state-sector staff and worker employment in the TAR reached a peak of 89 percent in 1982, just after Hu Yaobang encouraged Tibetanization of government employment and permitted the departure of a large contingent of Han cadres. This high level was quickly reversed the following year and fell back down to around 70 percent, where it remained throughout most of the 1980s and 1990s. The change in 2003 was therefore unprecedented throughout the entire reform period and marked a distinct departure from the implicit institutional compromise between Tibetan elites and the state in the first two decades of reform. As such, it reflected a fundamental shift in the dominant paradigm of minority-area governance in China, at least with respect to Tibetan areas. The timing also suggests that the sudden shift was linked with the new leadership of President Hu Jintao in Beijing. As party secretary in the TAR during the Lhasa uprisings in March 1989, Hu ordered martial law and was associated with the end of the relatively liberal period of the 1980s. The shift cannot be justified in terms of a deficiency in Tibetan capabilities given that Tibetan education levels were much worse on average in the 1980s and 1990s than in the 2000s, although it is plausible that the government viewed the composition of state employment in the early 2000s as inappropriate for its development plans under the OWC and beyond. Thus, even though these shifts in state-sector employment were obviously related to national policy trends, the implementation of labor market reforms in Tibetan areas reinforced the widespread perception that the Chinese leadership does not trust the loyalty of Tibetan cadres and is antagonistic to even the limited forms of political space that had been maintained up to the early 2000s. As noted in the previous chapter, we have no idea of these trends since 2003 because this particular minzu disaggregation of the staff and worker data was discontinued after TSY (2004) and replaced by a sex disaggregation. Since 2003, staff and workers in state-owned units in the TAR increased from 136,646 people in 2003 to 170,748 in 2008, and by 2010 there were 207,267 employees in state-owned urban units (out of a total of 222,116 employees in all urban units). We do not have any information of the degree to which the increase was filled by local Tibetans versus non-Tibetan out-of-province migrants. However, given that the end of guaranteed employment for graduates, as described below in the case of Qinghai, was apparently implemented in the TAR in 2006, in addition to the wage inequality trends analyzed in the previous chapter, it is very likely that similar exclusionary dynamics were at work later in the decade, as also



The Education-Employment Nexus of Urban Exclusion in Tibet 275

suggested by various qualitative reports. If so, these dynamics would definitely provide much insight into the outburst of protests that took place in Lhasa and elsewhere in March 2008. Employment Reforms in Qinghai The Tibetan areas in Qinghai provide another particularly illustrative case of these dynamics related to the implementation of the national reforms of state-sector employment. It illustrates similar exclusionary processes at the upper strata of Tibetan employment as those in the TAR despite the absence of the more extreme particularities of the TAR. According to my discussions with a variety of officials, scholars, teachers, and students,34 the phasing out of guaranteed graduate employment that started elsewhere in China in the early 1990s had been planned in Qinghai since 1997 and was implemented in 2001 for university students and in 2002 for high school students. Prior to this time, employment for graduates had been more or less guaranteed by either county governments or various work units (i.e., a medical student would be given a job placement by the health department, a student from the “normal” teachers’ schools by the education department, etc.), albeit students could opt out if offered private-sector employment. Employment allocation since 2001 became more competitive and thus purportedly more meritocratic and rationalized. Each work unit has a fixed number of posts and hires through open competition.35 The delayed implementation of these reforms was in line with the typically gradualist manner of applying national reforms to the western or poorer provinces more slowly than in central or coastal provinces, and in deference to discretion at the provincial level.36 Those affected most by the reforms tended to be graduates with a medium quality of senior secondary or tertiary education. For instance, a senior Han official from Malho Prefecture noted that the bottom range of students tends to drop out as early as lower middle school and return to work in farming or herding, or else engage in self-employed trades or small-scale business. The top students (mostly Han) tend to leave the area, either to universities elsewhere in China or to jobs in larger cities such as Xining. The next top students, those who might not have succeeded in entering universities outside the province yet who nonetheless have a strong provincial university education in the mainstream (Han) system, tend to return to the area and compete relatively well in the new labor market. The middle range of students, particularly those from the minority high schools and universities, is squeezed by these new competitive conditions. This range represents the bulk of Tibetan (and Muslim) students, especially those from rural areas, who are attracted (or restricted) to the minority schooling track either because of a choice to study in this stream (because

276

Chapter 6

of greater Tibetan-medium content, for instance) or because of the lower fees and typically lower entry requirements and competitive conditions in the minority system. As background, a system of minority education exists in the Tibetan areas outside the TAR, from primary up to tertiary levels, parallel to the mainstream Han schools.37 The minority education system is largely viewed (in the eyes of the state) as a means to gradually phase the assimilation (or “acculturation”)38 of Tibetan students into a Chinese-dominated system and there is a humanities bias in minority education that derives from this official attitude.39 Nonetheless, despite this assimilationist state intention, Zenz (2010; 2013) carefully documents how there has been considerable expansion in both the offering of Tibetan-medium streams and Tibetanmedium content within these streams, for the most part led by the initiatives of Tibetan educators in Qinghai since the 1980s at the tertiary level and since the early 2000s at the secondary level. Some initiatives have also taken place in the Tibetan areas of Gansu and northern Sichuan, suggesting a strong Amdo connection. However, the question remains whether these ground-up initiatives represent an institutional momentum from a previous period that will have difficulty being sustained under the current conditions. Zenz (2013) suggests that the initiatives at the tertiary level had apparently survived the 2010 education agenda of the Qinghai government, which was aimed at increasing Chinese content in the minority system as low as the primary level (officially referred to as “bilingual education”). In particular, the Tibetanization initiatives have been in tension with the demand for increased Chinese content from many quarters due to the competitive pressures induced by the employment reforms. Indeed, these tensions were brought to the fore with the 2010 protests in Qinghai against the government’s education agenda. Moreover, subjects that would be useful for the transitional context in these Tibetan areas, such as vocational, managerial, or scientific education, are undersupplied in the minority system. These issues were not pressing problems in the past so long as graduate employment was guaranteed, but under the new competitive system of employment the value of degrees from minority schools has been depreciated both by the mainstream school system and by employers, given that competition places a premium on the perceived quality or status of schooling (defined by the standard—especially linguistic standard—of the mainstream Han culture).40 Thus the two-tiered education system, which since the early 1980s had been an important part of the political compromise offered to minority elites in minority autonomous areas, was increasingly perceived by the state in the 2000s as becoming redundant with respect to addressing the changing labor market conditions, even despite the impressive development of Tibetan-medium education by Tibetan educators.41



The Education-Employment Nexus of Urban Exclusion in Tibet 277

It is important to note that many of these upwardly aspiring Tibetan youths possess among the highest levels of education in their respective communities and come from families that can afford such levels of education, particularly at the tertiary level, where tuition and related fees are quite expensive relative to average rural per capita incomes in these areas.42 In the case of rural students, many are the first of their household to reach a secondary or tertiary level of education and they therefore carry their own as well as their family’s aspiration to join the relatively elite workforce of public employees. The senior official mentioned above admitted that it was precisely this middle range of students that experienced the most dislocation under the new system, given that they had difficulty competing for higher-status jobs but they did not want to work in lower-status jobs, particularly in jobs considered unworthy of a senior secondary or tertiary education. These perceptions were confirmed to me repeatedly in numerous interviews and conversations with scores of Tibetan high school and university students or recent graduates from the minority education system, who consistently stressed that it was very difficult to obtain appropriate employment that corresponded to their level and type of education. Interestingly, a similar tension exists between Tibetans in the remote pastoral areas of Golok or Yushu in Qinghai, where better-educated non-local Tibetans from the farming areas of Rebgong, Chentsa, or Xunhua squeeze out less educated locals in competition over local public employment. Much local conflict in these remote areas pertains to local and non-local Tibetans, symbolized by the divisions between “herders” and “farmers.” Non-local “farmers” tend to be advantaged in competing for public employment in these areas given that they tend to be better educated, more sinicized, and better connected than locals due to the proximity of their home counties to Xining.43 Thus, in the context of increasing migration, large educational lags combined with inferior cultural or social connections to regional centers result in severe disadvantages for local Tibetans in competing for public employment, educated urbanites included, unless they are already part of one of the few elite families dominating such positions.44 This is not to mention the increased competition from already-resident and better-educated local Han, many of whom have been born and raised in the region and aim to find stable employment there. Zenz (2013) also notes the new phenomenon since the late 1990s of large numbers of students from eastern China coming to study in the universities of Qinghai—including the minority universities—in large part because of the less competitive conditions relative to eastern China. Some of these students then subsequently compete for local employment in the province, again with relative advantage compared to locals, even though they would have been disadvantaged in their places of origin relative to the much more competitive standards in both education

278

Chapter 6

and employment. Presumably the same processes are occurring in the TAR, as suggested by the disproportionate numbers of Han Chinese students in the Tibet University in Lhasa. In this context, many of my interviewees commented that wealthy Tibetans, and in particular Tibetan cadres, are increasingly sending their children to mainstream Chinese schools, which have higher entry criteria and are more expensive than minority schools (albeit even minority high schools and universities are very expensive for the average rural Tibetan). Those who cannot afford the mainstream schools are left with the reinforced impression that secondary and tertiary minority education is of diminishing value, given that it no longer helps to secure coveted public-sector jobs and is thus no longer worth the investment. As a result, in the face of increasing tuition fees and the chronic unemployment of graduates, many poorer rural families are opting out of secondary education or, at the very least, placing it at a lower level of priority within their household livelihood strategies.45 Therefore, the issue of educational inequality has increasingly become intra- as well as inter-ethnic. These various disjunctures present a catch-22 between the choices of integration or segregation, both of which ultimately place educated Tibetans at a disadvantage within the context of intensified competition over limited employment opportunities. In recent years, there has definitely been a strong and self-aware push from within Tibetan society to educate Tibetan young people. However, Tibetans educated in Chinese and in mainstream schools are disadvantaged within these schools on linguistic and cultural grounds, as well as by less supportive conditions for competing in school.46 Yet, while the minority education system is more accessible to Tibetans, it is less adapted to the needs of an increasingly competitive commercial economy, especially in linguistic terms, which leaves graduates at a strong disadvantage after graduation compared to graduates from the mainstream schools. This has particularly been the case since the employment reforms, which intensified competition over the limited employment opportunities available to the graduates from both systems. Zenz (2013) problematizes this scenario further by noting that those Tibetans who opt for the mainstream Chinese-medium education system might end up with stronger Chinese language skills and more options for employment, but they also face more intense competition within these expanded options, in comparison to graduates from the minority system who face less competition within a more limited range of options. This insight leads Zenz (and many of his informants) to question whether the mainstream Chinese medium route has indeed turned out to be the more privileged one (at least, so long as certain niches of public employment continued to be reserved for graduates from the Tibetan-medium tracks, which has become an increasingly tenuous condition, as Zenz himself points out).



The Education-Employment Nexus of Urban Exclusion in Tibet 279

Furthermore, since the reforms, official government employment strategies for unemployed high school and university graduates, as well as for laid-off staff or workers, have largely focused on encouraging self-employment or small private business start-ups. For instance, during a visit to Chentsa in June 2004, I interviewed one young unemployed male Tibetan university graduate who was attending a county-run workshop on how to set up and run a small business, based on ILO course material on this subject translated into Chinese. Attendance at these workshops was required in order to receive unemployment benefits of 150 yuan a month. This system of welfare conjoined with training was new, coinciding with the end of guaranteed employment. Within the interethnic context of Qinghai, this strategy effectively promotes a downward move into stigmatized intermediate positions of the labor hierarchy, such as in private urban services and commerce. In particular, these intermediate positions in Tibetan areas are precisely those that were traditionally and newly dominated by Muslims. In light of Tibetan stigmatization of these types of activity and of Muslims, as discussed in the next two chapters, these programs likely added insult to unemployed status or upward aspirations by implying that such relatively privileged Tibetans should rub shoulders and compete with Muslims over common types of activity deemed the domain of the uncultivated. Combined with the more general disadvantages faced in obtaining employment, such experiences create a powerful fusion of feelings of discrimination, dislocation, and indignation. Indeed, in recent years, the fusion has often mutated into forms of anti-Muslim activism by Tibetans in Qinghai, as discussed in chapter 8.47 The combination of all these factors has increased insecurity among those rural and urban households that continue to base their livelihood strategies on the education and subsequent employment of one or two children in the public sector. The situation has also intensified out-migration from these Tibetan areas. For instance, five different rural families that I interviewed in Qinghai in summer 2004 were planning to send one of their teenagers to India even though the teenagers were already successfully progressing in their secondary schooling. All five families explained that this was due to the cost of education. When I noted that the cost of illegal refugee migration to India was even greater, two answered that “the Dalai Lama” was taking care of the students in India through his schools, while three replied that the county-level secondary education was worthless for securing employment after graduation. Four of the families also noted that the minority education system was causing their children to lose their culture. While these views are not necessarily representative and express biases that might be poorly founded, they nonetheless capture the sense by which many perceive and respond to the catch-22 between integration and

280

Chapter 6

segregation. More generally, the impasse reinforces the impression that development has been undermining Tibetan society, particularly through the assimilation of its most educated youth into mainstream Chinese society while at the same time marginalizing them from what are deemed to be respectable employment opportunities. These dilemmas, which were described to me in a very similar manner by most of the Tibetan teachers, scholars, and officials I interviewed, led most of these respondents to the logical conclusion that increased autonomy to determine employment policies at the county and prefecture levels is required in order to reverse such disadvantages in local employment. This should be done in ways that are in accordance with the China National Minority Regional Autonomy Law (adopted in 1984). For instance, Tibetan language proficiency could be made a requirement for employment in local governments, which would be an affirmative action stipulation permitted under the 1984 law.48 Ideally, this would be supported by an integrated regional Tibetan education system and curriculum, consolidated across all the Tibetan areas in China in order to achieve a scale necessary for institutional self-sufficiency. Notably, in explaining these solutions, my respondents explicitly or implicitly echoed the advocacy of several Tibetan leaders inside Tibet, particularly the late Panchen Lama, who argued in the 1980s for precisely these applications of the National Minority Law as a means to support effective rather than merely nominal minority nationality autonomy in China.49 They also echoed the ongoing advocacy for increased Tibetan autonomy by the Dalai Lama, including his emphasis on the importance of modern secular education for preserving Tibetan culture.50 In the course of my fieldwork in 2004, one event occurred that is representative of such advocacy. On 21 September 2004, a demonstration of at least 200 mostly Tibetan students started outside local government offices in Dawu, the prefecture capital of Golok in Qinghai. The protest was peaceful and was supported by local officials through the provisioning of tents and food (RFA 2004). Based on my own further inquiries with Golok Tibetans in Xining, the protest continued into November. The purpose of the protest was to demand employment, particularly considering that the students’ families had been encouraged to pay high college fees with the promise that their children would secure good jobs after graduating. The students therefore claimed that officials had failed to deliver on these promises and were instead giving jobs to outside candidates brought into the area. They demanded that local applicants be given preference (RFA 2004). The contestation underlying this event refers precisely to the employment reform in 2001, which was implemented while these students were studying, as well as to austerity in public employment over the same period. A similar event took place in Lhasa in 2006, in conjunction with the beginning of the labor reforms in that province. According to RFA (2006),



The Education-Employment Nexus of Urban Exclusion in Tibet 281

a group of graduates protested after only two of one hundred public jobs where given to ethnic Tibetans following an open exam competition (conducted in Chinese). Ironically, in the face of marketizing labor reforms, these events appear to signal the predilection of relatively elite Tibetans within China to advocate for a return to earlier Socialist legacies of labor market regulation as a means for assuring preferential and representative treatment in public employment.

Conclusion: Dislocating Hierarchies and Nationalism This chapter focused on exclusionary processes in both the lower and upper strata of urban employment in the Tibetan areas. The first section analyzed the lower end of the labor hierarchy through an analysis of education differentials between locals and migrants in the TAR and how this has led to both structural and institutional disjunctures between the employment demands of growth and the actual skill levels among local Tibetans. Skill levels in this sense are represented by literacy and schooling rates, interpreted with the implicit understanding that Chinese fluency is generally obtained in Tibetan areas only at the senior secondary level and above, with some exceptions. Polarized dependent growth further exacerbates these disjunctures given that opportunities in the dynamic sectors of the economy have been increasingly instituted along lines that confer advantage to linguistic or cultural attributes deriving from the dominant group, including Chinese fluency, Chinese work cultures, and connections to government or business networks in China proper. Indeed, these attributes are some of the important cultural dimensions of how education confers advantage within a setting of occupation, cultural hegemony, or disempowered development, in addition to its role in generating the appropriate skills required by changing structures of employment. Exclusion can also be observed at the upper end of the labor hierarchy, among Tibetans with senior secondary schooling and above and who are thus the most likely to meet the demand for higher skill as well as the cultural attributes of advantage. This was discussed in the second section with a focus on the TAR and Qinghai. The fall in Tibetan staff and worker employment in the TAR between 2001 and 2003, despite the public sector’s being one of the fastest-growing sectors of the economy, was discussed. Whether or not this was intended discrimination, it took place through the increasing institutional emphasis on rationalization and standardization in public employment that accompanied the OWC across western China. This resulted in a notable disadvantage for elite Tibetans in these coveted areas of employment, again due to qualitative educational and linguistic modes of bias, among

282

Chapter 6

other probable reasons such as political distrust for Tibetan cadres or simple prejudice. Similar processes were analyzed in the case of Qinghai, with a specific focus on the phasing out of the system of guaranteeing public employment to high school and university graduates in 2001–2002 (the dynamics of public employment in the TAR may also have been related to this policy). Hence, while originally devised to offer preferential treatment to minorities, the changing nexus of minority education and public employment systems in the Tibetan areas has increasingly evolved toward a catch-22 between assimilation and segregation for relatively elite or upwardly mobile Tibetans. On one hand, Tibetans have been educationally disadvantaged in the mainstream system due to linguistic or cultural biases, or even economically disadvantaged due to the higher costs of such education compared to the minority system. On the other hand, the segregated option of minority education has ultimately led to later disadvantages in urban employment in the context of the employment reforms, given that the reforms effectively resulted in a devaluing of minority education by the dominant (Han Chinese) group that controls most urban employment opportunities that are accordant with a tertiary level of education. This imbroglio has been accentuated by the fact that employment reforms curtailed guaranteed or preferential employment in favor of more meritocratic principles (defined according to the dominant group), combined with intensified inter-group competition over employment opportunities due to increased migration. Improving schooling outcomes at the primary and junior secondary levels does little to address this institutional disjuncture between education and employment systems; if anything, it exacerbates it by increasing the supply of young Tibetans entering into senior secondary and tertiary education. Indeed, the end of preferential employment practices risks accelerating the redundancy of the minority system precisely at a time when education policy is pushing for an increase of enrollments and education outputs within this same minority system.51 Within this context, urban employment exclusion increasingly pivots on axes determined by the changing qualitative aspects of education demanded by urban employers, such as language ability or status of degree. The impasse highlights the very valid need to design educational strategies in conjunction with urban employment strategies, along with the need for effective political representation and decision making by minority groups within these designs. Notably, this analysis of exclusionary processes has been focused on structural and institutional disjunctures that have emerged from the heavily subsidized, externalized, and polarizing development strategies implemented within a context of open migration, rather than on specific agentive practices of exclusion, bias, prejudice, or outright discrimination practiced by Chinese toward Tibetans. This focus is not meant to underplay



The Education-Employment Nexus of Urban Exclusion in Tibet 283

these latter considerations, which obviously play a huge role, even if their discussion remains taboo within China. Indeed, research into practices of discrimination in minority regions such as Tibet and Xinjiang is sorely needed in the literature. However, the focus here is meant to offer some insight into how effective discrimination can emerge from the exclusionary consequences of various structural and institutional dynamics, regardless of the intentions or prejudices of the various actors involved. Moreover, as pointed out in chapter 5, this is not to say that all Tibetans are excluded, or that none benefit. From a poverty perspective, living standards have definitely improved, although the polarization that underlies these marginal improvements in wealth or poverty reduction simultaneously exacerbates dislocation and insecurity across the social hierarchy, particularly in the middle strata. Indeed, from the perspective of a largely agrarian workforce, much of urban labor can be considered middle strata. Urbanizing rural labor can be best understood as upwardly mobile, attempting to enter these middle strata. Thus, similar to a Polanyian notion of double movement, as discussed in the introduction, dislocation happens in a broad cross-class manner. Local Tibetan cadre elites are increasingly threatened by policies of rationalization, which pits them and their descendants against highly educated and politically connected young cadres from elsewhere in China. Tibetan businesspeople are threatened by migrant traders from elsewhere in China. Rural migrants and the urban poor are squeezed out of desired employment opportunities by outside migrants and into less desirable ones. In contrast, perhaps the least dislocated by these exclusionary dynamics are either the very politically connected or else those who choose to remain in agriculture. Whether or not we should consider farming and herding as the lowest strata of the labor hierarchy is discussed in the next chapter, although these activities definitely produced the least value-added per employed person up to 2010, as analyzed in the previous chapter. In other words, we do not necessarily see the most exclusion in the worst performing sectors of the economy, nor among those who we typically identify as the poorest. This emphasis on exclusion and dislocation at the middle and upper end of the labor hierarchy is important from the perspective of conflict given that such exclusion is usually the most politically sensitive, even if it is not necessarily the most impoverishing or disequalizing. Exclusionary processes are particularly significant for Tibetan elites given the potential loss of both position and dignity. Many of these elites who have positioned themselves as junior partners within Chinese rule—or have aspired to occupy this position—have undoubtedly been better placed than many commoners to weather the dislocations caused by recent developments. However, the relative employment security that they managed to maintain up to the 1990s has since been considerably eroded, with no guarantee of similar security

284

Chapter 6

for their progeny. Instead, the future presents an omen of downward mobility for many in the local social hierarchy despite rising schooling levels and rising aspirations of upward mobility among younger generations. Government strategies of self-employment and small business start-ups potentially irk the dignity of these once-privileged Tibetans, as well as that of upwardly aspiring graduates, if not openly insulting them by suggesting that, after all their education and/or privilege, they should compete shoulder to shoulder on an equal level with Muslims and low-cultured Han migrants in a variety of socially scorned activities, as discussed in the next chapter. Thus, precisely as the state made stronger efforts to expand education, it also moved toward a stronger assimilationist position together with strong exclusionary pressures in labor markets. Ultimately, the situation breeds considerable frustration and alienation, while the additional elite option of full assimilation further accentuates class polarization among Tibetans themselves. The seemingly dead-end scenario reinforces the belief among many quarters of both common and elite Tibetans that the only solution is found in nationalist (although not necessarily secessionist) politics, that is, increased local control over education and employment in order to create the conditions whereby Tibetan-prioritized education can be supported by Tibetan-prioritized employment. In other words, within these modes of integration predicated by the assimilating drives of the state, there is a tight synchronicity between experiences of exclusion and contestation. These experiences are most notable at the middle and upper levels of the urban labor hierarchy, that is, among those who have been able to afford secondary and tertiary education (which mostly takes place in urban areas), or who can afford unemployed status (or to support an unemployed family member in an urban area). In contrast, perhaps those least faced by exclusion are those who have decided to remain in rural areas and to tie their fortunes to the fate of agriculture, and who are typically associated with being among the poorest in contemporary China. While such poverty is deserving of our normative attention, exclusion among the non-poor is deserving of our analytical attention precisely because of its importance in determining the trajectories of various socio-political processes such as resistance and conflict, as discussed in the last two chapters.

Notes   1.  As briefly discussed in the introduction, “institutionalist” here refers to the modalities of social ordering (i.e., the instituting of social order), per sociological or historical institutionalist approaches (see Hall and Taylor 1996). The definition of institutions that is popular in the recent political science literature is included



The Education-Employment Nexus of Urban Exclusion in Tibet 285

within this broader understanding, for example, the formal and informal rules and norms guiding regularized social interactions, although the liberal bias within this latter definition is also to be noted, in the sense that violations of rules and norms can play as much of a role in social ordering as their observance.   2.  For further discussion on these theoretical implications, see Fischer (2008a; 2011a).   3.  For analyses of these issues in China, see Qiang Li (2002), Guo and Chang (2006), Lulu Li (2008), and Liu (2008). For an excellent survey, see Roulleau-Berger (2009).   4.  At the time of the final edit of this book, I was asked to review Zenz (2013), which turned out to be a fantastic contribution on this issue. The book deals with very similar issues related to the interaction between education and employment systems among Tibetans in Qinghai, with similar conclusions albeit from a perspective drawing largely from the anthropology of education literature.   5.  See an excellent review of these debates in Zenz (2013).  6. I refer here to “spatial” in the sense implied by Sautman and Eng (2001) or Hu and Salazar (2008). See Yeh (2013) for a critique of these conceptions of spatiality.  7. This government statistic came from a presentation by Robert Barnett on the television content of Xizang TV, which broadcasts mostly in Tibetan, during a conference presentation on the political implications of technical choices in Tibetan television editing and composition, delivered in Vancouver, Canada, in April 2004 (conference entitled “Tibet in the Contemporary World,” 17-21 April 2004)   8.  Zenz (2013) notes that many of his Tibetan informants in Qinghai held the view that even having a high school education was not worth much in terms of career prospects and generally allows only for non-manual, albeit low-paid, servicesector jobs in urban centers such as Xining at best. Obviously, such views and biases must be understood relative to the work expectations of the relatively elite educational cohort that Zenz was interviewing.   9.  According to the 2000 population census, 63 percent the population in the TAR with their place of household registration based outside the TAR were from Sichuan, or 68,496 people out of a total of 108,669 (Tibet Census Tabulation, Table 7-2). Lower shares for Sichuan were measured by Ma and Lhundup (2008, 18) in their survey of 1470 migrants in Lhasa in September-October 2005, of whom 30 percent were from Sichuan and 24 percent from Gansu, including both Tibetan and Han. However, the degree to which their survey was representative of the broader migrant population is not clear. 10. For instance, the schooling rates of both Han and Tibetan temporary migrants in Lhasa in Iredale et al. (2001, 156), based on two surveys in the mid-1990s, and Ma and Lhundrup (2008, 24), based on a survey in 2005, are significantly higher than the corresponding national average rates or the local TAR rates for the same years. The authors do not make this observation; I have drawn the insight from comparison to official statistical sources, as analyzed in this section. 11.  My own observations in Qinghai apply most strongly to Rongwo Town in Rebgong County (Ch. Tongren) and Chabcha Town in Chabcha County, both located in farming regions close to Xining City and serving as cultural and educational centers. However, based on discussions with several Tibetan teachers and scholars

286

Chapter 6

and one Western scholar during fieldwork in 2004, it seems that there was still considerable resistance toward schooling in the more remote pastoral areas of Golok and Yushu. This was due to a combination of very poor conditions in the boarding schools of these remote prefectures, disappointment with the employment outcomes of many recent graduates, and the fact that families preferred to keep their children, particularly their sons, engaged in pastoral activities. Notably, students, teachers, and parents unanimously agreed that, once sent to boarding schools in county towns for several years, particularly at the secondary level, students rarely return to farming or herding. Nonetheless, even in these contexts, overall enrollment rates appear to have increased since the mid-1990s. 12.  In personal communications in December 2007, Melvyn Goldstein provided me age-specific education data from his team’s surveys of three villages in Shigatse Prefecture, TAR. These data definitely showed strong increases in education levels between 1998 and 2006 among cohorts aged 15 to 20 years. This being said, claims such as in Xinhuanet (2006), that efforts to reduce illiteracy in the TAR have focused on the population aged 15 to 50 and that illiteracy in this age group fell from 39 percent in 2000 to 15 percent in 2005, are clearly nonsense according to the official survey and census data presented here. 13.  Presentation made in Oslo, December 2008, and personal communications. 14.  This is one of the important fallacies of educational strategies that push enrollments within poorly funded and dysfunctional school systems without regard for the need for equally, if not more intensive, complementary efforts to improve and increase educational infrastructure. Notably, rural schools in much of rural China have been severely constrained in terms of funding due to the policy of requiring rural areas to be more or less fiscally self-sufficient (e.g., see Wong 1997). In this context, the strategy of simply abolishing tuition fees can be crippling to such schools given that it removes one of their few sources of funding. The strategy can ironically reinforce poor educational outcomes even though it is motivated by the aim to assist poor families in sending their children to school. This is not an argument for tuition fees but, rather, that policy needs to focus on infrastructural supply-side issues as much as if not more than the current mainstream emphasis on service-user demand-side interventions such as conditional cash transfers, vouchers, and other such policy approaches, such as those promoted by WB (2003). 15.  As explained in detail in chapter 3, more than half of the Chinese in the 2000 census were temporary residents and thus they would not have been surveyed in the non-census years. 16.  These rates from the 2000 census are calculated using the urban and rural illiteracy data in CPSY (2001, table 2-12), with the rates properly calculated with the population aged fifteen and older taken from table 2-7. 17.  These data are analyzed in detail in Fischer (2005a, 145-49). In brief, the schooling rates of all Tibetans in China were almost identical to those of the TAR in 2000, despite a more than doubling of the head count. The same applies with illiteracy rates. This indicates that schooling rates among Tibetans outside the TAR were essentially the same as inside the TAR. Moreover, the Tibetan primary schooling rates the worst among the large ethnic groups in China, and the fifth worst among the entire set of 56 nationalities. The second- and fourth-to-last places were taken by the Monpa and Lhopa (Ch. Moinba and Loba), two miniscule indigenous groups



The Education-Employment Nexus of Urban Exclusion in Tibet 287

of the TAR that are closely related to the Tibetans, with populations of less than 10,000 people each. The Bao’an, an equally miniscule Muslim group indigenous to Qinghai, scored third-to-last place. The poor education performance of Muslim groups specific to the Qinghai-Gansu region is also revealing. The Salar Muslims, based almost entirely in Qinghai, were the sixth worst in terms of no-schooling rates, and their schooling rates were only marginally higher than the Tibetans’. The Dongxiang, another Muslim minority concentrated mostly in Gansu and also neighboring Tibetan areas, had by far the worst schooling rates of the country, placing last among the fifty-six nationalities of the PRC. In educational terms, the northwest Muslims represent an ethno-religious group that is as deprived as the Tibetans. 18.  For reasons not explored here, I avoid using the terminology of cultural capital. For some discussion of this, see Yi (2008). 19.  This example was recounted to me by a senior Chinese professor of Tibetan studies in Beijing in June 2004. It was repeated to me by several other Tibetan and Chinese scholars since that time, suggesting that it plays an iconic role in the official scholastic rhetoric. 20. Interview, Lhasa, December 2005; based on information supplied by the Lhasa taxi association. 21.  Unemployment data in China is also problematic. See Hussain (2003, 21-24) for more detail on the problems with official unemployment reporting in China. 22.  See an excellent discussion of this in Zenz (2013). 23. Zenz (2013) offers some interesting ways of measuring and representing these comparative qualitative outcomes between different types of schools or cohorts of students at the senior secondary and tertiary levels. 24.  For an example of a clear statement on these issues, see PRC (2001b). For an academic version, see Sautman and Eng (2001). 25.  Coincidentally, I published a report on these data in early 2005 (see TIN/ Fischer 2005) on the basis of data provided in TSY (2004). The subsequent TSY (2005) no longer reported this data. 26. For instance, see Saich (2004) for a discussion of the national reforms in 2003, which cut 1.5 million state-sector jobs, mostly at higher administrative levels and in the richer provinces. 27.  Similar dynamics plausibly underlie the trend toward harmonization of international aid among bilateral and multilateral donors at the international level. 28. See Zenz (2013) for a detailed discussion on the very limited degree to which Tibetan language was actually included in exams for public employment in the Tibetan areas of Qinghai (during the time of his fieldwork from 2006 to 2008) and how, even when included, the requirement was easily disregarded or overridden by other formal and informal selection criteria (including the extensive role of corruption). 29.  In particular, this was noted to me repeatedly in interviews with INGO staff and Tibetan scholars in Lhasa during fieldwork in November 2004, no doubt due to the temporal proximity of the reforms at that time. Similar dynamics were noted in interviews in Qinghai in 2003 and 2004. 30.  For instance, this was noted in one interview with three Han government officials in Sichuan in September 2007. They therefore suggested that these problems that I was describing in the Tibetan areas of Sichuan were not particular to Tibetans

288

Chapter 6

but were a general quandary throughout China. This is true, although the educational asymmetry between urban permanent residents and migrants is not. 31.  Again, for detailed analyses of these issues in China, see Li (2002), Guo and Chang (2006), Li (2008), Liu (2008), and Roulleau-Berger (2009). 32.  For instance, see Solinger (1999; 2002), Zhang (2001), Hussain (2003), and Lee and Warner (2005a; 2005b). 33.  Emily Yeh made this argument at a conference in Oslo in December 2008 based on research with Han residents in Chengdu who had recently returned from Lhasa. For similar earlier expressions, see Yeh (2007). 34. In particular, I had several key interviews between June and August 2004 with one senior Chinese official of Malho Prefecture, two senior Tibetan officials of Rebgong County (both meetings in Rongwo Town, Rebgong County), and a group of three Tibetan officials of Chabcha County, four Tibetan high school teachers and one Tibetan graduate student in Chabcha Town. In addition, over the course of my fieldwork I had numerous occasions to discuss these issues with university and high school students and graduates, and teachers in Rongwo and Xining. 35.  See an excellent discussion of the current system in Zenz (2013). 36.  Apparently, these reforms were delayed in the Tibet Autonomous Region, the most sensitive of all provinces in China, until 2006 (personal communications with Melvyn Goldstein, October 2007). 37.  I am grateful for extensive discussions with Adrian Zenz in early 2013 after reviewing Zenz (2013). Both the book and the discussions greatly enhanced my understanding of some of the intricate dynamics occurring in this minority system in Qinghai. Also see Yi (2005a; 2005b; 2008) for excellent studies of these parallel systems in Rebgong. 38.  See a discussion of the difference between these two terms in Zenz (2013), drawing from Harrell (2001). 39.  On this point, all of the Tibetans who I discussed these issues with during field work in 2003 and 2004 expressed the same contention, including over twenty teachers, over one hundred students, ten scholars, eight officials from county or prefecture education departments, and five western scholars and INGO workers with long term experience working in education in Tibet. This perception was shared in all Tibetan areas that I visited. I therefore take this insight as common knowledge, deriving from the fact that the minority education system does have a humanities bias. Also see Bass (1998; 2005) and Yi (2005a; 2005b; 2008). 40. These schools have typically been of lower quality than the mainstream schools, although Zenz (2013) also provides very interesting evidence on how several Tibetan schools in Qinghai came to achieve high standards in the 2000s and were competing well with the mainstream schools, to the extent of even surpassing them in certain criteria. 41.  In Zenz (2013) and in personal communications, he notes that there is nonetheless a tension here as the expansion of Tibetan-medium majors came directly out of a reasoning of expanding Tibetan employment opportunities. Hence, he argues that there are different developments occurring. Central and provincial governments push for homogenization even while diverse options remain on the ground. 42. For instance, in the case of one family interviewed in Chentsa County in 2004, their daughter had just entered the Qinghai minority nationality university in



The Education-Employment Nexus of Urban Exclusion in Tibet 289

Xining, for which they were paying a total of 7,000 yuan per year, which probably included room and board. They financed this through borrowing. The average per capita rural household income in 2004 in Qinghai was 1,958 yuan. 43.  Among other insights, this was particularly noted to me by Susan Costello during a field visit to Darlak, Golok Prefecture, in May 2004. 44.  Despite reforms, the local allocation of public employment can be broadly characterized as clientelistic or even nepotistic, particularly in remote Tibetan areas or at the county-level and below, where certain (Tibetan) families often dominate most of the cadre positions. This was noted in several interviews with Tibetan and western scholars. For instance, one western scholar who was working in Yushu and who I interviewed in Xining in June 2004 reported to me that most of the local county-level government posts in the county where she did her research were occupied by one (extended) family from one village in the county. 45.  This point was widely noted to me during my fieldwork by Tibetan teachers, scholars, and officials. Also see this point in Zenz (2013). 46. For instance, Chinese middle-class children in Xining or Rebgong at the secondary level typically receive private tuition throughout their summer holidays, whereas most of the rural Tibetan students whom I interviewed, particularly the young women, were required to perform farm or herding work during these same holidays and few besides those from elite families could afford extra tuition. 47.  Also see Fischer (2005b; 2008b; 2008e). 48.  See an excellent discussion of this in Zenz (2013, chapter 2). 49. Much of my understanding of the Panchen Lama’s non-secessionist nationalist advocacy comes from extensive conversations in 2000 and 2001 with an exiled ex-monk from Tashi Lhunpoe, the head monastery of the Panchen Lama in Shigatse, TAR, during my time in the Tibetan refugee community in India. This was supplemented with many more discussions with Tibetans in Amdo during my fieldwork in 2004, such as those mentioned in the notes above. For references, see Schwartz (1994), Hilton (1999), Barnett (1998; 2006), and TIN (2004c). 50.  Again, my knowledge of this comes mostly from having attended numerous teachings by the Dalai Lama during my time in the Tibetan refugee community in India from 1995 to 2001. In particular, several thousand Tibetans from Tibet usually legally or illegally cross the border to attend his annual Spring Teachings or his occasional teachings and empowerments held in various locations in India. During these events, the Dalai Lama usually talks at length about the importance of modern secular education as a means of preserving Tibetan culture and identity. 51.  Again, see Zenz (2013) for strong evidence of the rapid increase in Tibetan enrollments at universities in Qinghai, especially at the minority university or minorities departments and even in Tibetan-medium tracks.

7 Subsistence Capacity and the Material Foundations of Resistance

The analysis of this book has thus far portrayed fairly standard development transitions in Tibet, such as demographic transition, transitions out of farming and herding, urbanization, and rising prosperity, alongside associated perversities such as rising and/or high levels of inequalities and exclusionary processes across all strata of the urban labor hierarchy. Despite the perversities and associated tensions, the commonly held view is that such development transitions are nonetheless broadly progressive and indicative of upward mobility. Hence, if and when people come to have strong grievances in the midst such transitions, these are often explained as being caused by obstacles to upward mobility, particularly in a context of rising aspirations and rising living standards. Indeed, the analysis of the previous chapter regarding exclusions operating at the upper end of the labor hierarchy in Tibet could be understood in this way. Or else, grievances with state control might be understood, à la James Scott (2009),1 to drive some people to choose marginality and subsistence as strategies of resistance or escape from state control. Both interpretations associate grievance with deprivation, whether deprivation causes grievance or else grievance drives people to make choices that result in deprivation. Moreover, both rest on the assumption that transitions out of farming and herding represent upward mobility in most cases and, hence, would be seen as progressive from the perspective of rural people in the absence of the factors causing grievance. This understanding, however, becomes problematic if rural people perceive such labor transitions as downward displacements within local labor hierarchies, as understood through local perceptions of status and dignity. In that case, marginalization could be 291

292

Chapter 7

perceived to occur through processes of integration and transition rather than through exclusion and retrenchment. For instance, from the perspective of data on per capita GDP, household income, and expenditure, Tibetan farmers and herders are among the poorest in China, as analyzed in the previous chapters. They are also the poorest according to a variety of human development indicators, particularly those relating to education and health. These data seem to indicate that rural Tibetans are at the bottom of the labor hierarchy in China and that their increased integration into low-wage employment would be progressive. This is supported by findings that wealthier Tibetan rural households derive more of their income from off-farm labor than their poorer neighbors, as evidenced in Goldstein et al. (2003, 776). Indeed, in my earlier work, such as Fischer (2002; 2005a), I argued that Tibetans are in urgent need of expanded (albeit protected) low-skill, off-farm employment that matches their low literacy and schooling levels. This conclusion is obviously based on the assumption that, if and when such employment were to be available, farmers and herders would welcome it. However, several conundrums continuously nagged me in this deduction. The first, which I confronted before my fieldwork in Tibet, is that rural Tibetan households are not necessarily the poorest in China according to the official data on rural assets. In fact, they are the richest. One might argue that this is simply a statistical reflection of different agricultural systems, whereby pastoralism is much more asset intensive without necessarily implying greater wealth, and thus Tibetans may still be the poorest of China, all things considered. However, this logic was not necessarily confirmed in the field. Compared to the average farmer in Sichuan, the average rural Tibetan often appears fairly well off, particularly in terms of housing or conspicuous consumption, or their ability to take leisure, to remain unemployed for extended periods of time, to travel on long pilgrimages, or to offer free labor to monasteries, among many other examples. Accordingly, I confronted the second puzzle in my fieldwork. At the risk of making a gross generalization, according to my typical observations, rural Tibetans, particularly in pastoral areas, do not act as if they are the poorest of China, at least not in ways one might expect in terms of their wage and non-wage expectations in comparison to Chinese workers. Instead of readily accepting menial work at the lowest rungs of non-farm employment, many rural Tibetan households appear to voluntarily abstain from these stigmatized forms of work while targeting more coveted employment options through the selective education and migration of certain family members. While these attitudes may well derive from cultural notions of dignity, the freedom to act in accordance with these notions is notable, rather than being forced into undignified labor at any wage, as we would expect of those who are truly compelled by poverty.



Subsistence Capacity and the Material Foundations of Resistance 293

The combination of these observations would seem to suggest that the relative asset wealth of Tibetans trumps their relative income poverty when it comes to the factors influencing their employment behavior. In much of the Chinese literature, as represented by the influential work by Xiaoqiang Wang and Nanfeng Bai (1991 [1986]), this is often explained as an aspect of irrational backwardness among Tibetans or other minorities, which undermines productive accumulation and development. Although less explicit, similar assumptions also underlie much of the implicitly derogatory attitudes toward “subsistence agriculture” within both policy and academic circles of development studies, insofar as subsistence is usually seen as either a cause or a symptom of endemic poverty in the developing world, Tibet included. In both cases, much effort is made to explain why peasants would act in ways that preserve their basis in subsistence agriculture even though such a course of action is ultimately a dead-end strategy, doomed to result in fragmented landholdings and chronic poverty.2 This chapter assesses these various interpretations by examining the role of subsistence in rural household livelihood strategies within the context of rapid economic and social transitions out of agriculture. I argue that subsistence is valued by these households precisely because it provides the material foundations, or what I call “subsistence capacity,” from which they can choose to act in a variety of strategic ways in response to dislocating change. When faced by rapid change, such households aim to maintain an asset base at a level sufficient to meet minimum subsistence needs so as to preserve this capacity to act strategically. Furthermore, these strategies are not necessarily contradictory to market integration. Rather, subsistence-based strategies can often be complementary to market integration, although they generally aim to avoid subordinated forms of labor commodification. This emphasis is different from conventional theories of functional dualism that interpret subsistence as serving to subsidize capitalist processes of labor subordination.3 It is also to be distinguished from the idea of subsistence in The Moral Economy of the Peasant by James Scott (1976), for instance, insofar as the idea underlying subsistence here, in accordance to the way it was conveyed to me by farmers and herders in Tibet, is one of wealth rather than poverty. In this sense, subsistence capacity serves as a crucial material foundation for allowing households to respond to rapid social and economic transitions in ways that move beyond mere survival and coping, and seek to innovate or negotiate alternative forms of integration into changing matrices of wealth and power. Along these lines, notions of dignity generally concur with types of labor that allow households to prioritize the maintenance of subsistence capacity. Conversely, much of the grievance and contention with development is with respect to policies and processes that undermine such capacity. In other words, the proclivity for subsistence among “peasants” and other

294

Chapter 7

rural folk derives not so much from a survival logic engendered by the condition of living on the verge of destitution, in accordance with the idea of subsistence in Scott (1976), but instead from an implicit and culturally endorsed understanding of the crucial role of subsistence in supporting wealth within rural livelihood strategies, more or less akin to the way that monetary wealth provides freedom within a modern, non-rural setting. This point is important because, following the argument of the previous chapter, the Polanyian impetus for double movement, that is, the impulse for social protection and resistance to the dislocating changes wrought by rapid development, is most likely to emerge out of the middle or upper parts of a social hierarchy, that is, from those people most at risk of losing subsistence capacity and thus of relative demotion in the labor hierarchy. This argument is made in four sections. The first examines the apparent paradox between income poverty and asset wealth. The concept of “subsistence capacity,” in part inspired by the work of Arthur Lewis and Polly Hill, is used as a device to understand aspects of wealth that are difficult to capture through conventional measures but that have an important bearing on wage and employment expectations, and thus on livelihood strategies. The second section explores further certain aspects of these Tibetan employment and livelihood strategies in the context of rapid economic transition, as commonly observed in the field. The third section contends with the argument of Wang and Bai that these strategies are symptoms of “backwardness.” Instead, they can be seen as strategies aimed at resisting demotion in dislocating labor hierarchies. As such, subsistence-maintenance strategies are better understood as forms of wealth protection rather than as stemming from an anti-development logic. The chapter concludes with a brief discussion of government resettlement strategies in the pastoral areas of Qinghai, Gansu, and Sichuan through the lens of this argument, and a reflection on how the material conditions of resistance to exclusion are more likely to apply under conditions of wealth rather than poverty.

Relative Income Poverty, Productivity, and Relative Asset Wealth This section briefly describes the dynamics of rural Tibetan wealth since 1990 as represented in the official Chinese statistical sources, summarized from my more detailed analyses presented elsewhere.4 As usual, the TAR can be used to broadly represent all the Tibetan rural areas in China. There are primarily three sets of data of interest: per capita household income (or the closely related data on expenditure), as already analyzed in the second section of chapter 5 (see figure 5.7); productivity; and per capita productive fixed assets. Compared to the same measures elsewhere in China,



Subsistence Capacity and the Material Foundations of Resistance 295

these measures of wealth should in principle provide some indication of the underlying material conditions affecting relative reservation wages, as discussed in the next section. However, these three sets of data depict contradictory patterns of rural Tibetan wealth relative to the rest of rural China. An attempt to resolve this is made through an analysis of subsistence. Rural Household Incomes As discussed in chapters 2 and 5, average per capita rural household incomes in the TAR apparently collapsed in the 1990s in real terms (i.e., according to purchasing power, after discounting for rural consumer price inflation). After 1997, they slowly rose, albeit surpassing their previous 1992 peak only by 2003. In relative terms, they fell to the lowest of all Chinese provinces from 1997 to 2002. One surprising result from figure 5.7 (see chapter 5) is the fact that average rural incomes in the TAR were apparently much higher in the early 1990s than in other western provinces and even higher than the national average (in real terms, indexed by the respective rural CPIs for each province). This comparison somewhat overestimates the situation in the TAR given that price inflation in the TAR was the highest in China during the 1990s.5 Nonetheless, even in current prices (i.e., prices in the year of measurement), average rural incomes in the TAR exceeded the national average in 1992 and were higher than in most other western provinces besides Xinjiang throughout the 1980s and early 1990s.6 Inflation notwithstanding, the relatively high incomes up to the early 1990s reflect the fact that rural Tibetans benefited quite well from the first decade of the reform period compared to other western provinces. The introduction of individual patterns of land use in the early 1980s combined with other pro-agrarian policies of the early reform period provided an immediate recovery from the debacles of collectivization in the 1970s (as discussed in chapter 2), although the precise factors that boosted rural Tibetan wealth are difficult to decipher from the data of that period and will not be analyzed here. Goldstein et al. (2003, 764) found that 94 percent of the farming households in the TAR that they surveyed in the late 1990s felt their livelihood had improved since decollectivization. Official data seem to indicate that the improvements took place mostly in the 1980s. In contrast, the rural economic stagnation in the TAR during the 1990s that appears in these official income data was exceptional among all of the Chinese provinces. If accurate, it reflects that rural Tibet was particularly vulnerable to inflation and structural economic transformations over this period. The statistical appearance of stagnation may also be in part related to measurement issues, such as changing methods of measuring subsistence,7 or new sources of income that might have been poorly recorded in

296

Chapter 7

these data, such as the largely unrecorded boom in the caterpillar fungus trade (see Winkler 2008a; 2008b). However, many of these issues would have presumably impacted the income reporting in other western provinces as well. The strong rise in rural incomes after 2002, as shown in figure 5.7, was due largely to increases in wage and salaries, household operations, and transfers (per the official household income data). As discussed in chapter 5, these increases reflect a variety of rural development initiatives, such as those discussed by Goldstein et al. (2008; 2010) and Childs et al. (2011), alongside the more general context of rising and heavily subsidized affluence. However, average per capita rural household incomes of the TAR nonetheless remained in the range of the poorer provinces. For instance, by 2010 they were still considerably lower than those of Sichuan, the source of the majority of Han Chinese migrants to most Tibetan areas (rural household income in 2010 was 4,139 yuan per person in the TAR versus 5,087 yuan in Sichuan). Moreover, the comparison underestimates the relative income poverty (or overestimates the relative purchasing power) of rural Tibetans in the TAR given that the cost of living in the TAR is considerably higher than in other western provinces.8 Hence, while it is likely that there has been substantial reduction in absolute income poverty in the TAR (and other Tibetan areas) since the late 1990s, it is also likely that income poverty rates in the TAR are still among the highest in China (according to official data, which has not been made available since TSY 2000). In 1999, the last year when rural household income distribution was made available in the TSYs, about one-third of the rural populations in both the TAR and Qinghai were living below or slightly above the then national absolute poverty line of 865 yuan per person per year (Fischer 2005a, 96-110). This proportion would have subsequently fallen due to rising incomes, particularly from 2003 onward, although the impact of rising incomes would have also been compensated for by rising rural inequality, as discussed in chapter 5. While it is not possible to estimate the poverty trends in the TAR since 1999 given the lack of publicly available income distribution data and the fact that the government also substantially revised the official national absolute poverty line several times since the late 1990s, it is very likely that income poverty rates have remained among the highest in China given the available data on average household incomes. Productivity The stagnation of Tibetan rural incomes over the 1990s and their relative lagging cannot be explained by differences in agricultural productivity with the rest of western China. In fact, as indicated in table 7.1, productivity of



Subsistence Capacity and the Material Foundations of Resistance 297

cereals per hectare in the TAR in 2010 was among the highest in western China and close to the national average. This is consistent with the data in 2004, when cereal productivity per hectare was higher than the national average (see Fischer 2008e, 11-12), as well as with earlier data in the late 1990s (see Fischer 2002, 20-23). The only western provinces that were consistently more productive in grains than the TAR were Xinjiang, known for its intensive agro-industry, and Sichuan, known as the garden of China with three harvests a year in the Chengdu basin. Productivity of rapeseeds per hectare was the highest in the country (except in Jiangsu, not shown, which was two kilograms higher than the TAR), which is no doubt reflective of the fact that high-altitude rangelands have proven quite effective for this crop. Considering that the TAR yields are achieved within a short and often harsh growing season (whether in terms of crops or grasslands), they represent impressive productivity.9 According to the data on output per capita, the TAR appears more western average in farm output such as grains and oil seeds, although still well above average in per capita meat and milk output. In per capita meat output the TAR was surpassed only by Inner Mongolia (86 kg in 2010), and by Inner Mongolia, Heilongjiang, and Ningxia in milk output. The differences with the per hectare output in farm goods is probably reflective of a greater labor intensity of farming in the TAR than in other provinces, although it is not clear whether the per capita measures used here are specifically based on the population working in the production of these products or else of a more general measure of the rural population.10 In either case, the reduced performance of the TAR according to these per capita measures (in farm output, not herding output) would be reflective of a more agrarian rural population, that is, that a greater proportion of the rural labor force Table 7.1: Agricultural productivity in selected provinces and national average, 2010 TAR

Xinjiang Sichuan Qinghai

Gansu

Yunnan

Guizh.

China

Per hectare yields, kg Cereals

5,430

5,969

5,557

3,756

3,772

4,172

4,978

5,524

Rapeseed

2,442

2,158

2,166

1,948

1,816

963

1,077

1,775

Per capita output, kg Grain

309

539

397

182

369

334

306

409

Oil seeds

20

31

33

61

25

8

17

24

Meat

84

49

67

49

30

62

45

46

Milk

79

59

9

47

14

11

1

27

Sources: CSY (2011, tables 13-16 and 13-21). Meat includes pork, beef, and mutton.

298

Chapter 7

works in agriculture than in China, with a greater intensity of labor used per unit of land. The same principles would apply with regard to animal husbandry except that the data are not as easily comparable given very different production systems. It is likely that the high meat and milk output of Heilongjiang is based much more on agro-industrial processes than in the subsistence-based herding systems of the TAR. Indeed, the problem of these per capita measures in the TAR is that many households are involved in both farming and herding, in which case, the lower per capita or per laborer output in one activity might simply reflect the allocation of household labor across both activities. As noted above, this performance of the TAR has been consistent since at least the 1990s, with productivity measures increasing in tandem with national agricultural productivity. In other words, farmers (and, presumably, herders) in the TAR have been at least as successful as the predominantly Han Chinese provinces in bringing about increases in agricultural productivity, even if their activities have remained largely subsistence based with minimal state assistance up until the late 1990s (agricultural extension programs and rural infrastructure improvements have nonetheless featured more prominently since the early 2000s and would have also supported increases in productivity). Therefore, the stagnation of per capita rural incomes in the TAR up to the early 2000s and their lag with all but the poorest provinces in western China was not necessarily related to deficiencies in the productivity of Tibetan farmers and herders. Income stagnation was more likely related to the collapse in the prices of the main commodities produced by Tibetans (wool and grains) throughout the 1990s, alongside a shortage of rural employment opportunities outside farming and herding, and faster rural population growth than elsewhere in China. In essence, collapsing prices were compensated for in most other regions of China by a rise in off-farm rural employment, which was one of the strongest factors driving growth (and inequality) in rural incomes.11 In contrast, the fate of rural Tibetans was disproportionately determined by the fate of farming and herding, relative to other rural populations in China. The situation led naturally to the policy conclusion that increased wage employment was necessary for the TAR to catch up. This situation changed quickly in the TAR since the early 2000s with rapid transitions out of agriculture and strong increases in off-farm sources of employment and income, as analyzed in chapter 5. For instance, the proportion of average per capita rural household incomes derived from wages and salaries in the TAR increased from 17 percent in 2000 to 27 percent in 2010, when it started to reach a proportion that was comparable to Sichuan and China in 2000 (the proportion increased over the same years in Sichuan from 32 percent to 44 percent, and in China from 31 percent to 41 percent). The other large proportionate increase in the TAR was from transfers, which



Subsistence Capacity and the Material Foundations of Resistance 299

rose from 4.7 percent in 2000 to 13.3 percent of average per capita rural household income in 2010 (versus a rise in Sichuan from 3.8 percent to 8.5 percent). Correspondingly, income from household operations (i.e., farming and herding) declined as a proportion of average net income from 70 percent in 2000 to 56 percent in 2010 (it is not clear to what extent this includes income from caterpillar fungus, as mentioned previously).12 As elsewhere in China, the overall increase in average per capita rural household incomes in the TAR was much more closely related to these transitions in the sources of rural incomes than to the gradual increases in agricultural productivity. Productive Fixed Assets The relative income poverty of Tibetans co-exists with relative asset wealth. Rural households in the TAR were in fact the most asset intensive in China in 2010 (see figure 7.1), as they were in previous years since as long as these data have been reported, with more or less a consistent ratio to other provinces and the national average (cf. Fischer 2008e, 13). The next most asset-intensive provinces were Ningxia and Xinjiang, followed closely by Inner Mongolia and Qinghai, all provinces where pastoralism is prevalent. Most of these productive fixed assets were based in agriculture, as elsewhere in China with the exception of Ningxia. In the pastoral cases, this obviously includes a large component of livestock, as evidenced in the first two rows of table 7.2. Notably, most land-use in the Tibetan areas is either pastoral or mixed pastoral-farming, and livestock feature prominently even within pure farming areas.13 However, Tibetan households also had a much higher value of nonlivestock productive fixed assets, as shown in the last six columns of table 7.2. There were 3.3 times the number of motorized vehicles per hundred households in the TAR than on average in China in 2010, 1.1 times the number of large and medium tractors, and 2.4 times the number of mini or walking tractors. The number of motorized threshing machines and pumps was less than the national average and much less than in Sichuan, reflecting the fact that these assets are more related to intensive farming. Conversely, the higher value and number of transport-related productive fixed assets is obviously related to the much greater distances that are required to travel in the Tibetan areas. Older yearbooks also show that about half of the original value per rural household in the TAR was in the form of non-livestock productive fixed assets such as industrial machinery or buildings for productive purposes (CSY 1999, table 12-11). Similar to productivity, the ratios with other provinces or the national average have been more or less consistent since the late 1990s, with broadly proportionate accumulation of assets in all the cases shown albeit with some shifting within categories. For example, there was a per capita drop in the number of draft animals

300

Chapter 7

Figure 7.1: Original value of productive fixed assets per rural household, 2010

Source: CSY (2011, table 13-10)

Table 7.2: Major productive fixed assets per 100 rural households, end of 2010 (unit) TAR Ningx Xinj/g In. M. Qing. Gans. Yunn. Sich. Guiz. China draft animals

227.1 32.5

58.3 44.8 61.5

59.2

20.0 57.8

23.4

commodity animals

665.0 35.3 568.8 192.6 157.0 38.8

66.7

48.4 26.3

62.3

motor vehicles

56.2

7.9

6.0

2.8

1.8

3.5

1.5

2.7

1.5

2.1

2.4

large and medium tractors 4.2

4.2

8.1

4.2

4.2

6.6

1.2

0.4

0.4

3.7

mini and walking tractors 47.5 55.8

28.1

52.0 56.8 30.7

10.5

2.7

1.4

19.5

motorized threshing machines

6.9

4.1

2.1

4.6

6.1

4.2

8.8

26.0

8.7

10.6

carts with rubber tires

10.8

5.2

45.9

31.2

3.5

15.9

3.6

1.6

0.9

8.4

pumps

n/a

22.0

3.6

37.7

0.3

12.1

13.2

31.4 10.0

25.9

Source: CSY 2011, table 13-11.

in the TAR and an increase in commodity animals compared to 2004 (cf. Fischer 2008e, 13). This seems to indicate that rural Tibetan households have been accumulating productive capital at a similar rate as elsewhere in rural China. Even at this relatively much higher asset intensity, the numbers of assets represented in table 7.2 might be considered as still quite poor in absolute terms (2.3 draft animals and 6.7 commodity animals per household). However, it should be noted that these represent provin-



Subsistence Capacity and the Material Foundations of Resistance 301

cial averages. In the TAR this represents a mix of pure farming regions with mixed farming-pastoral and pure pastoral regions. Moreover, even though pastoral households account for most of the land use in the TAR, they only account for a minority of households, given that pure farming areas are much more densely populated.14 Thus, the actual asset intensity of pure pastoral households would be much higher than shown in table 7.2, while that of pure farming households would be somewhat lower, although probably still higher than the national norm given their larger landholdings, as discussed below. It could be disputed that these asset measures do not accurately reflect comparative asset intensity given that the most important asset for farming households is land, which is not included in these data. Therefore, the livestock and related asset intensity of Tibetan pastoral households could be offset by the value of farmland elsewhere in China (and in the pure farming Tibetan areas). However, this line of argument overlooks the fact that rangeland is also highly valued by pastoralists given its direct relationship to the potential size and quality of a herd. Indeed, both rangeland conflicts and the emergence of pasture renting in many areas are sure signs that land is valued.15 The fact that the per-unit value of rangeland might be lower than farmland, due to a lower per-unit intensity of use, is compensated for by the fact that pastoral households have much more land than farming households. Therefore, even if land values were included in the fixed asset data, it is probable that the TAR would again be ahead of the national average, even within this category, as well as pastoral areas vis-à-vis farming areas within the TAR itself. These variations in landholding sizes between the Tibetan areas across the plateau and the lowland areas of China proper represent broad differences in population density relative to land availability. The population density of the TAR was about 2 people per square kilometer in 2000, and about 7 for Qinghai (including the densely populated northeast corner of the province, and thus much less in the Tibetan areas of Qinghai), whereas it was 132 for China and 172 for Sichuan. If we remove the western half of Sichuan that is composed of Tibetan autonomous areas, we are left with a population density of 344 people per square kilometer in eastern Sichuan, versus about 9 in the Tibetan areas of Sichuan (calculations made from CSY 2001, tables 2-2, 2-13, and 2-21; CSY 2002; and Marshall and Cooke 1997, 2487-89). Obviously, arable land is much more limited in these Tibetan areas; it is roughly estimated that only two percent of the land area in the TAR is arable, although about 40 percent is used as pastures (these proportions would be higher in the Tibetan areas of Sichuan). Nonetheless, the huge disparity in population density compensates for this, leaving the Tibetan highlands with much more land per person than the lowlands of China.

302

Chapter 7

Recent data provided in CSY (2011, table 13-12) offer some insight into this assertion, although these data need to be treated cautiously due to complications in classifying types of land (e.g., when grasslands are used for cultivating rapeseed or forage), or in comparing land use in two very different topographic regions. In the table, the area of “cultivated land under management” by rural households at the end of 2010 is recorded as 2.0 mu per person in the TAR, versus 2.1 mu in Qinghai, 2.7 mu in Gansu, 1.1 mu in Sichuan, 4.8 mu in Xinjiang, 9.7 mu in Inner Mongolia, and 2.3 mu in China overall.16 The “area of grassland” under management was 35 mu per person in the TAR, versus 23 mu in Qinghai, 0.2 mu in Gansu, 8.1 mu in Xinjiang, 127 mu in Inner Mongolia, and 4.2 mu in China (no data on grasslands is provided for Sichuan although we can presume that it is substantial given that half of the province is composed of Tibetan highlands). The averages for Qinghai, Gansu, and Sichuan are complicated by the fact that they include both sparsely populated Tibetan highlands and densely populated Han Chinese lowlands. An example from my fieldwork in Xunhua County in Qinghai illustrates these differences in typical holdings of farmland at a more local level. Typical holdings of farmland in the Tibetan highlands of this county in 2004 were about two to three mu per person, whereas landholdings averaged about half a mu per person in the lowland Salar areas.17 These average farmland holdings of Xunhua Tibetans were very representative of other farming areas that I visited throughout Tibet, such as in the pure farming areas of Chentsa, Rebgong, Chabcha, or Mangra Counties in Qinghai; Kardze, Tranggo, or Dawu Counties in Sichuan; or Lhasa, Lhoka, and Shigatse Counties in the TAR, in all cases typically ranging from two to four mu per person. Similarly, outside the Tibetan areas, landholdings in the Han or Muslim farming areas of western or central China tend to resemble those of the Salar in Xunhua, as suggested above with respect to the average holding of cultivated land in Sichuan of 1.1 mu per person. The mixed farming-pastoral Tibetan zones also have rangeland holdings in addition to the household farmland holdings, and household rangeland holdings in pure pastoral areas can be considerable. In Sogwo County in Qinghai, a fairly densely populated Mongolian pastoral area that I visited in 2003, typical rangeland holdings were around 500 to 1,000 mu per household, or, depending on the size of the household, more than 100 mu per person.18 Some of the pastoral areas that I visited in 2004 in Kardze Prefecture in Sichuan had denser populations, resulting in greater pressure on rangeland holdings and smaller herds. Susan Costello, a scholar with extensive field experience in Golok, reported to me that the landholdings of the pastoral households that she studied in Golok ranged from 1,800 to 2,400 mu, which would also be similar to landholdings in the pastoral areas of Yushu, given that both prefectures in Qinghai have abundant land, albeit



Subsistence Capacity and the Material Foundations of Resistance 303

at higher altitudes and therefore slower growth of pastures.19 I was not able to visit any pastoral areas in the TAR given severe government restrictions on rural travel, although discussions with several Western scholars working in pastoral areas of western and southern TAR confirm these general observations from elsewhere in Tibet. Given the extent of these differences in landholding sizes between farming and pastoral areas in Tibet, it is very plausible that the higher fixed-asset values of pastoral areas correlate with higher land asset values, if the latter were measured. Finally, it could be further argued that these differences in asset values simply reflect different agricultural systems, whereby pastoral systems require more fixed and land assets in order to achieve a basic level of subsistence or absolute income. For instance, one might find drought-stricken Somali pastoralists possessing more fixed assets than farmers in more productive neighboring agricultural areas. Moreover, assets play an important risk insurance role and thus cannot necessarily be treated in the same way as an income stream. Indeed, one of the rationales for maintaining large herds is to ensure against environmental hazards such as harsh winters: a large herd increases the chances of surviving such hazards with a herd size that can still maintain subsistence. This in turn helps to explain why pastoralists tend to be risk averse and thrifty with regard to asset management, just as subsistence farmers tend to be risk averse with regard to crop cultivation.20 These arguments notwithstanding, there is a general consensus in Tibet among local people and scholars that pastoralist households tend to be wealthier than farming households on average, particularly in terms of disposable wealth or durable consumer goods.21 This was also probably the case in previous periods of history (as it was probably the case in predrought Somalia). However, this is not reflected in the household income statistics in part due to the fact that pastoral savings (e.g., an increase in yak herd size) are treated as productive fixed assets and, hence, are not reflected in the data on pastoral household incomes (only slaughtered livestock would appear as income). In contrast, savings of grain output (in kind or converted into cash) are calculated as part of farming household incomes because a harvest is treated as income. Therefore, the asset data need to be combined with the income data for an overall evaluation of comparative levels of wealth across regions or different agricultural systems. However, given that it is unclear how assets and income should be combined, in what proportions, and under what circumstances, there is considerable ambiguity in comparative wealth assessments. In certain cases there may even be an inverse relationship between income and assets. A remote pastoral household might appear poor in terms of income (and in terms of education and health) yet possess substantial livestock assets, enough to offer considerable subsistence consumption (even before considering the windfall opportunities provided by a lush supply of caterpillar fungus, as

304

Chapter 7

is the case in many remote high-altitude pastoral areas). In contrast, a pastoral household located closer to a populated area might have much more circulating income due to its integration into the “commodity economy,” even while its asset base is depleted due to a shortage of quality pastures or to stronger state enforcement of livestock limits.22 Furthermore, asset depletion might be reflected by income increases, such as when a herder is hard pressed for cash in one year and sells off a larger part of his or her herd than he or she normally would.23 In these cases, an increase in income wealth may in fact represent considerable subsistence impoverishment.24 All these considerations nonetheless hint toward the fact that there is considerable wealth generated and stored in subsistence-based rural economies that is often poorly captured by conventional household income measurements. As a result, Tibetan rural areas and, in particular, Tibetan pastoral areas may be considerably wealthier than the official income data suggest. Moreover, these conventional measures are poorly conceived for capturing the wealth dimensions of transition from more subsistence to more monetized or commoditized modes of wealth circulation; smooth increases in real disposable incomes may not necessarily reflect actual changes in overall wealth. In other words, rising incomes might hide the erosion of asset-based subsistence through population and economic pressures, or through government policies that undermine this asset base, such as resettlement or land reforestation programs, or the imposition of livestock limits. Subsistence as Wealth One way around these measurement problems is through an evaluation of subsistence. In deference to Polly Hill (1986, 16-21), I use the term “subsistence” with caution (for reasons discussed in the third section), although I have generally found that rural Tibetans do operate within a subsistencesurplus mode of production and consumption, in the sense that they do consume much of the output they produce. In support of this suggestion, Goldstein et al. (2003, 767) estimated in their survey of thirteen farming villages in the TAR that 77 percent of households produced enough grain for their consumption needs or a surplus. Indeed, this is also supported by the official data: although rural Tibetans in the TAR had among the highest per capita output of meat and milk in China in 2010, they apparently had only national average per capita sales of mutton or beef, or else far below the national average per capita sales of milk, at an almost negligible amount of 0.9 kg of milk sold per person, versus 11.3 kg nationally or 140.5 kg in Inner Mongolia (CSY 2011, table 13-23). Given that these incongruencies have been consistently reported since the 1990s (cf. Fischer 2002, 22), the implication is that milk, in particular, is mostly consumed



Subsistence Capacity and the Material Foundations of Resistance 305

or bartered by rural Tibetans (i.e., 79 kg of milk was produced per person in 2010, whereas only 0.9 kg was sold per person), in contrast to Inner Mongolia, where milk output appears to be more commoditized (i.e., 370 kg produced versus 140 kg sold per person). Moreover, less of per capita household expenditure was in the form of cash expenditure than typically observed in China. Cash expenditure amounted to 78 percent of total household expenditure in the TAR in 2010, versus 88 percent for China as a whole, although this share in the TAR had risen substantially since 2004, when it was only 57 percent.25 Most of the difference—in both 2004 and 2010—was accounted for by food expenditure, implying that the non-cash expenditure was either “own-consumption” or else bartered food. Goldstein et al. (2008) also argue that there was a very rapid shift between 1997 and 2007 in the villages they surveyed in Central Tibet from a predominantly subsistence agricultural economy to a new mixed economy in which non-farm income plays an increasingly dominant role. This is supported by the rise in cash expenditure as a proportion of total household expenditure in the official data. Nonetheless, even despite these changes, rural Tibetans in the TAR still derived much less of their income from salaries and wages than the national average or most other western provinces in 2010, only starting to reach the 2000 national average proportion by 2010, as noted above.26 Hence, while the situation is changing quickly, Tibetan rural areas during the period analyzed by this book can still be broadly considered to have been operating in a “subsistence” mode of production.27 In evaluating subsistence, I propose the concept of absolute and relative “subsistence capacity” as a means to evaluate relative levels of wealth measured in terms of an ability to subsist on household production. “Absolute subsistence capacity” refers to the ability of a household to produce a surplus above the subsistence needs required to reproduce itself economically. This meaning is essentially synonymous with food security at a household or community level.28 “Relative subsistence capacity” refers to this absolute subsistence capacity in comparison to other households or communities that have an impact on labor supply and wage rates within a regional economic system. The relative comparison made here refers to inter-regional comparisons. It does not necessarily fall into the trap of homogenizing the “peasantry” within each region due to the exceptional transitory circumstances of land tenure in China in the 1980s and 1990s. As described previously, variations in landholding sizes between Tibet and China proper broadly represent differences in population density relative to land availability. Within each region, the individual household responsibility system in China has led to a uniformity of landholding size at county or township levels that is exceptional in the developing world.29 Inequality in land assets at the local level is thus largely driven by differences in land qual-

306

Chapter 7

ity rather than size. This situation does seem to be changing rapidly, as noted above with respect to rapidly rising inequality since the late 1990s. It therefore appears to be an exceptional transitory outcome in the move from Socialist to market forms of rural economic organization during the first twenty years of the reform period. These exceptional characteristics of land tenure in China and Tibet offer a window of comparative insight into the underlying material factors conditioning labor processes across these regions. In most other situations in the developing world, these factors are much harder to decipher given that large intra-community inequalities tend to dominate inter-regional differences. If we focus on the case of pure farming areas, the ability of the average Tibetan household to subsist on household production becomes quite clear. For instance, abstracting from supplementary foods, it is generally considered that a person needs approximately 180 kg of staple grains to subsist per year.30 Depending on the quality of farmland, Tibetan farmers typically produce anywhere from 100 to 300 kg of grain per mu per year (one crop, mostly barley).31 With a typical range of two to four mu per person, yields would therefore range from 200 kg to 1,200 kg per person. For the sake of simplicity, a conservative yield of 150 kg per mu on three mu of averagequality land per person yields 450 kg of barley per person per year.32 In this typical case, the household produces a surplus of about 270 kg of grain per person above subsistence grain needs, which is then sold, bartered with pastoralists for meat and butter, offered to monasteries, stored, or used for making barley wine. In comparison, consider the typical Salar rural household in Xunhua. Average per mu yields are higher than in the Tibetan highlands due to proximity to the Yellow River, although they are limited by land degradation and urbanization. An estimate of 300 kg of grain per mu on half a mu per person yields only 150 kg per person per year, that is, less than the subsistence requirement.33 In order to meet subsistence food needs, farming households in Xunhua have no choice but to seek work outside household agriculture. In other words, typical landholding sizes in the Tibetan farming areas are not sufficient to generate any substantial surplus wealth, although they are generally sufficient to cover subsistence needs, whereas those of the Salar or other Muslim and Han areas are not even sufficient for subsistence. Notably, this comparison is between pure farming areas, thereby overcoming the argument (as discussed above) that these differences merely represent different production systems (i.e., pastoral versus farming). In both cases, off-farm labor is sought for a variety of compulsions (e.g., to pay for shortfalls of subsistence, or to pay for education and health fees) and incentives (e.g., the desire to consume beyond subsistence needs or to accumulate wealth). However, the difference between a typical Salar and a typical Tibetan household in Xunhua is that the former is com-



Subsistence Capacity and the Material Foundations of Resistance 307

pelled to seek off-farm work in order to survive at subsistence, whereas the latter is not. The typical Tibetan enters the search for work due partly to the need or desire for cash income, but he or she is not compelled to do so from the point of view of basic subsistence. From the point of view of subsistence, he or she is entering the labor market with a greater degree of freedom and security relative to the typical Salar (or Hui or Han) member of a farming household. This degree of freedom, achieved by reaching a basic threshold of absolute subsistence, constitutes a subjective premium for the opportunity costs of labor rooted in such subsistence and can serve to support cultural or other preferences. Relative Subsistence Capacity and Relative Wages Subsistence capacity in turn can be understood as influencing wage expectations. This perspective is inspired by the theory of factoral terms of trade elaborated by Arthur Lewis (1954). According to the classic contention of Lewis, in a context of “unlimited supplies of labor”—that is, where labor is in surplus supply at a particular wage rate and/or can be taken out of agricultural production without lowering the total output of agriculture— wages in the “capitalist” sector (including capitalist agriculture) are set by the productivity of staple food production for domestic consumption in the subsistence sectors of such an economy (or else in the major sources of labor emigration to this economy), rather than by labor productivity in the capitalist sector. This is because labor productivity in staple food production (constrained by supply of land) determines the reservation wage rate required to draw workers out of such subsistence production and into wage employment in the capitalist sector. Lewis suggested that wages in the capitalist sector would need to be about 30 percent higher than the implicit wages of household subsistence production. The way I have conceived “subsistence capacity” here follows a similar logic, except with the added dimension of compulsion (when subsistence capacity is not reached) or freedom (when it is reached), which is difficult, if not impossible, to value in objective quantitative terms. Lewis assumed that the labor productivity of staple food production would set wages in barter terms, in accordance with the classical economics approach that he adopted, and that this would then set prices. For instance, if a farmer were to be able to grow 1,000 kg of peanuts on his or her land, he or she would need a wage of at least the equivalent of 1,300 kg of peanuts in order to be drawn out of subsistence peanut production, and then prices for the commodities subsequently produced would be determined by this 1,300 kg peanut-equivalent wage rate.34 However, if we add the dimension of subsistence capacity and, hence, degrees of freedom from the compulsion to work, this clearly includes an important subjective element of valuation

308

Chapter 7

that cannot be simply measured in monetary or even barter terms. Thus subsistence capacity cannot be simply approximated by the market value of “subsistence” output. This is not to say that, because these subsistence considerations cannot be measured, they are irrational or merely cultural. While deeply enmeshed with cultural preferences, the subjective dimension nonetheless also reflects an important economic function and, in accordance with this function, plays a key role in influencing the wage and employment preferences of farmers or herders. This obviously presumes that such farmers and herders can act on the basis of their wage and employment preferences and are not forced into labor commodification through either economic forms of coercion (e.g., poverty or lacking sufficient subsistence capacity) or non-economic forms of coercion (e.g., forced dispossession of land). Lewis applied his analysis to the issue of factoral terms of trade between temperate and tropical country exports. He argued that the relative prices of these exports were set in the nineteenth and early twentieth centuries by respective differences in the productivity of staple food production for domestic consumption in the major sources of labor migration for each region, rather than by differences in the productivity of the export commodity per se. This was because “for temperate commodities the market forces set prices that could attract European migrants, while for tropical commodities they set prices that would sustain indentured Indians” (Lewis 1978, 14). Thus, the low levels of labor productivity in Indian and Chinese peasant agriculture set the lower prices for tropical export goods given the mass migration of about 50 million Indians and Chinese across the tropics during this earlier period. Similarly, the higher levels of labor productivity in European and, especially, North American agriculture set the higher prices for temperate export goods, again due to the mass migration of about 50 million Europeans to temperate colonies during this earlier period. “This analysis,” Lewis notes, “clearly turns on the long-run infinite elasticity of the supply of labor to any one activity at prices determined by farm productivity in Europe and Asia, respectively” (1978, 16). If we adapt this reasoning to the case of Tibet, it is possible to conceive how disjunctures would occur between the reservation wages of local Tibetan farmers and herders and the actual wages on offer in the regional economy. The former are determined on the basis of local subsistence capacity and the subjective valuations of the freedom that subsistence capacity can provide. The latter are determined externally, on the basis of similar considerations elsewhere in western and central China where most of non-Tibetan labor emigration originates. The openness of the Tibetan economy and its tiny size with respect to the rest of China effectively results in the inability of Tibetan expectations to influence actual wages. As a result, local wages are set lower than local rural wage expectations and without relation to local capacities.



Subsistence Capacity and the Material Foundations of Resistance 309

This helps to explain the resistance of rural Tibetans to enter into lowwage employment, as discussed in the next section, precisely because, abstracting from other important factors such as cultural preferences, wages in such employment are set below a level that would be required to attract them out of subsistence production activities. The fact that herders are generally considered to be wealthier than farmers in the Tibetan areas offers an additional explanation for why Tibetan herders tend to be much more resistant toward manual wage labor than Tibetan farmers, not only in terms of reservation wages but also in terms of types of work. Resistance is further supported by the fact that the average rural Tibetan is not compelled to seek wage labor in order to meet basic absolute subsistence, and this relative freedom is additionally valued for the dignity that it confers. This is not to say that only these material factors contribute to such work expectations. These obviously interact with inherited cultural norms of what constitutes dignified work. However, subsistence capacity can also be seen as a conditioning material factor in the formation of these values overtime, such as in the development of “working class” attitudes among workers in nineteenth-century England following their dispossession of land assets through the enclosure movements, as analyzed by various classic authors such as Marx and Polanyi.

Employment and Rural Livelihood Strategies As analyzed in chapter 5, rural Tibetans are currently undergoing a very rapid transition out of agriculture, perhaps one of the fastest, albeit latest of such transitions observed during the reform period in China. The concept of subsistence capacity helps to shed light on some of the material aspects that appear to be conditioning Tibetan attitudes toward off-farm employment within the rapidity of this transition. This section offers some observations of these attitudes. Employment and Wage Expectations At the risk of making a gross generalization, one may say that Tibetan farmers and herders are not unduly inclined to engage in long-term, low-wage employment and generally have higher wage expectations than Han or Muslim migrants of a similar skill level. Moreover, when they do engage in wage employment, they tend to be selective about their employment options. Even in farming areas, where wage employment is more actively sought than in pastoral areas, targeted occupations tend to be in the middle to upper strata of the local labor hierarchy, such as drivers, carpenters, teachers, or officials, that is, those that do not offend notions of dignity and/or that bear

310

Chapter 7

status.35 These attitudes contrast with their apparent income poverty, their very poor levels of schooling (among the worst in China), and the obvious need for low-skill off-farm employment in the Tibetan areas. A similar disinclination is commonly observed among many Tibetans with respect to reliance on trade or commerce as primary sources of livelihood. For instance, while petty trading forays might be a popular pastime among many farmers and herders, many rural Tibetans whom I interviewed expressed the prejudice that cheating and trickery are required for success in business. Business as a full-time occupation therefore tends to be stigmatized as un-Buddhist, the unworthy preserve of Muslims and other migrants (including Tibetans from other counties, such as Ngawa traders in Golok). There are obviously exceptions and deviations from the norm, particularly within the context of rising inequality, such as in Ngawa County (Ch. Aba) in Sichuan, which is known for its successful business community. Indeed, there is a very successful Ngawa business community in Lhasa. I met one Ngawa businessman from this community who had made his money building accommodation for Chinese railway construction workers. While there are probably many similar local exceptions throughout Tibet, the general rule consistently expressed to me was that business—especially small-scale commerce—is not considered as a particularly dignified activity among Tibetans, although some business activities, such as those related to transportation, are definitely considered as more appropriate than others. These generalizations are meant to capture the gist of broad labor trends that I observed among Tibetans during fieldwork and that were consistently confirmed through numerous interviews with a wide variety of respondents, from farmers and herders to businesspeople, scholars, and government officials, in the Tibetan areas of the TAR, Qinghai, Gansu, and Sichuan. However, data is not available to quantify these observations given that wage data below the level of relatively privileged employment in urban units is not available in the official statistics. There is also little, if any, ethnic disaggregation of wage or employment data, and temporary migrants are not usually included in any case. I must therefore proceed through my qualitative fieldwork reflections. This employment behavior was most apparent in the pure pastoral areas, such as Golok and parts of Yushu in Qinghai, or Nagchu in the TAR, all of which are among the least developed of the Tibetan areas in terms of infrastructure, education levels, per capita GDP, and other conventional development measures. However, herders in these areas by and large systematically avoid menial wage labor or petty trades. As a result, most of the low-skilled occupations at the lower end of the labor hierarchy are filled by Han migrants, typically from Sichuan, and to a lesser extent by Chinese Muslims or Tibetans from farming areas. This includes rudimentary processing activities of pastoralist output such as the stretching and tanning of



Subsistence Capacity and the Material Foundations of Resistance 311

yak and sheep hides. Muslims also tend to occupy the intermediate positions of petty trade, commerce, and service occupations such as catering and butchery, as discussed in chapters 2 and 8. Some Han also fill these intermediate roles alongside their obvious dominance in positions of authority. Indeed, Han who are involved in business are often related to those in government.36 Perhaps more significantly, these observations in the pastoral areas hold even when employers are Tibetan, thereby countering the argument that these patterns are not the results of “choice” but are instead caused by various subtle forms of discrimination. Discrimination does exist, such as in the contracting of government infrastructure projects to Han construction companies as discussed in previous chapters, although here we are considering how Tibetans themselves tend to discriminate in their employment choices. For instance, many of the relatively simple manual tasks related directly to Tibetan activities of animal husbandry that I observed during visits to Golok, Yushu, and Ngawa in 2004, such as the building of mud enclosure walls for corrals, were being performed by Han Sichuanese laborers hired by Tibetans. Numerous examples of monastery or temple construction that I visited throughout Qinghai were also based on Han Sichuanese laborers. Monastery managers, lamas, or senior monks typically oversaw the operations, but work was generally subcontracted to Han construction companies. An assistant of a monastery manager overseeing the construction of a particularly large prayer hall in a monastery near Darlak in Golok, which I visited several times in spring 2004, explained to me that Tibetans were not involved in the construction because the idea of working for the monastery for a wage is considered undignified. Notably, in this same case, a large group of Tibetan pilgrims from Gyarong (in Kardze, Sichuan) were involved in the construction of a prayer wheel wall in the same monastic compound. However, their work was voluntary: work for the monastery was provided so long as it was seen as a religious act of offering, along the same lines as offerings of food or other forms of wealth. I observed similar patterns in the construction of new temples farther north in the mixed pastoral-farming areas, such as in Rongwo Town in Rebgong. Likewise, in many Tibetan businesses, menial wage jobs were often filled by Han Chinese laborers. For instance, during a visit to Ngawa County in May 2004, I met a confidant of one prominent local Tibetan businessman who ran a popular hotel and restaurant in the county town. The businessman, a known nationalist who had recently been detained, nonetheless employed mostly Han migrants as cooks and waiters, while employing Tibetans, most of them from his extended kin, in managerial positions. His confidant explained to me that local Tibetans are not interested in the more menial range of jobs, at least not for long-term employment. Instead, they are more interested in quick cash-generating activities, such as digging for caterpillar

312

Chapter 7

fungus or ad hoc petty trading forays to Chengdu.37 He also noted that there is a widespread perception among Tibetan employers that locals cannot be relied on, even if they are interested in the menial jobs, given that they would most likely be hired through family networks. This would make them more expensive and more difficult to discipline than Chinese migrants, who are perceived as cheaper and more reliable. Therefore, for the menial range of jobs, this Tibetan businessman preferred to employ Chinese migrants.38 From another angle, conceptions of status also appear to play a strong role in guiding employment behavior. For example, in one successful Tibetan-owned car repair business in Golok that I visited in 2004 and 2007 and that employed Tibetan workers, the owner had previously worked in the Public Security Bureau of the county, was married to a Chinese woman from Xining, and, as it happens, had a daughter who had been recognized as the reincarnation of a yogini. In discussions with numerous Tibetan farmers, herders, workers, and scholars in Qinghai, the consensus was that if a Tibetan (i.e., Tibetan man) were to undertake the discipline of regular wage employment, the status of the employer would be vital. Working for a small business, whether Tibetan or Chinese run, would be deemed degrading, whereas working for a powerful strongman, a monastery, the government, or above all, an international NGO, would better fulfill a sense of status, even if the wage earned in some of these cases might be less than the wage offered by the small business (albeit wages in the government or in international NGOs were also some of the highest on offer). The one consistently glaring exception to these observations regards gender: young women appear to transgress the stigmatism directed toward low-wage menial employment in the service sector or in small businesses much more readily than the general Tibetan population, particularly before they have married or started to have children. For instance, in the case of the Ngawa restaurant mentioned above, the one Tibetan working as a waitress in the establishment was a young woman from another county who had been orphaned and had shown strong dedication to engage in disciplined work. Similarly, young women were employed in the hotel for room service. This also extends to non-Tibetan businesses; many, if not most, of the Han businesses operating in the Tibetan neighborhoods of Lhasa or Shigatse, such as small restaurants or shops, typically employ one or several young Tibetan women or girls, no doubt to service the Tibetan clientele patronizing these businesses. Young men are rarely if ever seen occupying the same positions. This gender dimension deserves further research and is not explored here, although it appears to stem from the expected social roles and status of girls and young women, which facilitates their entry into varieties of urban service work, such as domestic care work or sex work. It is also plausible that such subordination of young women’s work serves to subsidize and buttress patriarchal



Subsistence Capacity and the Material Foundations of Resistance 313

notions of dignity, including the ability of Tibetan men to desist from work, particularly at a time when such patriarchal notions are under strain in the context of rapid socio-economic transition. The clear ethnic stratification of employment breaks down somewhat in the farming areas, where Tibetan farming families generally show more willingness to participate in low-wage employment. Nonetheless, even in these cases certain types or conditions of wage employment are deemed more acceptable than others. One common area of low-skill employment in the farming areas is construction work, such as road construction, or in one case that I observed in Mangra County in Qinghai, the construction of a small hydroelectric project. In these cases, work was performed outdoors, it was seasonal or temporary, and it involved large parts of the local community, thereby conforming to a sense of communal labor, akin to a harvest season, where all are involved and thus none are stigmatized. However, even in these cases, Tibetan workers generally appear to have higher wage expectations than Han Chinese migrants, as well as higher expectations of non-monetary conditions of employment, such as shorter working days, less intense work discipline, more frequent holidays, and tolerance of absenteeism, as mentioned above in the case of the Ngawa restaurant. Higher wage expectations are less evident in the TAR due to the high salaries offered to Chinese migrants on many worksites (such as the railway) from which Tibetans are more or less excluded. Nonetheless, differences in wage expectations are much more obvious outside the TAR, given that typical wage rates for low-skilled Muslims or Han in places like Haidong in Qinghai or Linxia in Gansu can be very low. Indeed, one Hui high school graduate whom I interviewed in Linxia Town in July 2004 was earning 200 yuan a month as an office assistant for a local construction company, whereas his co-worker, a foreman with ten years of experience working for the company, earned 800 yuan a month. The relatively skilled Salar factory workers in Xunhua County discussed in the next chapter earned about 500 yuan a month for working eight hours a day, six days a week. In comparison, unskilled Tibetans (men and women) working on the dam construction in Mangra County mentioned above earned 540 yuan a month under fairly flexible conditions, which was a typical wage for most Tibetans I met in rural Qinghai doing unskilled construction work in 2004. These observations were instantiated by a construction project that I investigated in Lhasa in November 2004. The government had started a renovation project along some of the main tourist streets of Lhasa in September 2004, which involved replacing the outer facades of new, recently constructed buildings with molded Tibetan-style fascia.39 The work was contracted to Chinese construction companies and employed only Han Chinese migrants, even for jobs such as painting the Tibetan-style motifs on the new facades, despite the fact that the skill level was perfectly suited for

314

Chapter 7

many of the local Tibetan artisans accomplished in the construction of Tibetan facades and the painting of authentic Tibetan motifs. The contracting of such projects was obviously biased toward these Chinese construction companies. However, one Western INGO worker who had been living in Lhasa for several years, was fluent in Chinese, and had done some fieldwork among these Chinese companies and workers explained to me that another reason for the lack of Tibetan involvement was that Tibetan construction companies and their workers or artisans are more expensive and take longer to complete a job than the Chinese companies and migrants, even though their work might be of higher quality. Subsequently, I questioned several Tibetan officials and scholars about this and they all generally concurred. While these considerations do not excuse the blatant discrimination against Tibetan businesses in construction contracting and the lack of preferential treatment for Tibetan workers in Lhasa, they nonetheless reveal interesting aspects about the wage and work expectations of Tibetan artisans and construction workers in Lhasa. A similar problematic also faces various Tibetan attempts to run factories, such as textile or carpet-making factories, particularly in the face of increasing migrant competition in these activities.40 It is not the case that Tibetans cannot run effective companies or factories (as is often argued). Rather, many Tibetan businesspeople exiled in Kathmandu have set up very effective, competitive, and profitable carpet or textile factories. Yet in these cases, most of the manual labor is performed by Nepali workers, or even by Biharis migrating from India, whose wage and work expectations are unrelated to the conditions of relative affluence prevailing in the Tibetan exile community.41 This point is explored further in the next chapter, although the basic insight is that the wage and other work expectations of the typical Tibetan workers available in these cases would render such operations uncompetitive, and hence such workers are generally avoided by these Tibetan businesspeople, in deference to cheaper and more disciplined Nepali or Bihari workers (or Han in the case of western China). These dilemmas were succinctly explained to me by Tashi Tsering, an enigmatic Tibetan intellectual, entrepreneur, and philanthropist in Lhasa.42 In arguing that Tibetans were unable to compete with Chinese migrants in Lhasa, he noted: The shoe menders around my house have been replaced in recent years by Chinese shoe menders. Since the 1960s the shoe menders had been Tibetan. There are several reasons for this. One is that the Tibetans are lazier, which comes from social legacies, or the money consciousness of Tibetans versus the Chinese. Another is the hardship of the work that Chinese migrants do. Tibetans cannot compete with the incredibly hardworking Chinese shoe menders. The Chinese shoe menders are incredibly hardworking. Now they might be earning



Subsistence Capacity and the Material Foundations of Resistance 315 30 to 40 yuan a day, although they would have started by earning much less. Now they might stop for a short lunch, whereas before they were not even stopping for lunch. This process of replacement by Chinese has happened to every Tibetan trade, including the shop and restaurant businesses.

Whether or not this contention of Tashi Tsering was an accurate representation of the actual processes of ethnic substitution that had occurred in various sectors of employment in Lhasa, it is an opinion that was shared by many of the other educated Tibetans whom I interviewed (all of whom, of course, would never consider working as a shoe mender—the class judgment in these cases is notable). It is nonetheless notable that Han or Muslim Chinese migrants have ended up occupying much of the low-skilled off-farm employment that has been generated by recent economic growth in Tibetan areas.43 Garbage collection offers an extreme example. In all the Tibetan towns of Qinghai that I visited, this trade was almost exclusively plied by Han migrants. The only exception I saw was in the Tibetan neighborhoods of Lhasa and Shigatse, where uniformed Tibetan women worked as street cleaners. However, they were evidently public employees, and thus their work was relatively privileged given the benefits that can accrue from urban public employment in the TAR (however, they may have also been among the contract workers who were facing severe retrenchment in the TAR in 2002 and 2003, as discussed in chapters 5 and 6). Thus, while the dominance of non-Tibetan migrants in low-skilled employment and small business has given cause to accusations of population invasion and employment discrimination, to a certain degree Tibetans have been complicit in such stratification. Indeed, in another example from my fieldwork in Rebgong, a Hui Muslim high school teacher told me that local Muslim businesspeople in fact borrow from monks and lamas in the nearby monasteries, or else from wealthy local lay Tibetans.44 Such exchanges reveal how Tibetan elites with surplus capital, monasteries and lamas included, prefer to act as financiers rather than engage in the messiness and visibility of business. It could even be argued that this arrangement confirms a sense of hierarchy wherein Tibetans (at least, elite Tibetans) are at the top, being served by Han and Muslims. Livelihood Strategies and Ethnic Labor Stratification These attitudes toward work are reflected in typical allocations of family labor within rural Tibetan households since the reform period. A stereotyped, albeit common, caricature illustrates this, based on many of the rural families I visited in Qinghai, Gansu, and Sichuan and supplemented with my experience of living among newly arrived Tibetan refugees in India

316

Chapter 7

and Nepal from 1995 to 2001. If a household has four or five children, typically one child is kept on the homestead. This child, male or female, receives little schooling (currently this would be limited to the minimum required by the government), marries a similar person in the community, and together they maintain the household and care for the elderly. One or two children pursue schooling further, depending on the size, attitudes, or wealth of the family, with the hope that they complete secondary school and obtain a respected job, such as working in the government or as a teacher. One child obtaining such work can effectively secure the future livelihood of the family. Another child might be sent to a monastery or nunnery, or else into exile with the hope of meeting the Dalai Lama, receiving an education in one of the “Dalai Lama’s schools,” or else making it abroad through contact with Westerners (or Japanese), and possibly even marrying one along the way. Beyond the political irony that such households can end up with a monk, a public employee, and a refugee in the same family, these typical strategies carry several other implications. First, although each household might have one or two children pursuing post-primary secular education, this only leads to a rate of about half of the children receiving such schooling. In other words, this strategy is not contradictory with sustained high rates of illiteracy. Second, quick cash-earning activities, such as caterpillar fungus–digging or seasonal construction work, do not necessarily obstruct these strategies, so long as they do not constitute long-term employment or undermine the maintenance of the asset base. Finally, long-term low-wage employment is generally not on the list of options despite low levels of education, unlike in poor rural Han or Muslim households, where children are typically expected to enter wage employment or commerce after a certain amount of education.45 These strategies do target long-term wage employment, albeit at a higher level of the labor hierarchy or with higher wage and non-wage expectations than would be expected of relatively income-poor rural households with very low levels of schooling on average. In other words, a “proletarianized” working-class attitude is typically not observed among most rural Tibetans, whereas it is among many rural Han Chinese or Chinese Muslims. Deeply rooted cultural conceptions of hierarchy and dignity obviously play a central role in these employment attitudes. Tibetans, particularly in Amdo and Kham, and especially Tibetan herders, essentially conceive of themselves as the farmer/herder-warriors, guardians of the land, and are ready for mobilization into warfare whenever need be.46 This realization actually came to me most strongly in the course of a debate between an Australian woman and one male Amdo Tibetan high school teacher in July 2004 in Chabcha about the gendered division of labor in Tibetan households. His rationalization for the fact that Tibetan men do little or



Subsistence Capacity and the Material Foundations of Resistance 317

no housework was that the responsibility of men was to go to war (or otherwise protect the household, village, or tribe). Surprising though this was given that he and his wife were both high school teachers and he could barely do a single chin-up, for me it was a remarkable gendered expression of a Tibetan sense of hierarchy. And despite his own lack of physical robustness (albeit he did enter into a bar brawl a few weeks after this conversation), the point about going to war still holds much salience in the rural areas today, in light of ongoing feuding in many pastoral and mixed pastoral-farming areas.47 Accordingly, within this sense of hierarchy, trade and commerce are considered lower positions to farming and herding, and menial wage employment is downright degrading (unless it constitutes service for a monastery, a powerful leader, or a status-bearing organization). While these cultural conceptions of hierarchy help to explain typical employment preferences of rural Tibetans, the importance here is not in the cultural variations of attitudes toward work per se, but that Tibetans generally have the relative freedom to choose on the basis of cultural criteria, rather than being driven into alienating wage labor through impoverishment. This is in contrast to many areas of the world where groups might have held similar notions of labor hierarchy yet have been compelled to demote by poverty. In other words, Tibetan herders and farmers manifest a capacity to choose according to their cultural preferences. This appears to concur with the analysis of the previous section that assets and subsistence capacity seem to offer better indications of the wealth factors underlying these employment decisions than conventional income measures.

Subsistence as Wealth This question of what drives the employment decisions of rural dwellers touches on debates over the rationality of “peasant” behavior in much of the development literature. The standard position taken by much of the (officially sanctioned) Chinese literature on Tibet is that, similar to modernization theories of both conservative and Marxist flavors, the behavior of rural Tibetans is rule bound by customs and institutions, which adapt poorly to modern market mechanisms due to an absence of rational or possessive individualistic norms. The absence of such “market economy” norms in turn inhibits investment and accumulation, particularly in comparison to their Chinese lowland cousins who have been much more adept at taking advantage of market incentives in the reform period. The implication is that Tibetans need to be tutored by the example of their more “advanced” Chinese cousins, primarily through education and migration-induced emulation.48 Debates among Chinese scholars or policy makers generally do not dispute the underlying assumptions of this narrative, but focus on whether

318

Chapter 7

tutelage should take place gradually, in a carefully regulated manner, or through fast-track, leap-over strategies of uninhibited integration into the rest of China.49 As clarified in the first section, official statistics do not appear to support the assertion that, on average, Tibetan farmers invest any less than Chinese farmers elsewhere in China. Thus, whatever the cultural preferences of Tibetans, these preferences do not appear to obstruct thrift and productive accumulation, at least not within agriculture or any more than they would elsewhere in rural China. Instead, the relatively greater asset wealth or subsistence capacity of rural Tibetan households suggests that the various dimensions of wealth in subsistence-based rural economies are not being measured properly by conventional income measures and that Tibetans are actually wealthier than they appear in the income statistics, thereby helping to explain their apparently paradoxical employment behavior described in the second section. However, even when these points are considered, there is still contention in the literature over these alternative dimensions of wealth given that they are not seen as appropriate bases for capitalist accumulation, or accumulation that is based on an increasing productivity of labor rather than simply focusing on preserving asset wealth or else accumulating wealth in an unproductive manner. Examples include the tendency often observed among pastoralists of maintaining large herds with limited culling rather than moving toward commercial forms of ranching, or among farmers of investing agricultural profits and even migrant wages in rural housing and conspicuous consumption rather than in agricultural modernization or business development. Obviously, these are stereotypes; evidence from fieldwork, such as in Goldstein et al. (2008), indicates that considerable investment in rural Tibet has also taken place in various off-farm businesses, particularly in those related to transport or housing construction. Substantial household investments in education should also not be overlooked. Nonetheless, the question is often asked, why do we often see rural folk maintaining or even reinforcing their basis in subsistence when given the freedom, even though we (and they) know that this is a dead-end option with respect to the future of wealth generation in the economy? This question underlies much of the theoretical presumptions that influence policy making in Tibet, as it does in much of the developing world. Rationality and Culture The standard approach taken in the Marxist-inspired Chinese literature to explain rural development in Tibet usually follows a culturalist or “stages of growth” modernization theory paradigm. This is perhaps best represented by the work of Xiaoqiang Wang and Nanfeng Bai (1991 [1986]), as



Subsistence Capacity and the Material Foundations of Resistance 319

discussed in chapter 4. In their groundbreaking book, The Poverty of Plenty, they refer to a “two-way initiative” that started in the early reform period. One way was toward the “commodity economy,” that is, intensified capitalist modes of accumulation, and the other way retreated back into the socalled “natural economy.” The latter initiative, which they argue characterized most of the ethnic minorities in western China and especially Tibetans at that time, was due to “the intrinsic determinant of backwardness” or “the poor quality of human resources” (23). “This is the immature social system and its vicious circle of poverty” (65). The more a region is backward, the less likely catching up will be achieved through simply “relaxing controls” (175), given that the resources needed for the development of a commodity economy will be diverted to the natural economy, thereby undermining wealth creation (32). They therefore advocate that the real challenge in China’s backward regions constitutes “reversing the attitude of the local inhabitants towards social wealth and changing their traditional ways of exploiting natural resources” (92). They also explicitly address the paradox of income poverty amid a strong asset base, albeit in a derogatory tone. They note that impressive figures of wealth can be achieved in backward regions “at the expense of ruining resources through traditional means of exploitation. From these indicators it is impossible to tell if the economy is being transformed from a natural economy, or what the social implications of individual prosperity in backward regions might be.” They substantiate this assertion with several anecdotal cases from Qinghai.50 For example, a herder who was very wealthy by Chinese standards “never traded his livestock and could not even afford a new pair of shoes, while all he lived on was gruel” (91).51 Similarly, a Tibetan county that was apparently the richest in China in the early 1980s on the basis of per capita livestock holdings had a “way of life and mode of production among the people [that] showed little change from the remote past” (91). This assessment of Tibetans remains common in China, as it is in the idealized caricatures of Tibetans as peaceful non-materialist ecologists in much of the Western media. Wang and Bai relate such Tibetan naïveté to the ongoing influence of “such negative religious ideas as withdrawing from the world and fatalism,” causing them to reject modern agricultural techniques due to a variety of superstitious concerns, or else using new wealth to build temples in competition with other newly rich (34). Many of these views were echoed in my fieldwork by a variety of Chinese scholars, officials, or laypeople, and some modernization-advocating Tibetans, who decried that poor Tibetan rural communities consistently pour their wealth and savings into monasteries or religious monuments rather than productive investments, thereby perpetuating their cycle of poverty. Officially, the Chinese government also adopts this rhetoric by explaining that rural

320

Chapter 7

poverty in Tibet stems from a slow cycle of low output, low investment, and low accumulation, leading again to low output (PRC 2001, question 22). These perspectives generally mistake the symptoms of distorted development for causes, given that the impacts of disempowerment are utterly ignored. For instance, Wang and Bai’s far more rigorous analyses of the failure of aid and various economic projects in the TAR, as discussed in chapter 4 (which they also blame on Tibetan backwardness), deals mostly with the state-run economy from the 1960s to the early 1980s. Notably, they ignore the late introduction of communes in the TAR from the late 1960s onward, followed by failed attempts at large-scale crop diversification, ecological disaster, and widespread hunger by the late 1970s. They also ignore the earlier and more radical introduction of communes in the mid-1950s in the Tibetan areas outside the TAR, which took place despite the fact that these areas had not yet undergone the intermediate stages of so-called “democratic reforms” as elsewhere in China. This was followed by widespread Tibetan uprisings in Sichuan, Gansu, and Qinghai and an intense government counterinsurgency. Moreover, Qinghai, Gansu, and Sichuan were among the provinces in China worst hit by the famine of the Great Leap Forward in 1959-1961.52 In this context, it is difficult to argue that an adult male herder interviewed in Qinghai in 1984, in a county that had experienced uprising, counterinsurgency, and famine only a few decades earlier, not to mention the subsequent Cultural Revolution, and that was currently in the midst of de-collectivization, had a way of life that “showed little change from the remote past.” Indeed, the individual household responsibility system itself was a fundamentally new form of rangeland management, based as it was on individualized land use and thus differing from both traditional and commune pastoral systems, which were both organized on collective principles.53 In contrast, farming areas generally had some experience with individualized land use prior to collectivization, like elsewhere in China. Moreover, most Tibetans were themselves completely divorced from any sense of agency within these processes of radical transformation, which would tend to reinforce a conservative response to change, in contrast to China, where both revolution and reform were indigenously generated processes par excellence. In addition to these Tibet-specific considerations, Wang and Bai do not consider the effects of the price scissor model of urban-rural development, whereby agricultural prices were suppressed until the 1980s in order to subsidize urban industries. The associated policies are widely regarded as having caused heavy productive disincentives in western China throughout this period. Nor do they consider the uneven character of regional development policy, even though these were subjects of heated politicized debates during 1980s, as discussed in chapter 2. Therefore, the inefficiencies of the



Subsistence Capacity and the Material Foundations of Resistance 321

highly state-centralized and militarized economy of the TAR and other Tibetan areas in these years can hardly be attributed to the entrepreneurial disability of local Tibetans. Perhaps the most significant refutation of Wang and Bai’s argument is the fact that rural Tibetans are thrifty and productive, as discussed in the first section, and they are price responsive with regard to both commodity prices and wages. Tibetans can be seen to quickly shift their livelihood strategies in response to price changes, such as in the case of the booming caterpillar fungus trade or in relation to collapsing wool prices versus rising prices for meat and hides. Indeed, based on his fieldwork in 2005 in a pastoral area of the TAR where illiteracy remains very high, Melvyn Goldstein noted that nomads had been using a subsidized loan program to innovate seasonal ranching, purchasing young animals in the spring, fattening them in the summer, and selling them for meat and skins in the fall. The reason for this shift to a commoditized form of ranching was precisely because prices for meat and skins had been very favorable in recent years, whereas the prices for wool had been very poor. He also noted that local township officials had been actively involved in promoting and supporting the changes.54 This latter institutional dimension is suggestive of a Polanyian view that the behavior of rural Tibetans—or of anyone else— cannot be judged independently from the broader context of governance and/or state control. In other words, given the right incentives, there is no particular reason why Tibetans would not display a certain degree of market rationality, even under conditions where they continue to have a “poor quality of human resources.” Rather, this analysis suggests that Tibetans will be economically rational (in a neoclassical sense) when price signals are in harmony with their cultural norms, particularly with respect to labor hierarchies. Cultural norms in turn are crucially rooted in the principle of maintaining a sufficient degree of subsistence capacity. Moreover, subsistence modes can co-exist or switch back and forth with more market-oriented economic behavior, even within a single household. This seems to imply that subsistence production is part of a larger set of choices within calculated livelihood strategies, rather than simply constituting a deterministic outcome structured by physical, human or social capital, culture, and so forth. Subsistence and Strategy Returning to the discussion of subsistence capacity at the end of the first section, if and when a condition of absolute subsistence is met, household members have the individual or collective freedom to allocate their labor as they see fit, regardless of immediate considerations of productivity or marginal returns. Conversely, the incapacity to meet subsistence needs would

322

Chapter 7

compel farmers or herders to accept wage labor, in accordance with various theories of labor commodification. However, several distinctions must be made from the classic European cases of labor commodification that informed such theories. The classic cases were in the throes of industrialization and were at the colonial core of the international economy, with the political and administrative power to control and restrict migration from tropical regions, all of which worked to support the higher wage expectations in these regions. Tibet holds an inverse position: it is on the periphery of Chinese industrial power, with a paucity of economic opportunities outside agriculture, and no ability to limit in-migration. Thus, there is a gaping asymmetry between higher wage expectations and the sheer scarcity of higher remunerated economic opportunities that would conform to these expectations. Moreover, these disjunctures are exacerbated by conditions of disempowerment. In this sense, there is some validity in the insight of Wang and Bai that de-collectivization produced a forward movement into commoditization in the Han areas and a backward movement into the “natural” economy in the Tibetan areas. However, the mistake of Wang and Bai was to attribute these movements to “intrinsic backwardness.” Instead, the introduction of the individual household responsibility system in the early 1980s effectively caused a “peasantisation”55 of Tibetans; for the first time land use was divided up into small, relatively equal individuated parcels, something that had never existed under either the collective economy or the “old society.” This obviously happened all over China, but elsewhere in China the individuated plots were generally not large enough to sustain subsistence needs, even if labor productivity was higher on average. Plots in Tibet have generally been large enough to sustain subsistence, thereby helping to sustain rural Tibetan conceptions of labor hierarchy during the reform period despite the rapid marginalization of rural Tibetans from the dominant sources of wealth and accumulation in the regional economy. Along similar lines, an extension of the “commodity” or cash economy is not necessarily antithetical to these processes. Tibetans are involved with the commodity cash economy and thus they cannot base their livelihoods solely on subsistence household production (if ever they did). Today, this is partly because they purchase, consume, or invest in modern goods, be they mobile phones, motorcycles, or fertilizers, among many others. Perhaps more importantly, rapidly increasing fees for education and health care (up until the mid-2000s, particularly outside the TAR) have also acted as strong compelling forces for Tibetans to move into cash-earning activities, particularly in light of household livelihood strategies that target post-primary schooling of at least one child. A scarcity of rural cash-earning opportunities in turn drives rural to urban migration.



Subsistence Capacity and the Material Foundations of Resistance 323

However, the cash imperative is often over-emphasized as a driver of labor commodification, given that it can easily co-exist with subsistencebased livelihood strategies. In fact, the combination is ideal. If subsistence capacity is sufficient, it offers choice as to how and when to engage in cash-earning activities. It offers autonomy from a dependence on regular low-wage employment, or from the forced selling of produce at inopportune moments, both of which can have impoverishing implications. The ability to wait thereby confers an ability to access better opportunities if and when they arise. In this sense, the subjective proclivity for subsistence within rural communities is not merely cultural or symbolic. It also carries important instrumental wealth functions, not only in terms of the risk-averse insurance principle as argued by authors such as James Scott (1976), but also in terms of providing a position from which market and other forms of economic opportunity can be engaged advantageously. Indeed, although Scott tries to explain the rationality of the peasant’s proclivity for risk aversion, he nonetheless portrays this rationality as resulting from a survival logic engendered by the condition of living on the verge of destitution. The interpretation here differs, along the lines argued by Polly Hill discussed below, in that the reinforcement of subsistence capacity can serve as an important wealthsupporting strategy within rural transitions into more marketized forms of production, exchange, and employment. In this light, the standard derogatory association of subsistence with poverty or traditional pre-market economic systems may actually be the inverse to how subsistence is subjectively experienced by rural dwellers. This is precisely the point of Hill (1986, 18-20) in her criticism of the conventional usage of the concept of subsistence in much of development economics. She argues that it is usually only rich rural households that can hope to attain a degree of self-sufficiency and she therefore critiques the common assumption that traditional agriculture is characterized by degrees of subsistence, which are inversely related to market integration, given that this assumption ignores the role of inequality within such agricultural systems. She points out that it is precisely the wealthy households that achieve some level of “subsistence” and that also enjoy the most lucrative non-farming occupations. This also relates to the point made by Hirschman (1970, 5-15) that the counterpart of society’s ability to achieve a surplus over subsistence is that society is not bound to a state of “permanently taut economy” but can afford and tolerate a certain degree of slack and lapses from full efficiency, contrary to the implicit assumptions underlying neoclassical economic models of perfect competition. From these perspectives, it is clear how achieving a minimum threshold of subsistence capacity offers a freedom to act, whether for profit or non-profitable social approbation, and that this freedom is both instrumental for and symbolic of wealth.

324

Chapter 7

Moreover, the commodification of certain product markets can actually reinforce resistance to labor commodification. For instance, the current boom in the caterpillar fungus trade in the Tibetan areas has certainly brought about a commodification of caterpillar fungus and significantly increased the cash income component of Tibetan rural incomes, particularly in remote pastoral areas that previously were the least integrated into the cash-based economy. This commodification of caterpillar fungus has thereby buttressed the ability of these same households to maintain a degree of autonomy from outright dependence on wage employment precisely because it provides a lucrative source of cash income in a manner that avoids wage employment and that does not undermine their subsistence asset base. Thus, to the disappointment of market fetishists, the extension of commodity markets will not necessarily cause a “capitalist” transformation of social structure. Rather, per the insight of Karl Polanyi (2001 [1944]), along with much of Marxist political economy, the fundamental transformation underlying market society is not found in the existence of commodity markets, which have existed for millennia, but in the creation of labor markets, which requires labor commodification. Historically, this has usually taken place through various structural and institutional processes that undermine rural assets and associated livelihoods, including the outright expropriation of assets, and thus undermines the ability of rural households to achieve absolute subsistence, thereby compelling them to search for wage employment at any wage and for any type of work.56

Conclusion: Subsistence, Resistance, and Double Movement While it is not clear to what extent rural Tibetans are poor, long-term dependence on low-wage employment does not seem to be a panacea. It will probably not compensate for their marginalization from the rapid economic growth in their regions, as discussed in previous chapters. It also certainly represents a demotion for many rural Tibetans within their conceptions of local labor and wealth hierarchies in comparison to the recent past, including in comparison to the collective period when most Tibetans were still farmers and herders and collectivization did not necessarily dramatically alter their traditional patterns of land management, which were also broadly collective, especially in pastoral areas. More importantly, participation in low-wage employment only marginally helps to overcome the growing gap between the average means and ends of rural households, given the escalating costs associated with post-primary education and health care, among other factors. The caterpillar fungus commodity boom has somewhat offset



Subsistence Capacity and the Material Foundations of Resistance 325

the impact of sharply declining terms of trade in traditional rural commodities during the reform period although, like all commodity booms, it might only be postponing adjustment to a later date. From this perspective, it is understandable why many rural Tibetans resist low-wage employment and place their bets on getting at least one child into some form of more gainful employment, at best in a privileged public-sector job, even if the odds for this option are increasingly stacked against them in such jobs, as discussed in the previous chapter. Against this backdrop, in certain pastoral areas of Qinghai, Gansu, and Sichuan, government strategies of poverty alleviation and environmental protection started to promote the direct dispossession of land assets from the early 2000s onward through a combined emphasis on the resettlement of pastoral communities into urbanized resettlements located in township and county towns. The more dramatic and highly publicized cases of such resettlement programs have been taking place mostly in Qinghai, particularly in the ecologically strategic Three Rivers Source area (Ch. sanjiangyuan). Admittedly, there has been a lot of confusion regarding these resettlements, such as the degree to which they are forced or voluntary, coercive or incentivized, complete or partial, permanent or reversible. There has also been conflation between these resettlements and the much more lenient “Comfortable Housing Program” in the Tibet Autonomous Region, which does not necessarily involve resettlement. Nonetheless, the resettlements in the Three Rivers Source area have been of the most severe varient, that is, partial or full relocations of pastoral communities to urbanized settlements in small towns. Similar, although more gradualist, programs have been taking place in the pastoral areas of Gansu and Sichuan, as well as in other parts of Qinghai.57 The poverty dimension of this policy derives from the conventional wisdom in China that greater exposure to off-farm employment leads to substantial improvements in rural incomes, as discussed in the first section. Similarly, flooding in lowland China is usually blamed on overgrazing and rangeland degradation in the highlands (i.e., Tibet). Ergo, urbanization and urban expansion have logically come to play a central role in poverty alleviation strategies in the Tibetan areas since this time. However, as elsewhere in China and the world, urbanization makes sense as a strategy of poverty alleviation only insofar as urbanizing migrants are successfully employed upon reaching their destination, and at wages that compensate for their loss of rural livelihoods. Yet this employment consideration appears to be singularly lacking in much of these resettlements across eastern Tibet, particularly as the flurry of related construction activity has been mostly captured by Han construction companies and nonTibetan migrants. Rather, an almost perverse faith in “the market” predominates in local government circles, as well as among Chinese officials and scholars.58

326

Chapter 7

Exposure to the town environment and to non-Tibetan migrants is presumed to encourage herders to engage in trade and commerce, leading a lot of them to become petty entrepreneurs, despite the almost complete absence of productive secondary activities in the towns themselves, as discussed in the following chapter. Adjustment in the interim is to be supported by the wealth generated from selling off herds and supplemented by basic welfare contributions. Besides the fact that these views ignore or wrongly interpret a wide range of social and ecological issues, forced urbanizations are unnecessary given that Tibetans are urbanizing in any case, as discussed in chapters 3 and 5. They do not need government policy to encourage them to see the necessity of this course of action. According to the typical rural household livelihood strategy described in the second section and given the scarcity of non-agricultural employment opportunities in the rural areas, often more than half of the working members of a typical rural household are already effectively urbanized, at least on an occasional or seasonal basis, and this trend will most certainly increase over the course of the next generation. The key issue in this regard is the degree to which households can navigate these transitions in an autonomous and strategic manner, in ways that are deemed advantageous and dignified. The maintenance of subsistence capacity is crucial to allow for such navigation and is not necessarily contradictory with these transitions. The imposition of state paternalism, however, has the effect of potentially short-circuiting these self-determined indigenous livelihood strategies. To the extent that they undercut the subsistence basis of these strategies, government poverty alleviation and environmental policies effectively undercut the very means by which such rural Tibetans might transition in a relatively autonomous and dignified manner. Indeed, a much more effective way to promote urbanization along the pathways already chosen by rural households would be to heavily subsidize post-primary education and to return to a policy of guaranteeing public-sector employment for at least some of the high school and university graduates, particularly those from pastoral areas, as argued in the previous chapter. Instead, the current market rhetoric of creating a level playing field amid an absence of any serious consideration for sustainable employment creation runs the risk of creating a series of small-town ghettos in eastern Tibet, similar to the urbanization of aboriginal communities in Canada and the United States. Ironically, where ghettoization involves the wholesale sell-off of subsistence assets or marginal gains in wage employment and petty trade, it may actually manifest as decreasing income poverty given the weakness of conventional income measures to capture such economic transitions. In this perspective, the subsistence-based strategies that rural Tibetans devise and value themselves might actually be quite logical, least-worst



Subsistence Capacity and the Material Foundations of Resistance 327

options to ride through the dramatic transitional restructuring that has occurred in their regional economies since the mid-1990s. To clarify this point, the first section argued that standard income measures are not well conceived to reflect subsistence or asset-based wealth, nor transitions from more subsistence to more commoditized modes of production and employment. That rural Tibetans might actually be fairly wealthy relative to the rest of rural western China—at least in terms of household-level subsistence capacity—helps to explain the commonly observed resistance among Tibetans toward a dependence on long-term, low-skilled employment, as described in the second section. Even though this is hard to quantify, it nonetheless helps us to understand how there is an implicit disjuncture between local wages on offer and local wage expectations. Rural Tibetans certainly sense this disjuncture and resent it. This in turn leads many nonrural and non-local economists to condemn Tibetan resistance as irrational or backward, precisely because the disjunctures are hard to measure and thus not amenable to understanding through standard methodologies of economic analysis. The third section synthesized these insights by arguing that the autonomy achieved by maintaining an asset base at a level sufficient to meet minimum subsistence represents a fundamental source of economic opportunity given that it provides an ability to act strategically, primarily by avoiding dependence on subordinated forms of employment. Hence, rural dwellers naturally seek autonomy from such dependence even though social and economic structural changes might simultaneously undermine the sustainability of subsistence-based household livelihood strategies. This autonomy-seeking behavior can be seen as a micro-expression of the Polanyian double movement at the household level, whereby people continuously seek to re-embed themselves in an asset base as a means to stabilize and protect themselves against the dislocating processes of labor commodification. Subsistence therefore serves an instrumental as well as symbolic (cultural) wealth function within rural communities. Contrary to arguments that such valuation of subsistence is only a characteristic of those living close to destitution, this idea that subsistence is an instrument for wealth is comparable to the way that monetary wealth provides freedom within a modern, non-rural setting. Few would contest that the rich can do what they like with their disposable wealth regardless of whether their actions are deemed rational or irrational; this is the privilege of wealth. Similarly, the ability to be unemployed in most poor countries is actually an attribute of middle- or upper-class status; the truly poor have no capacity to be unemployed. They therefore must work, whatever the work and whatever the remuneration. This understanding of what it means to attain subsistence (and thus, in principle, to be a “subsistence farmer”) may not

328

Chapter 7

accord with the ways that non-rural academics or policy makers expect the “disciplined” or “deserving” poor to behave, but this is more likely due to a misreading of the clues rather than any developmental deficiency on the part of rural dwellers. These aspects are poignantly evidenced in the context of certain government policies that undermine or even dispossess rural assets in the name of poverty reduction or environmental protection, such as with the resettlement programs implemented in Qinghai, Gansu, and Sichuan since the early 2000s. The strategies of resistance that people adopt in such situations might not result in street protests although it is striking the degree to which the occurrence of self-immolations in 2011 and 2012 was contiguous with this area (see Fischer 2012a). Rather, strategies more commonly result in various ways of resisting the processes of being disembedded from their subsistence base in the rural economy, such as by leaving their livestock with relatives to take care of, splitting up their families and putting some people in town and keeping others in the pastures, or even abandoning the resettlements and reconstituting their herds in the pastures.59 Moreover, these strategies are more likely to be taken up by those who have some minimum threshold of subsistence wealth and who therefore have the most autonomy to lose through labor commodification or welfare dependence, and who face an implied demotion within the local labor hierarchy through the process of resettlement. In contrast, truly poor farmers or herders who have no hope for subsistence might already be much more socially conditioned or adapted to placing their household labor into wage-earning activities at the lower end of the non-farm labor hierarchy, or else to becoming dependent on the welfare payments associated with these programs. Indeed, some research on these resettlements in Qinghai has indicated that relatively poor herders, with previous herd sizes too small to guarantee a minimum level of subsistence, were much less dissatisfied (or even happy) with resettlement, in comparison to richer herders who were the most disgruntled because they fell from an independent and dignified subsistence situation to one dependent on government welfare. This appeared to be the case even though the poorer herders were generally worse off in the small towns than the rich ones given that they had started with much less capital from the sale of their herds.60 This example demonstrates how we need to reverse our way of thinking with respect to the association between deprivation, grievance, and resistance. These insights are important from the perspective of this research given that the material basis for the double movement can be seen to derive from wealth rather than poverty. In other words, resistance is conditioned on having some basis of wealth from which to resist. Resistance is not necessarily an attribute of the coping strategies of the extremely poor but is most likely to take place among the social strata who have some wealth basis



Subsistence Capacity and the Material Foundations of Resistance 329

from which to resist and who potentially have the most to lose through structural transformations that undermine their asset base. It is precisely for this reason we must pay attention to the exclusions that occur among the non-poor in order to understand the social responses that exclusion might incur.

Notes   1.  This is in reference to Scott’s thesis in The Art of Not Being Governed (2009), which evolved out of elements in his earlier work in The Moral Economy of the Peasant (1976), Weapons of the Weak (1985), or Domination and the Arts of Resistance (1990).  2.  As noted in the previous note, much of Scott’s work was centrally concerned with this question. For a classic exploration of this issue in the Western Marxist literature, see Brenner (1976; 1977). Much of the contemporary Marxist-inspired literature on the “agrarian question” is concerned with the question of how and why the “peasantry” still exists despite theory suggesting that it should have been absorbed by the progression of capitalist development, as in classic capitalist transitions.   3.  For the classic expression of functional dualism, see de Janvry (1975; 1981), based on his study of rural development in Latin America and in a context where rural labor is semi-proletarianized, meaning that rural workers in the capitalist sector maintain a basis in subsistence agriculture.  4.  See chapters 2 and 4 in Fischer (2005a).   5.  I have inflated past values to constant 2010 yuan, which increases the real (or constant) value of past average rural incomes relatively more in provinces with higher inflation. Calculating from annual CSYs, the cumulative rural CPI from 1990 to 2004 was 203 for all China, 221 for Qinghai, 236 for Gansu, 224 for Sichuan, and 254 for the TAR (1990 = 100).   6.  The reported surge in official TAR incomes in the early 1990s might also reflect some data manipulation, given the political sensitivity of the region following the uprisings in Lhasa in 1989. TAR rural incomes were not as high relative to other provinces in the 1980s.   7.  For details on the changing measurement of “income in-kind” or “own production” in the 1990s, see Fischer (2005a, chapters 1 and 4).   8.  Provincial cost-of-living data is not publicly available in China. However, it is easy to ascertain from fieldwork that price levels in the TAR are significantly higher than in the rest of western China. For instance, using the starting fare of a taxi in 2005 as a very rough barometer, this fare was five yuan in Chengdu, six yuan in Xining, and ten yuan in both Beijing and Lhasa.   9.  Two agronomists working in the TAR, whom I interviewed during fieldwork in 2004, one Western and one Tibetan, confirmed these levels of productivity. 10.  This was not indicated in the explanatory notes of CSY (2011, chapter 13). Earlier yearbooks differentiated per capita output from output per agricultural laborer, with the latter showing much higher output, although with similar trends as the former (see Fischer 2008e, 11). In that case, the per capita output is probably based a more general measure of the rural population.

330

Chapter 7

11.  Most income decomposition analyses in China point to this clearly. For instance, see Khan and Riskin (2001, 30) or Goldstein et al. (2003). 12. These proportions are calculated from CSY (2011, table 10-22) and CSY (2001, table 10-19). 13.  On land use in the TAR, see Tashi et al. (2002). 14.  In discussions with Melvyn Goldstein, it seems that there is no reliable population estimate of the number of pure pastoralists in the TAR, although they appear to account for much less than half of the rural population. 15.  On rangeland conflicts, see Pirie (2005) or Yeh (2003b). I observed pasture rentals in 2003 in Sogwo (Ch. Henan) County, Qinghai, a pastoral Mongolian county bordering Golok in Amdo. 16.  Note that 15 mu = 6 acres = 1 hectare. 17.  Interview with two officials from the Xunhua agricultural department, Xunhua County, June 2004. 18.  Interview with two county officials, Sogwo (Ch. Henan) County, July 2003. 19.  Personal communication, 3 February 2006. 20.  For instance, on Sahelian pastoralists, see Swift (1977) or Niamir (1990). For a rich analysis of the relationship between assets and consumption in pastoralist Golok, see Costello (2003). For an extensive analysis of the importance of herd sizes and why pastoralists tend to keep the herd sizes they do, see Yundannima (2012). 21.  For instance, this was the opinion of Melvyn Goldstein, expressed to me in discussions during 2005 and 2006 about his recent fieldwork in both farming and pastoral areas of western Tibet. 22. Similar observations were also made by Rigdrol (single name), a Tibetan NGO staff from Golok, in a presentation at the 11th IATS conference in Bonn, August 2006. He also noted qualitative aspects, such as the fact that pastoralists herding in areas under ecological or population pressure also tend to have poorerquality herds, such as thinner animals and poorer-quality wool. 23.  These examples were also mentioned by Rigdrol (2006). 24.  For example, during fieldwork in Tranggo, Kardze Prefecture, Sichuan, which is a mixed farming-pastoral region that is fairly densely populated (compared to purer pastoral areas), one family (or mobile household) that I interviewed in August 2004 had about 25 to 30 animals (numbers cited were not precise), including 4 to 5 dzo (a cross between a yak and a cow) and 10 to 15 yak. This family did not characterize itself as particularly poor relative to the pastoralist norm in this region. Nonetheless, its livestock holdings were less than the official livestock poverty line discussed in the following section. 25.  CSY (2010, tables 10-26 and 10-27); see Fischer (2008e, 19) for the 2004 data, which was calculated from CSY (2005, tables 10-26 and 10-27). 26.  Calculated from CSY (2011, table 10-22); the provinces where rural households on average derived less of their incomes from salaries and wages than the TAR in 2010 were Inner Mongolia, Jilin, Heilongjiang, Hainan, Yunnan, and Xinjiang. 27.  Accordingly, see Goldstein et al. (2003, 769) for an interesting attempt to measure poverty according to subsistence-based measures, based on households not having sufficient grain either from their own fields or from wages, and having to borrow or receive welfare to meet their needs. This rate lay between the official income poverty rates for the TAR in 1999 measured by the official (9.1 percent) and



Subsistence Capacity and the Material Foundations of Resistance 331

absolute (24.5 percent) poverty lines (Fischer 2005a, 107). While these surveys did not cover any pastoral areas, the same team conducted longitudinal research on a pastoral area of western Tibet, which was also based on rough estimates of basic subsistence requirements. The findings were presented by Melvyn Goldstein at two conferences in 2006 (Bonn, August 2006; and Harvard, December 2006) and in personal communications from November 2005 to August 2007. They used the official county poverty line of 30 sheep equivalents per person (1 yak = 4 sheep; 1 horse = 6 sheep), which appears to be a standard throughout Tibet (although sometimes heads of animals are calculated without considering sheep equivalents). According to the sheep-equivalent line, they found that the proportion of poor households in their small pastoralist sample rose from about 5 percent in 1986 to 20 percent in 2005, while the proportions of middle and rich households fell. The rising poverty rates were due not to a falling average but rather to a sharp increase in inequality, particularly since the late 1990s. Goldstein also explained that, underlying these livestock-based poverty measures, the standard of living generally improved among these poor due to a variety of development and community interventions. In other words, their poverty line considered only asset (i.e., herd) wealth, not other forms of livelihood such as wages and government transfers that would have corrected some of the rising inequality in assets. Both poverty lines were not conventional in this sense, although they provide valuable insight into indigenous notions of subsistence wealth and poverty. One should of course be cautious in generalizing from their small sample or comparing their results across regions. 28.  A scattering of references to subsistence capacity can be found in the environmental literature, although only as a synonym for carrying capacity. 29.  For instance, see Khan and Riskin (2001) or Brandt et al. (2002). 30.  Interview, director of the local TAR office of the Swiss Red Cross, Shigatse, November 2004, and confirmed by a senior Tibetan official from the Tibetan Academy of Agriculture and Animal Husbandry Sciences (TAAAS), interviewed in Lhasa in November 2004. 31.  Typically, most farmers whom I interviewed in Qinghai or Sichuan estimated their yields at around 150 to 200 kg per mu. In one farming area of Qinghai where the soil quality was poor, yields were about 100 to 120 kg per mu. The Tibetan TAAAS official mentioned in the note above told me that farmland in the river valleys in Lhasa, Lhoka, and Shigatse Prefectures in the TAR can yield around 250 kg per mu, or even up to 400 kg with hybrid grains. Outside these river valleys, yields are more in the range of 100 to 150 kg per mu. 32.  Using 150 kg per mu as a measure, this leads to a per hectare yield of less than half of that cited in table 7.1, that is, 2250 kg. I am not aware of the official method of calculating the latter yields, although it possibly derives from a survey bias toward the higher-yielding pure farming river valleys in the TAR and against the mixed farming-pastoral regions where I was doing most of my fieldwork. Along with the often-noted official tendency to over-report yields, the bias is probably consistent across China. Alternatively, the per capita yields that I have estimated above, on the basis of my fieldwork, are more consistent with the official per capita yields reported in table 7.1. 33.  This yield estimate was reported to me by the country officials noted above. Most of the Han and Muslim areas in the northwest are not nearly as well endowed

332

Chapter 7

in terms of water supply, and land degradation is a serious and escalating problem. On the other hand, farm yields in Sichuan are probably similar to those in Xunhua. 34.  Also see Lewis (1978, 14-16) for an excellent synthesis of this earlier argument. 35.  My presentation at the 11th IATS conference in Bonn, August 2006, generated much discussion on these points, particularly from Melvyn Goldstein and Geoff Childs, both of whom had recently returned from rural farming areas in the TAR and remarked on the prevalence of job seeking among farmers. However, on further discussion, Geoff Childs agreed that such job seeking tended to be targeted toward middle-rung positions in the employment hierarchy, in work that would tend to enhance status in the rural community, such as carpentry or truck driving. 36.  I noted this on several occasions. For instance, in the Sogwo (Ch. Henan) Mongolian Autonomous County in Qinghai, in July 2003 I interviewed one Sichuanese woman who had migrated to the county town the year before with her husband and small child and opened up a small restaurant. They had decided to come because the woman’s sister, who was working in the local government, had told her that there were no Sichuanese restaurants in the town and thus business would be good. Both sisters had been born and raised in the county after their parents were transferred there in the 1950s. 37.  My observations in Qinghai were confirmed for Central Tibet in an interview with a senior Tibetan official from the Tibetan Academy of Agricultural and Animal Husbandry Sciences (TAAAS) in Lhasa in November 2004. The official told me how they had tried to offer construction work to farmers in Shigatse at 40 yuan a day—a very decent wage—but that many of the farmers were not interested and instead returned to their fields, being more concerned with digging for caterpillar fungus. However, Winkler (2008a; 2008b) notes that Shigatse is not a caterpillar fungus–producing area, so the official was probably constructing an illustrative example based on broad experience. Similarly, several NGO staff who I interviewed in Lhasa had experienced some difficulty with early attempts to interest locals from Nagchu, a pastoral region of the TAR, in vocational training. This was likely due to the stigmatized nature of the work that the vocational work was targeting. This being said, in personal communications I had with Emily Yeh in January 2013, she noted that she knows of a very successful vocational training project on motorcycle repair in Nagchu, so these cases vary (although the example of motorcycle repair again fits in with acceptable notions of employment among herders). 38.  See a similar discussion of Tibetan employers’ choosing to hire Han rather than Tibetans in Yeh (2013). Yeh stresses the issue of Tibetans’ being seen as more difficult to discipline than Chinese migrants because of social/cultural expectations that are absent when Tibetans hire Han employees. 39.  See my anonymous article on this issue in TIN/Fischer (2004). Given that TIN has been defunct since October 2005 and its website has not been functional I abrogate TIN’s policy of anonymity and assume my right as sole author of this article, along with other TIN articles that I wrote from 2003 to 2005. 40.  These observations are based on market research that I conducted for one textile factory in the TAR in 2003 and interviews in 2004 and 2005 with three Tibetan carpet traders and five INGO workers who had been involved in various attempts to set up carpet-making ventures.



Subsistence Capacity and the Material Foundations of Resistance 333

41. Interview with a Tibetan industrialist in Kathmandu, June 2003; and discussions with Malika Shakya, a Nepali PhD candidate at the London School of Economics who was in LSE researching textile industries in Kathmandu at the time. 42. Interview, Lhasa, November 2004. For background on Tashi Tsering, see Goldstein et al. (1997) and Siebenschuh and Tsering (2003). 43.  See, for instance, Ma and Lhundup (2008). 44.  At that time, the loans typically had interest rates of around 8 percent per annum. 45.  See Yi (2005a) for an analysis of these expectations among Muslims in Rebgong, Qinghai. 46.  A parallel might be made in this sense with the Kshatriya in the Indian caste system. 47.  Again, see Pirie (2005) and Yeh (2003b). 48.  From a different perspective but with a focus on similar issues, see the excellent discussion of the discourse of the “lazy Tibetan” in Yeh (2003a; 2007). 49.  Notably, as analyzed below, the need for Tibetans to undergo cultural change is often highlighted, along lines that accord with the Chinese view of higher “quality” (Ch. suzhi) human resources. Indeed, official insistence that Tibetans need “education” refers not only to conventional secular schooling, but also to what might be called “cultural education,” or education that changes these culturally bound forms of economic behavior. Again, see Yeh (2007). 50.  I use the word “anecdotal” because their fieldwork in Tibet appears to have been very short (perhaps a month), with no indication of ethnographic methodology and typical prejudices taken at face value. 51.  Presumably, they are referring to tsampa, or roasted barley flour, which is the basic (and nutritious) staple of Tibetans. It is typically mixed with butter tea, hand kneaded into a doughy paste, and then eaten directly without cooking. While the average Chinese person and most Westerners look on the hand-kneaded paste with much distaste, this is obviously not an issue for those raised on tsampa. 52.  See chapters 2 and 3. 53.  I am indebted to discussions with Ken Bauer and Melvyn Goldstein for these insights on the changing systems of rangeland management. 54. Conversation with Melvyn Goldstein, November 2005, Berlin, and subsequent communications. 55.  I use this term “peasantisation” with reference to Washbrook (1988). 56. Similarly, some authors in the literature on famine point out that vulnerability to famine or severe poverty emerges only after substantial asset erosion. In particular, see Swift (1977; 1989) and Rangasami (1985). 57.  In some cases these are of a different character, although during my fieldwork in 2004, one colleague researching in Machu County in Gansu reported to me that the county government was planning a wholesale resettlement of herders in the county. Later updates indicated that this goal had been substantially whittled down through the negotiations and contestations that inevitably arose through the course of planning and implementation. 58. These views were constantly repeatedly to me in interviews with a variety of Han (and some Tibetan) scholars and officials in Beijing, Lhasa, Xining, and Chengdu.

334

Chapter 7

59.  For an excellent study of these various strategies, particularly in interaction with the primary school consolidation policy in Qinghai in 2009, see Cencetti (2012). Also see Du (2012). 60.  I am grateful to Jampel Dell’Angelo for allowing me to use this example from his MSc dissertation research at the London School of Economics.

8 Boycotts and Religious Networks: Counter-Strategies of Integration into the Heart of the New China

This chapter, based largely on fieldwork in Qinghai and to a lesser extent in Sichuan, Gansu, and the TAR, from 2003 to 2005 Gansu and the TAR, offers some qualitative reflections on the conflictive or collaborative repercussions of exclusion and resistance as various aspects of double movement, that is, the impulse for social protection induced by the context of rapid, disembedding, and disempowered social and economic transition in the Tibetan areas. Whereas the previous chapter focused on rural subsistence as the material basis of double movement, the focus here is on Tibetan urban areas as the loci for both contestation and the formation of alternative indigenously rooted strategies of social protection. This is examined through the rise of anti-Muslim sentiment throughout the Tibetan areas during the reform period as well as the financial revival of Tibetan Buddhism through expanding lama-patron networks within China. The choice of these two patterns of inter-ethnic relations might appear surprising. However, these two cases are chosen because they capture the sense of double movement that appeared most poignantly to me through the course of my fieldwork. These involve the disruption or even reversal of traditional social and wealth hierarchies—in this case, ethnic hierarchies rooted in labor hierarchies—that the disembedding transitions have entailed, and some of the most potent indigenous Tibetan attempts to innovate new ways to reorder these changing hierarchies to their advantage, particularly through the leadership of traditional Tibetan elites. As discussed in the conclusion of chapter 2, tensions between Amdo Tibetans and local Muslims of northwest China have gradually re-emerged as one of the dominant foci of contestation within the transforming society and economy of 335

336

Chapter 8

the Tibetan areas during the reform period. These various local conflicts coalesced with sufficient momentum in the early 2000s to produce a regional boycott movement by Tibetans against Muslim businesses in Tibetan areas. Parallel to this, Tibetan relationships with the Han Chinese have been shaped by continued contestation of attempts by the Chinese state to forge hegemony out of supremacy, which has been the focus of many political studies on Tibet outside China. However, less analyzed has been the fact that contestation has been also complemented by increasingly close social collaboration between Tibetan and Chinese people through the vehicle of Tibetan Buddhism. The evolutions of these two aspects of contemporary ethnic co-existence, whether conflictive or collaborative, can both be understood as possible outcomes of the impulse for social protection, conditioned by the underlying forces of dislocation and relocation that derive from modern transformations, which in turn shape patterns of exclusion and hence the possible channels for new forms of inclusion within the intensified integration of the Tibetan areas into China. In addition, these two dimensions of contemporary inter-ethnic coexistence had rarely, if ever, been treated in the scholarship on Tibet at the time the material of this chapter was first analyzed and published in Fischer (2005b; 2007; 2008b), although some notable historical contributions had been made, such as Tuttle (2005). The lacuna in the literature around these issues has not been due to a lack of salience. Rather, it has been arguably due to the extreme political sensitivity of many of these issues, particularly with respect to Tibetan-Muslim conflict, from the side of both the Tibetan Exile Government as well as the PRC. For this reason, great care is taken in the reporting of my fieldwork, which is presented as a distillation of my field insights and experience gathered in both the exile community in India from 1995 to 2001 and then in Tibet from 2003 to 2007. I have not made an effort here to compare these insights with a few significant more recent contributions on the issue of Tibetan-Han religious interactions, most notably by Smyer Yu (2011), although I mention these where appropriate. The chapter starts with an examination of local economic restructuring and its asymmetrical impact across various ethnic groups in the Tibetan areas of Qinghai, in particular the de-industrialization of Tibetan towns outside the TAR in contrast to the success of certain neighboring Muslim counties in consolidating control over locally integrated processing industries. In combination with the employment reforms already discussed in chapter 6, the resulting confluence has arguably played a key role in fueling grievance, nationalism, and, in particular, prejudice against Chinese Muslims among Tibetans in this region, even despite the overall context of rapid economic growth and rising living standards. The second and third sections then examine Tibetan-Muslim ethnic conflict and Tibetan-Han



Boycotts and Religious Networks 337

religious cooperation as two different aspects of double movement among Tibetans within this context.

The Ethnic Political Economy of Economic Restructuring in the Tibetan areas of Qinghai During the period of opening and reform in China, from the early 1980s onward, there was an emerging, although historically rooted, perception among Tibetans that Muslims were increasingly dominating the economy of the Tibetan areas, particularly in the region of Amdo, but also as far south as cities like Lhasa. This perception requires a qualification: the Chinese state obviously remains by far the most dominant player in the economies of the Tibetan areas, as analyzed in previous chapters. However, parallel to this dominance, rising Tibetan-Muslim tensions relate to two economic dimensions that are central to local notions of ethnic hierarchy. One has been the marked economic advantage accrued during the reform period in the traditional niche specializations of Muslims versus those of Tibetans in the Tibetan areas, that is, trade, manufacturing, and services, versus farming, herding, and local public employment. The second has been the entry of Muslims into the new economic opportunities in the regional economy that have come about in the wake of state-induced growth, in contrast to the de-industrialization of Tibetan towns. These two dimensions compound the common Tibetan perception that Muslims are disproportionately advantaged and Tibetans are disadvantaged in the context of rapid growth and change, particularly within the context of intensified migration since the mid-1990s. The first dimension is poignantly summarized by the story of wool. As discussed in chapter 2, the dynamics in the terms of trade of wool (i.e., wool prices compared to other prices in the economy, and hence the purchasing power per unit of wool produced by a herder) capture the essence of how the main Tibetan niche sectors of the local economy were undermined as sources of monetary value by both national economic structural changes as well as international trade liberalization. The story of wool also helps to understand the subsequent repositioning of various groups and actors with regard to new economic opportunities emerging in the wake of the reform period and, in particular, the increasing presence of Muslims in Tibetan areas. Notably, nominal prices (i.e., the current yuan value) and relative prices (i.e., compared to other prices) of the main commodities produced by rural Tibetans had collapsed during the first two decades of the reform period. The collapse was rooted in the gradual economic reforms that began in the 1980s, in particular in the continued underpricing of raw materials (such as

338

Chapter 8

wool) as an implicit subsidy for coastal processing industries despite liberalization in other parts of the economy that also privileged the coastal areas, followed by liberalization of these raw material prices in the 1990s (with respect to both domestic and international trade). The regional discrimination that was perceived in these policies catalyzed the “commodity wars” of the middle to late 1980s across interior and western China or, more specifically, the “wool wars” in the major wool producing provinces of Qinghai, Gansu, Inner Mongolia, and Xinjiang (the TAR seems to have stayed out of the wool wars, probably due to the much tighter security environment, particularly following the protests from 1987 to 1989). The following account of the wool wars is taken mostly from the seminal work of Dali Yang (1997, chapter 4) supplemented with insights from my own fieldwork in Qinghai.1 National competition over the wool trade in the 1980s was fierce, involving producers, private traders, state trading companies, county governments, coastal and interior processing industries, as well as provincial and central governments. Local governments in the wool-producing regions were trying to stimulate and protect the development of local wool-processing industries as a means of taking advantage of their undervalued wool to promote local production, employment, and wealth, and they tried to block the export of wool from their jurisdictions for this purpose (an approach that became known as “local protectionism”). Meanwhile, coastal processing industries tried to maintain a secure supply of wool, with the effect that prices in private or illegal regional markets for wool were pushed far above the official state-controlled prices. The price differences undermined local government efforts to prevent the sale of local wool to private traders given that wool producers logically preferred higher prices and were not necessarily the beneficiaries of the local processing activities. Amid these contesting interests, the central government was trying to regulate the rapid proliferation of wool-processing activities, in part to guarantee a stable inter-regional supply of wool. It was precisely these conditions of intense competition, within a chaotic structure of fixed, negotiated, and market prices, that led the central government to designate wool as one of the first commodities to become fully liberalized with respect to national pricing and international trade. This was done to circumvent local protectionism and to guarantee supply to the coastal industries. The inflow of high-quality and relatively cheap wool in the late 1980s and 1990s, primarily from Australia and New Zealand, was therefore one of the main factors underlying the subsequent collapse of market prices for wool. As analyzed in chapter 2, the terms of trade for wool relative to rural consumer prices in Qinghai fell by around three-quarters between the late 1980s and 2004. Similar terms of trade reductions were observed in the main Tibetan farming commodities such as barley, wheat, and rapeseed. The resulting impact of these collapsing relative prices was enormous for



Boycotts and Religious Networks 339

rural Tibetans, particularly for those deriving a substantial portion of their livelihoods from herding or farming. It acted as a powerful downward pull on the real value of Tibetan rural incomes in the 1990s, as analyzed in the previous chapter, although since the late 1990s the collapse in the value of wool appears to have been compensated for by rising prices for meat and hides, as well as by the very sharp price increases for caterpillar fungus and the subsequent bonanza in many pastoral areas where caterpillar fungus is found.2 Indeed, in several of my interviews in Golok with pastoralists and one foreign scholar living there in May 2004, it was noted that sheep wool was seen as having almost no value once costs of production and transport were considered, and that pastoralists there were raising sheep mainly for meat rather than wool, especially since Goloks apparently prefer mutton to yak meat. While sheep wool was still sold as a “by-concern” of an otherwise planned trip to a town, possibly as a means to earn some pocket money, the main point is that wool output was not viewed as a major source of commoditized income generation given that low prices barely justified the costs of labor and transport.3 On the other hand, Chinese Muslims in Tibetan areas were not necessarily as adversely impacted by such dramatically falling terms of trade given that they were traditionally specialized in the trade of wool products and thus they needed only to maintain profitable trading margins. Obviously, Muslim farmers in locations such as Xunhua, traditionally a farming area, would have been similarly impacted by the falling terms of trade in farm goods. However, as discussed below, this reinforced the impulse within these communities to rely more on their other traditional niche specializations in trade and commerce, or else new activities such as wool processing, for which they would have also had more cultural affinity, in contrast to Tibetans, especially Tibetan pastoralists, who (with some exceptions) generally look down on trade and commerce. Although wool had been undervalued throughout the Maoist period, strict control over the economy via collectivization prevented such pricing issues from degenerating into local distributional conflicts. In particular, total state monopoly over the distribution of commodities removed Muslims from their traditional niche role as regional traders across the Tibetan areas, thus removing them from being the focal point of local Tibetan resentment. However, Muslims reverted to their traditional intermediary role with the lifting of restrictions on migration, markets, and individual entrepreneurship from the beginning of the reform period. In some cases, they almost completely took over the role of the state by buying up county-owned trading companies or abattoirs during waves of privatization in the 1990s and early 2000s.4 The liberalization of the wool trade therefore gradually ushered in an increasing presence of Muslims in the Tibetan areas, some of whom

340

Chapter 8

were regular residents in these areas, but many of whom were not settled within any particular Tibetan community, although they were generally native to the larger Qinghai-Gansu region.5 Therefore, by the time wool prices started to collapse in the 1990s in the midst of a range of other liberalization measures, the face mediating the collapse was Muslim, not the local Chinese or Tibetan officials, whose wool trading and processing operations were unraveling and eventually became defunct. Whether or not Muslims engaged in unfair trading practices, as is usually claimed by Tibetans, Tibetan discontent with the loss of purchasing power from their main tradable commodity was undoubtedly felt and expressed through their trading interactions with Muslims. For this reason, many of the accounts of fighting between Tibetans and Muslims that I collected during fieldwork in Qinghai, Gansu, and the TAR, particularly in pastoral areas, were in relation to the wool trade and, in particular, wool prices. For instance, on several occasions, Tibetans told me how Muslim traders would charge a price that then turned out to be less than the fair market price. As the story is told, the Tibetans involved realized this only once they traveled outside the Tibetan areas—typically to a city such as Xining, Lanzhou, or Chengdu—and realized that wool prices were in fact much higher. Of course, such price differences might simply represent the markups that are normal between farm gate, wholesale, and retail prices, and might not imply cheating, although this point was never taken into consideration by my Tibetan interlocutors. Furthermore, Tibetan herders are not necessarily an innocent lot either. On several occasions I heard—from both Tibetans and Muslims—of a fairly common practice of dirtying sheep wool in order to make it heavier before a sale, a practice that no doubt causes Muslim traders to underprice dirty wool. Similarly, many of the conflicts between Tibetans and Muslims over abattoirs in this region break out over questions of prices and procurement, even though Tibetans often framed these conflicts within a moral Buddhist discourse.6 The second economic dimension of Tibetan discontent with Muslims involves the entry of Muslims into new areas of regional economic activity. Again, this partly relates to the story of wool. Within the context of the wool wars of the 1980s, wool processing industries, such as textile and carpet factories, proliferated in many Tibetan counties of Qinghai and in other wool producing provinces. Tibetans were involved in these efforts, either at the level of workers or management. However, together with the collapse of wool prices in the 1990s, the national government implemented carrotand-stick strategies aimed at controlling the proliferation of wool processing, which led to a decisive regional consolidation of the industry. One of the disciplinary strategies involved the removal of subsidies for county-owned factories, in line with the national policy to streamline stateowned enterprises by shedding less profitable enterprises, particularly those



Boycotts and Religious Networks 341

in non-strategic sectors such as textiles. Unfortunately, most of the Tibetan efforts to enter wool processing had been led by county-run ventures, typically (and often poorly) run by local Tibetan (and Chinese and Muslim) managers selected by (and often related to) county officials. This insight was recounted to me several times in meetings with groups of Tibetan scholars from across Qinghai in 2004. On two occasions several of them described in detail how the management of these processing ventures was often characterized by high degrees of nepotism and diversion of funds toward other county government priorities, including kickbacks to officials and other forms of guanxi. These issues were accentuated by the fact that officials and managers were often related (and, in most of the cases cited, Tibetan). A further factor that was consistently noted was the lack of discipline among Tibetan workers in these ventures, particularly at wages that would make them competitive with Han or Muslim factories in northwest China producing similar types of products and with similar types of production techniques (i.e., handloom or machine produced), as discussed in the previous chapter. The removal of subsidies and related policies therefore forced the closure or privatization of most of these local government-run operations by the early 2000s. As a result, during my fieldwork in 2004, I was not able to identify a single Tibetan-run wool-processing factory in Qinghai that was still viable or running at full capacity. My search was aided by the insights of numerous Tibetans, in particular the Tibetan scholars mentioned above, during my fieldwork in several of the key Tibetan towns of Qinghai, such as Markuthang (Chentsa County), Rongwo, Chabcha, Jyekundo, and Dawu. In every case, factories that had been operational up until about 2000 or 2001, including non-wool-based factories such as leather processing, had since closed down or were operating at negligible capacity. One of these cases from Chabcha is described below. Some were still functional in Lhasa, such as carpet making, although these were largely dependent on supplying up-market tourist demand or else niche luxury markets in places such as New York, through networks supported by international NGOs and wellmeaning foreigners.7 This does not imply, however, that the wool-processing industry in Qinghai was underdeveloped. On the contrary, by the late 1990s it had consolidated into a small number of large private players. As it happens, these players were mostly Muslim, and even among these Muslims there had been significant consolidation. For instance, a Salar Muslim manager of a wool-processing factory in Xunhua who I interviewed in June 2004 informed me that “a few years ago” there were sixty-eight companies processing Tibetan yak wool in Xunhua County alone, whereas in 2004 there were only a few left. The company he worked for was the largest and most sophisticated, and had come to dominate the market for higher-

342

Chapter 8

quality wool in Qinghai. This private group was formed in the 1980s on the purchase of privatized government factories in Xunhua by four Salar Muslims, one of whom, although illiterate, was apparently the richest man in Qinghai in 2004. The company had been tightly managed by these four owners, together with considerable logistical support from the local Xunhua County government.8 The operations of the main factory in Xunhua, which employed 1,100 people in 2004 and generated a revenue of 120 million yuan in 2003 (two-thirds from exports), were based entirely on the processing of Tibetan (and Qinghai Mongolian) yak, sheep, and camel wool, although the group had apparently also expanded into real estate investments in Xining, among other activities. Notably, this industrial group came to dominate the wool-processing industry in Qinghai after they scaled up their operations in the mid-1990s by importing automated processing machinery from Japan and Germany.9 Smaller, less sophisticated private Muslim processing businesses also remained active in the region, capturing the supply of lower-quality wool sufficient for products such as carpets. As a result, the only processing operations that seemed to have survived this period of industrial restructuring were the private Muslim ones. Most of the trade in Tibetan wool in Qinghai was dominated by these players at the time of my fieldwork. It is interesting that industrial wool processing had not been a traditional niche of the Salar in Xunhua, who, like the Tibetans, had been mostly concentrated in agriculture up to the 1980s, and to a lesser extent in trade.10 The shift had been supported by activist industrial support by the county government, which had a strong incentive to promote Salar self-reliance given that Xunhua is the only Salar autonomous jurisdiction in China, amid a dominant society (both Chinese and Tibetan) that is perceived to discriminate against Muslims. This spirit of self-reliance suited them well for the privatizing conditions. Also, the labor force of these factories, mostly local Salars and to a lesser extent Hui Muslims from the region, many of whom were young women, was considerably more disciplined and cheaper than comparable Tibetan labor.11 Thus, the ladder for Qinghai Tibetans to progress into higher valueadded processing and marketing of their main commodity had been effectively kicked away by the late 1990s. Their position in the value chain as suppliers of raw materials thereby became increasingly difficult to move away from given the already high level of maturity in the provincial woolprocessing industry, and thus the high barriers to entry. The higher wage and related work expectations of Tibetan workers—reinforced by the escalating prices and wealth accrued from the caterpillar fungus trade from late 1990s onward—further undermined the competitiveness of Tibetan processing ventures. This ethnic segregation of the local value chain in wool production and processing thereby exacerbated various underlying feelings



Boycotts and Religious Networks 343

of grievance toward Muslims held by many Tibetans in Amdo (and other Tibetan regions). Industrial Restructuring in Malho and Tsolho Prefectures The resultant de-industrialization of the main Tibetan towns in Qinghai left Tibetans with fewer options for non-farm employment, at least for non-farm employment that was appropriate for their low average levels of schooling, and yet also in accordance with their wage, non-wage, and status expectations. Such employment was nonetheless sorely needed precisely because of the sharp decline in their commodity prices, compounded by relatively rapid population growth and rising basic schooling levels. More recently, subsequent to my fieldwork, a variety of factors that have accentuated urbanization have made this need even more pressing, such as various resettlement programs or else the consolidation of rural primary schools in county towns that took place in parts of Qinghai in 2009.12 Such de-industrialization was most evident to me during my fieldwork in Rebgong and Chabcha Tibetan Autonomous Counties. The main towns of these two counties (Rongwo and Chabcha) both serve as the administrative seats of their respective prefectures (Malho and Tsolho), which are among the most densely populated Tibetan prefectures in Qinghai, located relatively close to the core region of Xining. Both towns constitute the most important schooling and economic centers of all the purely Tibetan towns in Qinghai, and thus they also had the strongest base of factory development. Based on a series of extensive interviews with Tibetan county officials as well as one senior prefecture-level Han official in Rongwo over the course of several visits in 2003 and 2004, the extent of polarization in the local economy was very evident. Malho Prefecture in fact had a very strong secondary economic sector,13 although the resurgence of this sector in the middle to late 1990s was due to the construction and subsequent high value-added operation of one large-scale hydroelectric project along the Yellow River in Chentsa County that came into production in the late 1990s (the Lijiaxia dam, one of the largest in China with a 2000 megawatt capacity). A high (albeit declining) secondary share was then sustained in the early 2000s, up to the time of my fieldwork, through a variety of infrastructure projects in the prefecture, including the construction of several smaller-scale hydro-electric projects. However, these activities were concentrated outside the Tibetan towns and had few, if any, linkages to the local economy, particularly once construction phases ended. For instance, the senior Han official explained to me that very few Tibetans were employed at the Lijiaxia dam because they did not have the skills, besides basic construction work. Even in construction work, firms from outside the province hired mostly their own labor

344

Chapter 8

and the prefecture did not have any policy to promote local employment with outside investors. Furthermore, although located in Malho, the dam and its electricity belong to the Qinghai Electricity Bureau, which in turn coordinates the generation of electricity according to downstream water supply needs and the distribution of electricity according to needs within the national grid. The electricity generated in smaller private dams is also managed within this system of national coordination. As a result, similar to the Longyangxia dam in Tsolho Prefecture, also one of the largest in China, the electricity generated by the Lijiaxia dam is prioritized for several key energy-intensive industries, such as a large steel mill in Xining and an aluminum mill near Xining. Ironically, despite the enormous generation of electricity, the prefecture was required to purchase electricity from the provincial electricity bureau and often experienced shortages in the winter months. This in itself became a means of enforcing industrial rationalization. For instance, a small aluminum factory in Chentsa had closed down temporarily in spring 2004 because the government raised the price of electricity and the factory could no longer make a profit at the higher costs. The workers also asked for a wage increase, which might have precipitated the temporary closure.14 In contrast, the senior Han official acknowledged that there were essentially no manufacturing industries still functioning in Malho Prefecture, such as those in neighboring Xunhua County. There had been a private wool-processing factory, also run by a person from Xunhua, but it went bankrupt in 2002. Through interviews with other locals in 2004, I also identified one leather shoe factory that had been first privatized and then closed down the year before, as well as one unsuccessful cheese factory that had closed. The senior official suggested that the reasons for the failures of industrial restructuring in the Tibetan towns had been a lack of good production and management systems; lack of scale; and products of insufficient quality or reputation. As a result, most factories went bankrupt in the face of reduced subsidies and intensified competition. Similar patterns were also evident in Chabcha Town (Tsolho Prefecture). One Tibetan high school teacher and two Tibetan scholars, all from Chabcha and interviewed at different times from June to August 2004 in Chabcha or Xining, concurred that there had been one fairly successful Tibetan wool factory that was privatized sometime around 2001. The factory was sold to a Han man from Guangdong who apparently made a deal to buy first and pay later, but then absconded, leaving behind losses of about four million yuan. The factory closed down for a while. It was operating in 2004, albeit at a much smaller and intermittent scale than before, filling small orders as they came in. One successful small carpet factory was identified, but beyond this, most of the other factories that had once existed in the town had since closed down.



Boycotts and Religious Networks 345

Industrial restructuring in these Tibetan areas over the 1990s and early 2000s therefore essentially involved a reorientation away from townbased manufacturing and toward large capital-intensive energy and infrastructure projects. Although very productive, the latter were utterly de-linked from the local economy, thereby creating a sharp polarization in the local economic structure and a growing dis-synchronization between employment needs in the rapidly growing towns on one hand, and the supply of employment generated by rapid GDP growth on the other. The outcome was similar to that analyzed in chapters 5 and 6 regarding the TAR, even though through somewhat different processes (e.g., a much larger weight of secondary activities than is observed in the TAR economy). Insecurity and the Dislocation of Hierarchy Much like in the TAR, economic restructuring in the Tibetan areas of Qinghai intensified competition in economic activities outside farming and herding, which, old or new, came to be increasingly concentrated in towns (with the exception of caterpillar fungus harvesting), even despite deindustrialization of Tibetan towns. As described in the previous chapters, public employment is a coveted option among Tibetans, such as work in the government or as teachers, although this tends to be restricted to the educated few. Moreover, labor market reforms compounded by increasing numbers of Tibetan, Muslim, and Han graduates intensified competition for the limited opportunities that exist in Qinghai. As a result, Tibetans aspiring to upward mobility in the local labor hierarchy have been increasingly restricted on this path and increasingly pressured to rely on lower middle-rung positions. Hence, strong demands for middle-rung employment in Tibetan towns in the 2000s came both from below (rural-urban migration) and from above (Tibetan graduates excluded from public employment), precisely at a time when the industrial base of these Tibetan towns was disintegrating. In other words, those unsuccessful in obtaining public employment, along with laid-off staff and workers, and the bulk of Tibetans with low levels of schooling, were all pushed toward the lower skill and wage opportunities that came in the wake of the subsidized expansion of Tibetan towns, such as construction, petty trade and commerce, and a variety of services, in particular catering. In certain locations, tourism can be added to this list. Yet these are precisely the activities where a variety of Muslim and Han migrants also entered into the employment fray in Tibetan towns, taking advantage of their larger regional trading networks, their more astute business sense, and their lower wage expectations (particularly in Qinghai and other provinces outside the TAR). The culmination

346

Chapter 8

of all these factors resulted in the sidelining of Tibetans from even those residual lower skill opportunities. These processes of economic restructuring have been especially contentious for Tibetans because they have resulted in a profound displacement of ethnic hierarchies in Tibet, challenging deeply set notions of dignity within these hierarchies, which had ironically been preserved throughout the collective period and up until the 1980s. Moreover, the challenge works across social class, from the level of farmers and herders up the level of elites, as argued for the TAR in chapter 6. The new Tibetan elites of local government cadres and others who have positioned themselves as junior partners within Chinese rule have undoubtedly been better placed to weather these displacements within the local economy. Nonetheless, even their relative employment security up to the 1990s has since been considerably eroded, with no guarantees of similar security for their progeny, at least not without substantial patronage and guanxi. Instead, for many the future presents a likelihood of downward mobility in the local social hierarchy. Government strategies of self-employment and smalls business start-ups can be seen to potentially irk the dignity of these once-privileged Tibetan employees as well as the graduates aspiring to upward mobility, if not openly insulting them by suggesting that, after all their privilege and/or education, they should compete shoulder to shoulder on an equal level with Muslims and low-cultured Han migrants in a variety of stigmatized activities. Given that these elites are also keenly aware of the legitimate need for their non-elite rural compatriots to move up the economic hierarchy in the towns, moving into the economic spaces that have come to be dominated by these same Muslims, a powerful common purpose across class arguably coalesced within Amdo Tibetan communities in the 2000s, targeting the most vulnerable scapegoat: Muslim businesses.

Boycotts and Modern Myth Creation in Amdo15 The idea of boycotting Muslim businesses appears to have developed in different pockets of Amdo in the 1990s, if not before. Perhaps coincidentall similar ideas were being promote in the exile community in South Asia by the active promotion of the Gandhian concepts of Swadeshi (selfsufficiency) and Swaraj (self-rule) by Samdhong Rimpoche, the prime minister of the Tibetan Government in Exile, as possible strategies to take against the increasingly marketized Chinese rule during the reform period. Given the regular travel and exchange of Tibetans back and forth between Tibet and the exile community in India since the 1980s, I suspect that the resemblance of the Muslim boycott movement in Amdo to Gandhian social activism is not a coincidence, albeit not in the manner suggested by Beijing



Boycotts and Religious Networks 347

as if Tibetans are receiving orders from the exile government, but more as a matter of uncoordinated or “vernacular” ideational dissemination.16 Other influences probably include similar activism either in history or else in other locations across the Tibeto-Buddhist world, such as the Ladakhi boycotts against Muslims in Ladakh studied by van Beek (2000). However, the boycott only escalated into a regional movement following a clash between Tibetans and Muslims in Chentsa County in January 2003. One account of this clash was reported in the Western media (CNN 2003). In the CNN account, the clash began when a group of Tibetan and Muslim youths bickered over a game of billiards. However, several Chentsa Tibetans claimed to me that it had started when some Hui youths stole a motorcycle from some Tibetans. Fighting ensued until one fifty-year-old Tibetan was stabbed to death with a skewer by two Hui youths, and then hundreds of Tibetans went rioting in the town, destroying Muslim-owned property. The military was quickly called in. From my fieldwork in 2003 and 2004, I was able to ascertain that the older man who died was a village leader who had come to mediate the clash. All the Muslims who were fighting were from a village across the Yellow River in Hualong County, called Bagyathang, which is known as a contraband center for drug and arms smuggling. None of the Muslim residents of Chentsa County were involved. The security presence remained for several months. This event, which occurred around the same time as several other Tibetan-Muslim clashes in other parts of Amdo, appears to have acted as a catalyst and rallying cry for the boycott to be taken up as a regionally coordinated political movement, spreading as far south as Lhasa and Kham. As a result, Amdo came to function as an epicenter for Tibetan-Muslim confrontation. The standard rationale for the boycotts, which was told to me by a variety of students, scholars, and officials, and which was echoed in many conversations with Tibetan farmers and nomads, was that if Tibetans did not spend their money in Muslim businesses in Tibetan areas (which cater largely to Tibetans), then Muslim businesses would not survive and the Muslims would leave. Moral legitimacy was given to this argument by noting that the departure of Muslims would also reduce the amount of animal slaughter in the Tibetan areas, given that the abattoirs are run largely by Muslims. This latter argument is, apparently, often made by lamas. For instance, Gaerrang (2011; 2012) discusses the anti-slaughter movement led by the late Khenpo Jigme Phuntsog, a highly renowned lama throughout all of Tibet who was based in Serthar County in Sichuan (technically part of Amdo), as well as by Khenpo Tsultrim Lodroe and others at Serthar. As part of this campaign, a video had been produced with footage of Muslim abattoirs. The video decried the Muslim treatment of animals and advocated vegetarianism.17 Conversely, if Tibetans were to set up their own businesses and spend their money in them, they could capture the Muslim economic success.

348

Chapter 8

Curiously, while similar arguments have been made with respect to Chinese migrants, the boycott idea seems to have only taken hold with reference to Muslims. This could be a sign of a deeper cultural animosity toward Muslims or, more likely, it was an expression of opportunism, given that any overt animosity toward Han Chinese would be treated harshly by the authorities, while Muslims are generally discriminated against by both Tibetans and Han alike, and are scapegoated by both for a variety of social problems, such as increasing crime, or drug or human trafficking.18 Even a successful Muslim businessperson may be typically characterized as tricky or as a cheater.19 Moreover, there is much more Tibetan-Muslim competition across a variety of old and new economic activities. Muslims are closely integrated into the Tibetan economy, largely catering to Tibetan demand and plying their trades within Tibetan neighborhoods. They also follow lifestyles that are much closer to Amdo Tibetans than to those of the Chinese migrants (and arguably and ironically have made much greater efforts to integrate into Tibetan communities throughout history, such as by learning Tibetan language). As a result, Muslims are much more visible and proximate than Chinese migrants, who tend to cluster around the more modern sinicized sectors of Tibetan towns. From their own side, many Chinese migrants are directly or indirectly sustained by the local government, often attracted in the first place by relatives in the government, and any activism directed against them would be treated severely by the local authorities. In any case, whether due to some or all of these considerations, Muslims bore the brunt of economic activism in the early 2000s, whereas Chinese migrants escaped largely unscathed, or even benefitted, given that they attracted many of the clientele lost by the Muslim businesses. Prior to 2003, boycotts appear to have been localized events, concentrated in a few townships within certain counties.20 “Ngawa” was consistently identified during my fieldwork throughout Qinghai as especially important given the leadership of one lama who had been particularly vocal in advocating the boycott, although it was never clear whether this referred to the county or the prefecture of Ngawa.21 Interestingly, Ngawa County is known for its tradition of entrepreneurship in trade, and Golok Tibetans often complained to me, in an argument that resembled their resentment toward Muslims, that Ngawa Tibetans dominate most of the Tibetan-run commerce in Golok, taking most of their profits with them back to neighboring Ngawa.22 Given that the Ngawa business community also extends strong financial support to the local Ngawa monasteries, they might have been indirectly involved in sponsoring the boycott movement through these monasteries, if any had been involved in the boycott. Despite the local nature of these movements, they were well known throughout Amdo. Even in Central Tibet, far removed from Amdo, antiMuslim sentiment was reportedly strong as early as the mid-1990s, if not



Boycotts and Religious Networks 349

before, resulting in several clashes, typically involving Muslim restaurants or other businesses.23 One Western INGO worker who had been living there in the 1990s told me that in 1994 there was a rumor that “the Muslims are coming” to Lhasa, possibly referring to the liberalization of migration to the TAR around that time. In early 1995 some Tibetans claimed to have found a human finger in a soup served in a Muslim restaurant near the Jokhang Temple. This led Tibetans to claim that the Muslims were practicing cannibalism, a theme that was to return repeatedly during the later boycotts. The boycott movement gathered regional momentum in January 2003 following the clash between Tibetans and Muslims in Chentsa. According to my informants from Chentsa and Rebgong, along with several Tibetan scholars in Xining,24 Tibetans widely believed that the resolution of this episode was biased in favor of the Muslims, which in turn stimulated a call to action, an incitement to take matters into their own hands. There was a broad consensus on the events: the Muslim man who was responsible for the killing of the Tibetan village leader was let off after his village made a compensation payment of 10,000 yuan to Chentsa. They compared the treatment in this case to another event, which apparently occurred in early 2004, when a famous Muslim motorcycle thief from a nearby village in Hualong attempted a nighttime raid on a Tibetan village in Chentsa, during which he was ambushed and beaten to death by the villagers. Following his death, the Tibetan village had to make a much larger compensation payment to the thief’s Muslim village in Hualong, in addition to which twenty men from the village were arrested and detained.25 In any case, the January 2003 clash in Chentsa served as a call to action for Tibetans. It seems that the bravado of Tibetans had also been stoked up by the events of 9/11, the Afghan war, and the lead-up to the Iraq war.26 Soon afterward a regional boycott of Muslim businesses gained momentum and even extended into areas of Kham that had little Muslim presence. Although Tibetans in Lhasa were hesitant to talk about it, the boycott evidently reached Central Tibet as well, finding an accord with the anti-Muslim sentiment already built up over the previous decade. While the boycott was meant to target all types of Muslim businesses and trade interactions, including Muslim-owned buses, the symbolic focus became the Muslim restaurants that dominate catering throughout Amdo. The boycott appears to have been strongest, most cohesive, and most sustained in the more remote pastoral areas such as Golok, or in parts of Yushu and Ngawa Prefectures, in many cases continuing with force up until my field visits in late 2004. In Golok, the boycott took off with an event that occurred soon after the Chentsa clash, involving the poisoning of three Tibetans in a Muslim restaurant, one of whom died. The poisoning, which was recorded in the local newspaper, acted as a trigger for the boycott, although the question remains whether it took place after heavy-handed harassment

350

Chapter 8

of the restaurant by Tibetan youths.27 For instance, a Western scholar who was living in Golok at the time told me that some young Tibetan men, students from the secondary normal (i.e., teacher-training) school in Dawu, regularly harassed Muslim restaurant owners and other Muslim businesspeople by, for instance, eating and walking out without paying, telling the owners to follow them into the street. This was essentially an invitation to a street fight in which the Muslims would be inevitably outnumbered. Such Tibetan youth bravado apparently often involved the sons of county or prefecture officials, who acted with relative impunity. While I was in Amdo in 2004, several similar incidents in Golok were reported to me, such as one about a gang of Tibetans who destroyed the sign of a Muslim shop. The isolated shop owner simply had to watch, with little hope of police protection, given that most of the local police are also Tibetan and also quite biased against the Muslims. In any event, within a year of the poisoning, the local newspaper reported that the number of Muslim restaurants in the Prefecture of Golok had fallen from over seventy to just over twenty. Jigdrel Town, which was sacked by Ma Bufang in the 1930s, was apparently the first town of Golok to entirely rid itself of Muslim restaurants at that time. In the farming regions of Amdo that are closer to Muslim areas, such as Rebgong, Chentsa, Labrang, and Chabcha, the boycott started with a bang but petered out within less than a year. In many cases the boycott was enforced with some degree of moralistic coercion by Tibetan activists, many of them monks. For instance, Tibetans found eating in Muslim restaurants were shamed or even beaten up afterward on the street. In Rongwo Town, the seat of Rebgong County, it was reported that monks from the monastery would impose fines on people found eating in Muslim restaurants. I was even told that one Tibetan businessman from Kham who was visiting Rongwo was so enthused by the spirit of the boycott that he happily agreed to pay a fine of 1,000 yuan after being caught eating in a Muslim restaurant. In this way, Tibetans who were opposed to the boycott were silenced, thereby generating an appearance of consensus. Nonetheless, the boycott gradually ran out of steam in these more central areas due to the difficulties of boycotting neighbors in towns that have had a long history of ethnic coexistence and where Muslims play an important symbiotic role in the local economy. Even in Golok, many Muslim restaurants were simply taken over by Chinese migrants from Sichuan or by Ngawa Tibetan businesspeople. Chinese restaurateurs often showed their colors by advertising their new establishments as “pork restaurants.”28 While such restaurants are very popular as tasty alternatives to Muslim food, the displacement of Muslim restaurateurs by Chinese ones generally defeated the more instrumental purpose of the boycott movement, which was to allow Tibetans the chance to enter into these areas of the urban service economy that had been hitherto



Boycotts and Religious Networks 351

dominated by Muslims. Nonetheless, there appears to have been some expansion of Tibetan businesses in this regard.29 The Tale of the Muslim Cook and Other Modern Mythologies of Prejudice Parallel to the rapid spread of the boycott, a highly prejudiced mythology became widespread throughout Amdo. Fiction is mixed in with actual events to an extent that one is never certain whether a depicted event actually took place. The most typical is what I have come to call “The Tale of the Muslim Cook.” Its appearance, along with non-culinary spin-offs, seems to have taken place more or less around the time of the clash at Chentsa. In any case, whenever it was told to me, the time reference was always recent, that is, within the last year or two. A typical version, told in all seriousness by most of the common Tibetans whom I interviewed, including relatively educated Tibetans such as high school students,30 follows a standard formula, although elements varied from locality to locality. Here or there, a Muslim cook prepared the food he served to Tibetans using their leftover dirty bathwater, or used the bathwater of an imam. In the most extreme accounts, he put the ashes of the cremated bones of an imam, or even his parents, in the cooking water. It is said that according to Muslim belief, the use of ashes will cause the eater to convert to Islam or, at the very least, make him or her sympathize with Islam. However, all the Muslims whom I interviewed, including several scholars in Xining and one Hui Muslim secondary school teacher in Rebgong, insisted that there was no such belief among local Muslims. An almost identical story was prevalent in many counties. Some tellers sourced the exact event to “Ngawa,” others to the neighboring county, and yet others to their own county. In one rendition, a Muslim cook was working for a Tibetan boss and business was very good. One day the Muslim decided to leave, after which business worsened. The boss phoned the cook and requested him to come back, at which point the cook told the boss that he would never come back and that all the while he had been cooking with his used bathwater. It is said that this is how the Muslim practice of disrespect was uncovered. I was even told by two Tibetan high school students in Rongwo Town that this event actually took place within the main monastery of Rongwo itself, which apparently employed a Muslim cook and a Tibetan woman helper. The cook threatened the woman that he would kill her if she snitched on him, until finally she gathered enough courage and told the monastery. I was suspicious of this story given that monastery cooking is usually performed by the monks themselves. I asked a lama from the monastery about this and he denied that the monastery had employed a Muslim cook or that the event had ever taken place.

352

Chapter 8

There were several non-culinary spin-offs. For instance, in Yushu it was said that passing Muslim motorists throw the ashes of cremated imams in the air, which then lands on, or is inhaled by, hapless Tibetan pedestrians and has the same effect as eating a tasty noodle soup. It does not seem to perturb these storytellers that Muslims do not, in fact, cremate the bodies of their dead, but bury them according to prescriptions given by the Koran. Indeed, Rebgong County itself has been the scene of several well-known conflicts over Muslim attempts to rent or buy land for cemeteries, in which case it should have been obvious that even local Muslims bury their dead. However, none of the Tibetans whom I interviewed seemed to have realized this logical fallacy. On one occasion, a Tibetan high school student from Chentsa even started to argue that Chentsa Muslims do, in fact, cremate their dead as a local custom. All the evidence to the contrary did not seem to prevent ongoing embellishments of the ashy versions of the myth. These and similar conceptions of Muslims are not new among Tibetans. From discussions with other Tibet scholars, it seems that some of these prejudices have a long genealogy throughout Tibet. For instance, I was told by one researcher that rumors abounded in Lhasa in the mid-1990s. Tibetans were saying that Muslim bread was yellow because they urinated in it, that they put their scabs in their food, or else that they practiced cannibalism and ate children. Lipman (1997) discusses the evolution of prejudice against Muslims in China, and Yi (2005a) also offers a detailed account of prejudices against Muslims in Rebgong County. What is nonetheless striking in the Amdo events is the synchronicity between the boycott movement and the culinary myths, along with the rapid dissemination of both throughout most of Amdo within a matter of months, which was reported to me by numerous Tibetans and Western scholars. The most coherent explanation that I managed to assemble is that the Muslim Cook myth was initiated, or at least encouraged, by certain Tibetan elites. 31 These included secular intellectuals from some of the main urban centers in and around Amdo, such as Chabcha, Rebgong, Xining, and Lanzhou, along with religious leaders, such as the “Ngawa” lama mentioned above. The dissemination of the myth subsequently took place, knowingly or unknowingly, through students, businesspeople, or others who travel regularly between these key urban centers and the dispersed villages and towns of Amdo and beyond. Notably, the clash in Chentsa occurred just before both the Tibetan and Chinese New Year, when many Tibetans return to their home villages. These elite actors can be broadly considered as part of a Tibetan intellectual community that circulates within both secular and religious Tibetan institutions in Amdo. The secular include the minority universi-



Boycotts and Religious Networks 353

ties, research or cultural institutes, Tibetan newspapers, and similar organizations, all of which have functioned under the auspices of the PRC since the 1980s as part of the policy of promoting nominal “minority autonomy.”32 Many of these secular intellectuals maintain close formal or informal collaboration with monastic scholars, with whom they often also have family or other social ties.33 In turn, the same ties also bind both communities into broader elite networks, principally those of Tibetan cadres and businesspeople, who in turn act to various degrees within the limited space sanctioned by the state to support the projects of the secular and/or the religious intellectuals.34 On the religious side, many people have been actively involved in boycott activism, from lamas down to heavy-handed monks. Even state authorities tacitly acknowledged the moral authority of lamas, highlighted by the fact that several key meetings with religious leaders were reportedly convened by the Qinghai government in efforts to defuse the boycott.35 In one meeting in Golok, most of the leading lamas from the prefecture were assembled in Dawu, the prefecture capital, and officials requested them to tell the Golok Tibetans to eat in Muslim restaurants for the sake of preserving inter-ethnic harmony. It was said that one of the most senior lamas present at the meeting replied to the officials with the observation that Chinese and Muslim people never eat in Tibetan restaurants. He therefore concluded that they would tell their people to eat in Muslim restaurants only if the Chinese authorities told Chinese and Muslim people to eat in Tibetan restaurants. Whether or not this was an accurate rendition of the event, its popular repetition reflects the moral authority attributed to the lamas to dictate the terms and progression of the boycott. According to the explanations that I assembled, many of these elites had been brewing the boycott idea for a while. In the midst of the frustration and agitation that followed the Chentsa clash, the myths were disseminated in various versions, playing on local events as a means to incite Tibetans to embark on a sustained regional movement. Previously, events such as Chentsa were capable of mobilizing rural Tibetans into short-term bouts of localized retribution. However, these mobilizations had not catalyzed more far-reaching strategies. This required that Tibetans become motivated to strategically attack and take over the economic foundations of the Muslim advance into their local economies. However, in the view of various elite Tibetans, this logic had so far failed to motivate a largely under-educated Tibetan population. The myths were a means to this end: by instigating racist mania among the general population, which was possible due to pre-existing strong historical prejudices that Muslims were cheating, dirty, and dangerous, elites were able to mobilize collective action and consumer power to address issues of economic marginalization,

354

Chapter 8

and they were able to do this so long as the symbols of state power were not directly attacked. Instrumental Primordialism? The myth concoction therefore appears as a classic case of instrumentalism, that is, elites fomenting ethnic conflict in order to advance their own strategic agendas. Or, expressed in the theorized terms of the political economy literature on ethnic conflict, instrumentalism can be understood as grievances (such as economic marginalization) operating through structures of entitlements (which shape economic opportunities and the avenues available for political activism), such that the conditioned incentives facing elites lead them to foment open conflict.36 However, on closer examination, the direct benefits of the boycott to the Tibetan elite are not clear. Some Tibetan businesspeople probably benefited and it is interesting that one of the attributed sources of the boycott movement, “Ngawa” in Sichuan, is also known for its very successful business community (if this refers to Ngawa County). Nonetheless, the marginal benefits accrued from the boycott would have been counter-balanced by the damage done to Tibetan-Muslim business networks. For instance, one Muslim schoolteacher in Rongwo Town told me about the regular, although rarely mentioned, financial relations between wealthy Tibetans and local Muslim business people. Even wealthy monks and lamas from Rongwo Monastery are involved, lending money to local Muslims at interest rates of about 8 percent per annum. Furthermore, Tibetan businesspeople themselves are often, stigmatized and thereby could not—and did not—serve as popular rallying figures. In any case, businesspeople do not appear to have been the primary instigators of the movement, although they probably helped in its financing or circulation, as they would have done for any other nationalist or religious movement. Some members of the secular and religious educated elite, on the other hand, seem to have played primary leadership roles, yet their personal benefits are not clear. Most secular intellectuals are rooted in state- or para-state-sector employment that is generally not threatened by Muslims, yet which could be threatened if their activities were deemed contrary to state interests, as in the case of fomenting nationalism or ethnic conflict. Moreover, intensified competition over public employment in the Tibetan areas takes place largely with Chinese, not Muslims. Religious elites are also more or less immune to economic competition from Muslims, and their already tenuous relations with the state would only become further strained if they were associated with politically sensitive activities. From the instrumental side of the equation, most signs point to the Han Chinese as the main contenders for Tibetan elite rivalry, besides



Boycotts and Religious Networks 355

the fact that the Han Chinese are also overwhelmingly dominant in almost all aspects of the polity and economy. Effectively, an instrumental interpretation offers only a partial understanding of these events. This is because a strong normative dimension, framed in terms of Tibeto-Buddhist solidarity and morality, exists within these elite strategies. Ethnic animosity does not necessarily follow a purely economistic rationale, although certain economic aspects of a larger malaise become symbolic rallying points. These rallying points are those that touch most closely a nerve of dignity, rooted in conceptions of hierarchy, and conditioned by a sense of expediency. Muslims receive the brunt of Tibetan animosity because, on one hand, they are upwardly mobile in the economic hierarchy from a lower and stigmatized position, and on the other hand, they are not necessarily mobile in the social or political hierarchy, at least not within the Tibetan areas.37 Their economic rise comes to symbolize the exclusionary dynamics faced by both common and elite Tibetans within their local economies, while their social position makes it expedient for Tibetans to scapegoat them. Tibetans can construct contemporary histories of conflict with Muslims that draw on potent religious symbols, and this is easy for them to do precisely because of the primordialist nature of Tibetan-Muslim relations. Indeed, many Tibetans believe that the current global conflicts with Muslims were prophesied more than a thousand years ago in the Kalachakra Tantra commentaries. Many of the examples that educated Tibetans drew to make their case refer to informed historical analyses, such as the loss of Buddhist centers in Afghanistan to Islam, Iran, Kashmir, Khotan, and Sumatra. In this sense, it is interesting to note that even among Tibetan religious leaders who were opposed to or abstained from the boycott, many nonetheless shared a certain degree of consensus with other Tibetan elites on the need for Tibetans to take a more competitive stance toward Muslims. For instance, in extensive interviews that I carried out in summer 2004 with two lamas from different parts of Amdo, neither of whom overtly supported the boycott and could be considered moderates in their communities, both implicitly supported the broader principles underlying the boycott. They explained their positions through a subtle and skillful circular reasoning that mixed Buddhist logic with informed interpretations of history as well as rhetorical stereotypes that bordered on racism. In particular, both lamas portrayed the competition between Tibetans and Muslims in terms of a fundamental antagonism that has existed between Buddhism and Islam throughout history and which placed Tibetans on the defensive. Both focused on differences of orthopraxy to justify their cautiously racist characterizations of Muslims as naturally violent, tricky, and cheating, thereby explaining the reasons for Muslim success in the Tibetan economy as well as the reasons for Tibetan insecurity.38

356

Chapter 8

Thus, despite their opposition to the boycott in its overt aggressive form, both lamas ended by giving very similar justifications for the boycott. These ultimately returned to the economic issues of Muslim takeover that have been contemporary to the reform period. Both thought that the boycott could, in the end, have a positive outcome by stimulating Tibetans to start thinking about these issues, such as developing businesses and moving into new areas of the economy. These interviews corresponded with a series of discussions that I had with an ex-monk and close disciple of several prominent lamas in Golok in 2004. Apparently, some of the most senior lamas in Golok argued that Tibetans should not support Muslims on purely moral grounds given the way Muslims treat animals (as discussed previously with regard to the anti-slaughter movement in Serthar). The senior lamas avoided the politicized questions of boycotts, not necessarily because they were opposed to economic activism in principle but because it was important for a high lama to appear to be above the mundane concerns of politics. However, the moral ground that they provided implicitly condoned the stance of more junior and more politicized lamas, who explicitly connected the boycott to political economy concerns. In turn, the advice of the activist lamas was “do not fight, but do not give money.” Nonetheless, despite the apparent Buddhist principle of not fighting, Amdo Tibetans have not hesitated in the past to rise up in armed resistance against invading Muslim and Chinese armies (and against each other), with the religious leadership playing central mobilizing roles (see chapter 2). Pride in this martial heritage is still very much alive throughout Amdo, particularly considering that older generations alive today lived through periods of resistance against Ma Bufang in the 1930s and the PLA in the 1950s. In effect, the advice to resist Muslim economic dominance could have been easily misconstrued as a covert signal to take a heavyhanded stance toward Muslims. At the very least, this impulse would have been an interpretation that the lamas could only hope to attenuate, particularly that any open discussion of conflict would bring immediate retribution from the state, as it did in the case of events in Chentsa. Indeed, in an incident that broke out in Gande, Golok, in August 2007, just prior to my last visit there, an argument between a Muslim restaurateur and a Tibetan customer escalated into a clash involving hundreds of Tibetans (the Tibetan claimed he had found a human tooth in his soup, amid general complaints about mass poisonings). At the beginning of the clash, a local trulku (Chadrel Rimpoche) appealed to the Tibetans to stop, which eased the situation for a while, although the Tibetan crowd nonetheless rioted again, destroying and/or laying siege to a small mosque despite heavy police intervention (see RFA 2007).39 In response to the situation, a large armed police contingent was reportedly sent from Lanzhou. During my visit to the area in September 2007, I was told that at least one Tibetan



Boycotts and Religious Networks 357

had been sentenced to death for his leadership in the incident. The tense security environment that remained might help to explain why Golok was relatively calm during the protests that engulfed much of the surrounding region of Amdo in March and April 2008.

The Return of the Lama: Tibetan-Chinese Cooperation In contrast to these re-emerging Tibetan-Muslim fault lines of conflict, this section examines one important yet often overlooked dimension of contemporary Tibetan-Han relations: the financial revival of Tibetan Buddhism and its expansion into China through lama-patron networks. While most scholarly work on the political dimensions of Tibetan-Chinese relations focuses on ongoing political collaboration with or contestation of the Chinese state, whether violent or non-violent, or expressed in terms of secessionist or nonsecessionist nationalism,40 the issue of cooperative interactions with Han Chinese people serves to illustrate certain processes underlying modes of integration into the “new China” initiated by Tibetan religious elites. The brief discussion here is again mostly based on my fieldwork from 2003 to 2007, presented in Fischer (2007) and describing the situation prior to the protests in 2008. These findings are not updated or compared to the important recent contribution on this subject by Smyer Yu (2011). Rather, the focus here is to briefly explore the political economy dimensions of this issue in relation to the broader analysis of development and dislocation in this book. The reform period has witnessed a re-emergence of lama-patron relationships between Tibetan religious figures and private Han sponsors from mainland China, Taiwan, and other Chinese cultural areas. This pattern has followed the re-emergence of lamas (regular and trulkus) as significant nonstate actors within the local political economies of the Tibetan areas over the same period given that, despite several decades of Maoism, they still hold pre-eminence as indigenous sources of moral and legitimate authority among Tibetans, in particular, rural Tibetans. In this respect, the minority nationality laws, which were first conceived in the 1950s, abandoned during the Cultural Revolution and then revived from the early 1980s onward, included the rehabilitation of these ostracized or imprisoned traditional elites, together with the encouragement of minority languages, minority education systems, and the cautious liberalization of religious practice.41 One of the immediate effects of this opening was to allow for a religious and cultural revival among Tibetans and other minorities in China.42 This effectively opened up space for old and new Tibetan elites to begin cautiously and gradually exercising local authority, albeit the degree to which local autonomy was allowed in the Tibetan areas lagged far behind the Han

358

Chapter 8

areas of China given the political and strategic sensitivity of these areas. The trend of openness has also oscillated since the 1980s, depending on political trends in Beijing in conjunction with social and political conditions on the ground in Tibetan areas. Like political trends elsewhere in China, much of the lenience in the 1980s was curbed in the 1990s, particularly in the TAR, even while economic reform was accelerated. This being said, as noted in chapter 2 and argued by Robin (2011; 2012; forthcoming), this binary characterization of opening and closing is perhaps a simplistic representation of a dynamic that is better characterized as a constant exploration and testing of newly emerging, imprecise, ambiguous, and uncharted boundaries of the permitted within contemporary China. While many elite maneuvers can be understood as attempts to supplant the local state as the primary locus of legitimate authority, they ironically involve an important dimension of integration into China (and often degrees of rivalry between Tibetan elites). This particularly concerns the religious elites. Evidently, such maneuvers must be played out with care and political astuteness, for the ongoing activities of the religious elites are predicated on a certain degree of state lenience with respect to the religious revival. In other words, strategies to enhance local clout must themselves walk a fine line, accommodating the dominant authority of the PRC, which includes state intervention in the religious space when threat is perceived. Nonetheless, the accommodation of Chinese power by religious elites has also been characteristic throughout Tibetan history, which will be discussed further below. Such strategies of integration into China have been facilitated by the rising popularity of Tibetan Buddhism in “China proper,” reflective of the overall revival of religion in China more generally. This derives much from the fact that Tibetan Buddhism holds a similar fascination for the Chinese, especially among the wealthy coastal Chinese, as it holds for Westerners. An added dimension is the much closer cultural common ground between Tibetans and Chinese than between Tibetans and Westerners.43 This is especially apparent in terms of the social roles that patrons are expected to play toward religious teachers, and encounters between these two diasporas (Tibetan and Chinese) operate smoothly with an implicit understanding of the lama-patron relationship.44 Lamas have thus capitalized on the auspicious combination of curiosity and cultural affinity by traveling, teaching, and even living part-time or full-time in China, Taiwan, Hong Kong, or Singapore, or in Chinese communities in Malaysia, Indonesia, Canada, the United States, and Europe, and in other Asian countries such as Japan and Korea.45 The extent to which these conditions have been maintained or else closed down since 2008 is not clear given restrictions imposed on lamas to travel outside of Tibetan areas, although some reports indicate that the more general patterns have persisted in more discrete forms.



Boycotts and Religious Networks 359

Neo-Lama-Patronism Current Tibetan-Han cordiality is itself reflective of the long history of coexistence between Tibetan and Chinese powers. As discussed in chapter 2, political and economic relationships between Tibet and China came to be conceived through lama-patron46 conceptions of co-existence, which defined Tibetan notions of inter-state and inter-elite relations with the successive Chinese dynasties from the Yuan in the thirteenth century onward. Offshoots of such associations persisted into the Republican era, as discussed by Tuttle (2005),47 and were eclipsed only during the heyday of radical Maoism. While this lama-patron concept is somewhat complex,48 the point here is not to analyze the historical practice but to reflect on the appearance of analogous forms of relating to Chinese power by Tibetan religious elites in the present. While much has been said about the religious revival in Tibet, much less has been said about the increasing role of “Chinese” (broadly defined) in privately financing this revival, along with Tibetan efforts to integrate Chinese people into the organizations that lead the revival (the 2011 book by Dan Smyer Yu is a recent exception). “Chinese” in this regard refers to a transnational Chinese identity, including overseas Chinese, as noted above, as well as domestic Chinese from the wealthy coastal provinces and from the interior, such as from Chengdu, Lanzhou, and Xining. The domestic element also includes some members of other minority groups in China such as Manchurians and Mongolians. In many cases the domestic Chinese include officials, even members of the Communist Party and related paragovernment organs. For instance, this was noted to me by two members of a Chinese family in Chengdu with connections to both the Communist Party and to several lamas in Gansu and Qinghai. The two lamas mentioned at the end of the previous section of this chapter also talked to me at length about their disciples throughout China, many of whom were cadres and/or members of the Communist Party. Similarly, the ex-monk in Golok, also mentioned above, described at length the status of the Chinese patrons of the lamas he worked for. While records of donations between patrons and religious institutions might exist within monastic bookkeeping, these are not publicly available and I did not have access to any. Transactions generally take place informally as symbolic gestures of devotion or merit-accumulation by the giver. There is an implicit understanding within this exchange that services will be rendered by the lama to the donor, although not as a direct consequence of the donation (which would be tantamount to selling religious services). In effect, the exchange operates within an understanding of a hierarchical relationship that deepens the karmic bond between the two people, through devotion from the side of the patron and through compassion from the side of the lama.

360

Chapter 8

Given my lack of access to any records, some anecdotal examples suffice to provide insight into the depth of private financial relationships between Chinese sponsors and these Tibetan institutions (see the work of Smyer Yu for a book-length, non-anecdotal study). Most of the major monasteries or religious organizations that I came across in my fieldwork, and in particular most of the larger construction works at these institutions, had a large element of sponsorship from private Chinese sources. For instance, in Ngawa County in Sichuan, noted earlier for its successful and active Tibetan business community as well as its central role in the boycott movement, has a particularly high density of monasteries, many of which are funded by Chinese sponsors. The Jonang order, one of the smaller Tibetan Buddhist orders centered in the neighboring Dzamthang County in Ngawa Prefecture, had apparently been very successful in attracting Chinese donors and disciples due to the charisma of its late leader, and permanent Jonang “embassies” were maintained in Chengdu and Beijing at the time.49 Across the Sichuan-Qinghai border in Golok, there is strong evidence of deep networking by Golok lamas with domestic and overseas Chinese. One Western scholar living there and with close connections to a local monastery noted, during discussions in May 2004, that the gathering of Chinese disciples is a common activity for many Golok lamas. Chinese disciples, particularly those from the coastal provinces or from Taiwan, were known to give very generously to lamas whom they are devoted to, often offering individual donations in excess of 1,000 US dollars and in some cases more than 10,000 US dollars per visit. Groups of devoted Chinese disciples could thereby bring in considerable amounts of donations, in excess of 100,000 US dollars, within a single visit.50 One Nyingma monastery in Golok that I visited several times between 2004 and 2007 had several trulku lineages, two of which were permanently based in overseas Chinese communities, one in Taiwan and another who divided his time between Vancouver and California. A third lama, who was more regularly based in the area, made regular visits to Taiwan and had many domestic and overseas Chinese disciples, who were financing his many activities such as monastery improvements, an orphanage, a school, and the building of extensive religious monuments. In the same monastery, a large prayer hall with a seating capacity of 3,000 monks was being built in 2004 entirely from the donation of a single Chinese woman from Shanghai. Similar patterns can be seen throughout the Tibetan areas outside the TAR. For instance, lamas in the main monasteries of Kumbum near Xining, Rebgong, and Labrang are all known to have considerable numbers of Chinese disciples. In the main monastery in Rongwo, Rebgong, at least two of the major temples that were under construction during my fieldwork there in 2004, and several others that had been recently completed were either fully or partially financed by Chinese donors, some from Beijing. Further



Boycotts and Religious Networks 361

south in Kham (Sichuan), monasteries that I visited in Kardze and Tranggo Counties had received funding from domestic or overseas Chinese, and in each case had several lineages of lamas living abroad in Chinese communities. This pattern of Chinese sponsorship has also been developing with much vigor between overseas Chinese and exile Tibetans in India and Nepal. For instance, it is likely that one of the main sources of funding for the Tibetan exile community in India—or at least for many of the religious institutions there, including Namgyal Monastery, the personal monastery of the Dalai Lama—are also derived from the Chinese diaspora and other East Asian sources.51 Even in the much more sensitive TAR, private Chinese sponsors have been involved in funding construction or other activities in various monasteries, such as Drepung or Sera, although these involvements tend to be much more hushed given the heightened political sensitivity of the region. Lamas from the TAR have also been much more restricted in their travelling and thus not as successful as lamas from outside the TAR in gathering disciples in China and abroad. Nonetheless, they are one of the attractions of the currently booming tourist industry from China to the TAR, which increasingly provides them with a captive Chinese audience searching for Shangri-la. Notably, many Chinese devotees came to the TAR to have audiences with the young seventeenth Karmapa when he was still residing there, before he escaped into exile in 2000.52 Having one lama networked in such a way can be of critical support to a monastery. Lamas with Chinese disciples are often rich, even by Chinese standards. Depending on the lama, a large part of this wealth is directed to the monastery and community. Despite the noted presence of many charlatans profiting from naïve Chinese devotees, truly eminent lamas who are locally recognized for their scholarship and achievements are particularly famed among the Chinese faithful. Some lamas will take on additional social activist roles by using their Chinese links and their local clout to establish schools, clinics, and so forth, alongside their more religious activities, and are often supported in these activities by local Tibetan businesspeople and officials. Such activities have undoubtedly been restricted since the 2008 protests and especially since the wave of self-immolations from 2011 onward,53 although they have reportedly continued, even if discretely. Alongside these financial relationships there have also been subtle efforts to integrate Chinese into Tibetan religious organizations. At the higher level this includes recognition of Chinese or half-Chinese children as trulkus, several cases of which were reported to me during the course of my fieldwork. One scholar told me about a conflict in a monastery near Xining in 2004 over the choice of the reincarnation of the head trulku, with the monastery leadership selecting a Chinese child while another faction disagreed and forwarded its own choice of a Tibetan child. Indeed, this skillful use of

362

Chapter 8

“recognitions” by Tibetan religious leaders to forge allegiances and loyalties is also used in the West, such as with the 1997 recognition of the US martial arts movie star Steven Seagal as an “emancipation” by Penor Rimpoche, then the exile head of the Nyingma order. On a less status-oriented level, another phenomenon has been the increasing presence of Han Chinese monks within Tibetan monasteries. One of the most notable cases is the Serthar Institute in Kardze Prefecture in Sichuan, founded by the late Khenpo Jigme Phuntsok, which had an estimated one thousand Chinese monks, largely from mainland China, prior to the large demolitions conducted by the Sichuanese authorities in 2001 (see Germano 1998 and TIN 2001). More generally, it is common to see or meet a scattering of Chinese monks studying with various lamas in monasteries throughout Tibet. During my fieldwork, I came across several, often in surprisingly remote areas. While the principal motivations are obviously spiritual, one of these monks explained that attracting Chinese monks can also be a key for these lamas to make contact with Chinese society, given that the Chinese monks can act as translators or attract disciples and sponsors. At the more popular level, several times I came across Sichuanese Han during my fieldwork in Sichuan who had been told by a lama that they were a monk or even a lama in a past life (although the latter might be a misinterpretation, given that non-erudite Chinese and the Chinese press usually call monks “lamas”). This is an imaginary that is as equally potent for the Han as it is for Tibetans or even Westerners. Similarly, during my visits to Chengdu from 2003 to 2007, I regularly noted how many Han Chinese have a natural respect or devotion for the esoteric sides of Tibetan Buddhism and will resort to using lamas for a variety of activities such as prayers or divinations. Political Economy Implications While there is obviously a strong political economy motivation for Tibetan religious elites that runs in harmony with their religious objectives, such motivations are less obvious for the Chinese involved in these networks. Their investments in Tibetan Buddhist religious organizations can be understood in spiritual terms, whether based on a superstitious or profound understanding of Buddhist doctrine, but rarely in economic terms. Or, economic concerns might be present in “spiritual materialist” terms, to borrow from Chogyam Trungpa (1973), in the sense that it is common for Chinese disciples to ask lamas to bless their house, car, or business. In some cases, guilt about the past and the wish to make retribution on behalf of Chinese society might motivate some of these devotees. In other cases, giving is inspired purely by religious practice. For instance, I met a Fujian real estate investor in a pastoral area of Kham who was spending four months wandering



Boycotts and Religious Networks 363

the region looking for worthy religious projects to fund as part of his practice of generosity, inspired by his religious teachers. Overall, most of these motivations derive from the renewed search for spirituality among such Chinese. The interface of Tibetan Buddhism with the renewal of other faiths in China also increasingly offers space for consensual forms of cooperation between Tibetans and Han Chinese based on mutually understood and respected symbols and practices. This in turn possibly plays a role in reinforcing a psychological sense of national unity among these Chinese according to a discourse that is alternative to the mainstream state ideology of socialism or secular nationalism, which has alienated many of them, whether due to the mayhem of Maoism or to the rampant consumerism of the later reform period. This interpretation is similar to the discourses in the 1930s described by Tuttle (2005) and in chapter 2, whereby the Nationalists turned to a religious rhetoric in their attempts to forge a national consciousness. Indeed, Tuttle emphasizes that this earlier strategy was elaborated by Nationalist Party leader Dai Jitao, a devout Buddhist. Besides the fact that many officials themselves are increasingly religious, this sense might help to understand why the state tolerates or turns a blind eye to these increasing religious interactions today, given that they encourage and legitimate the tightening integration of Tibetan areas into China, albert not along lines deemed ideal by official ideology. From the Tibetan side, the instrumentalism of these interactions is obvious. They enhance the status and clout of religious elites as alternative sources of authority in the Tibetan areas, derived in part from the wealth that they can mobilize and bring into their communities. Outside sources of funding also allow some lamas to counter official state criticisms of religious elites as parasites, profiting from their spiritual exploitation of the masses, given that they allow these lamas to mobilize funds for various activities without drawing on local resources.54 However, as in the case of Tibetan-Muslim conflict, the lama-patron interactions cannot be understood primarily through an instrumentalist interpretation. Normative concerns again play a particularly strong role, such as the preservation and promotion of Tibetan Buddhism, and of religious society in Tibet as the best means to do this. In this sense, it is difficult to disassociate the instrumental from the normative, given that indigenous discourses clearly see the instrumental as legitimate means to serve these greater normative ends.55 Moreover, these interactions enact a sense of dignity and agency for Tibetans. For instance, as discussed in the previous chapter, Tibetan conceptions of hierarchy do not necessarily conform to the modern “underdeveloped” status that has been attributed to them. The religious leadership is central to a counter-perception of status, in the sense that they see themselves as having a vanguard civilizing role in the rest of Asia

364

Chapter 8

and even the world, and especially in China, where the culture has been broadly influenced by Buddhism although rendered “irreligious” and “amoral” by communism. While working within Chinese dominance, the lama-patron networks serve to buttress this collective self-perception of morally superior agency.56 There is thus an interesting dialectic between the definite nationalist dimensions underlying religious revival in Tibet and these Tibetan-Chinese religious interactions that carry an integrative role within overarching Chinese authority. For instance, while expanding their influence in China, many lamas and monasteries often, albeit discreetly, also maintain tight contacts with the Tibetan exile community in India, thereby not only disseminating nationalist ideas from India but also cultivating rapprochement between the exiles and domestic Chinese devotees. Keenly aware of this factor, the Dalai Lama has skillfully shown sensitivity to the dynamics of this rapprochement. For instance, I was told by one Falun Gong devotee in Chengdu in 2003 that the Dalai Lama strategically won much support in the Falun Gong community in China by publicly suggesting that the religious movement must be good given that so many people have devotion in it (in contrast to Chinese Buddhists, who criticize Falun Gong for plagiarizing Buddhism). Furthermore, the nationalism advocated by religious elites is obviously of a religious nature, reflecting the traditional ideology of power in the Tibetan areas, that is, the union of the spiritual and temporal under the authority of lamas/trulkus.57 In many respects, this form of religious nationalism carries an alternative vision of Tibetan nationhood, one that nominally and temporally accepts Chinese dominance while monopolizing legitimacy among the local Tibetan population, and using this legitimacy to negotiate increased spaces for local authority and power. This is similar to the historical posturing of monasteries, most of which were more concerned with maintaining their authority within the Tibetan social system and co-opting various foreign rulers to this end than with working toward a strong central Tibetan rule.58 Indeed, the contemporary fusions between religion, nationalism, and subaltern finance serve as an archetypal microcosm for the changing dynamics of pan-Tibetan nationalism more generally, which, as noted by Tsering Shakya (NLR 2008), reflect interesting synergies within pan-Asian religious solidarities, ironically centered in China but in support of indigenous and traditional forms of Tibetan leadership. Certainly, if the current political economy of the monasteries is considered as a beacon, the future of Tibet is within China, albeit with an autonomy buffeted by religious activism that leverages the advantages of Tibetan elites in this realm and that returns Tibetans to their self-perceived role as civilizers within Asia and abroad.



Boycotts and Religious Networks 365

Conclusion In the cases of both Tibetan-Muslim conflict and Tibetan-Han cooperative interactions, a religiously rooted nationalism can be seen as one of the primary channels through which many Tibetans articulate their experiences within the radical restructuring of their societies, polities, and economies since the beginning of the reform period in the early 1980s. In the first case, modernity in Tibet has driven a recalibration of the place of Muslims in local Tibetan society due to the breakdown of traditional economic hierarchies. The irony is that, even while the rise of Muslims challenges Tibetan conceptions of hierarchy, the space in which Tibetans must compete with Muslims also irks their conceptions of dignity. Tibetans are left with a disoriented sense of agency in the face of contemporary transformations, one which they have double difficulty to deal with given the iron grip of Chinese rule over any manifestations of “minority” nationalism. One legitimizing discourse that they have therefore turned to in their backlash against Muslims has been the moral discourse of Buddhism, tacitly supported by various elite and common members of the religious institutions in Tibet. Alternatively, religious revival in Tibet has also come to play a strategic role in creating a variety of social and economic networks within Chinese society, both domestic and overseas. To a certain extent these networks have been subsidizing the revival of Buddhism in recent years, through alternative non-state financial channels that are relatively independent from state patronage and rooted in the heartland of the Tibetan cultural space (monasteries). Leadership is based in the traditional Tibetan religious elite, most of whom had been persecuted during the Maoist period but have since made a slow and subtle comeback. These evolving networks carry an important implication in terms of contested inclusion and double movement. In essence, religious elites enhance their status and clout as alternative local sources of authority in the Tibetan areas by leveraging their particular “niche” specialization within China, simultaneously feeding a sense of purpose and mission that this symbolizes for many Tibetans, and demonstrating their ability to mobilize external sources of finance in support of locally determined forms of accumulation (at least this was the case up to 2008). In this way, religious revival serves as a means for traditional indigenous elites to buttress themselves with powerful but subtle alter-ideologies of power that carry far more local legitimacy than the state and with which they can bargain for informal power sharing at the local level despite official state animosity toward religion. Indeed, events related to the choice of the reincarnation of the Panchen Lama in 1995 and ongoing efforts by Beijing to legitimate its choice among Tibetans, as discussed in chapter 2, illustrate the futile and almost ridiculous attempts by Beijing to compete for legitimacy on this level. The tacit recognition of the author-

366

Chapter 8

ity of religious elites by the state with regard to Tibetan boycotts of Muslim businesses is one expression of this influence, as has also been the case with respect to the resolution of other conflicts, such as grassland conflicts, as discussed by Yeh (2003b) and Pirie (2005). The double movement of this situation is to be emphasized, in the sense of the protective impulse to re-establish hierarchical orderings in the midst of the dislocations wrought by modern development. While the contemporary religious revival carries a definite nationalist overtone, it simultaneously implies a strategy of integration within overarching Chinese dominance. Rather than rejecting integration, religious elites aim at redefining the mode of integration. Religious nationalism in this sense fuels indigenously led efforts to legitimate alternative terms of inclusion within the new China, on the basis of historical religious animosities or affinities, and in a manner that preserves an indigenous sense of hierarchy and dignity, at least in the sense defined by the dominant sources of authority within Tibetan society itself. In other words, in relation to both Muslims and Han Chinese, local political economy motivations interact with inherited cultural notions of self-manifesting power to produce a political double movement within integration. The implications of this will be discussed in the final chapter, the conclusion of this book.

Notes   1.  Yang (1997) does not engage much with the specific inter-ethnic dimensions of these “commodity wars.” For a Tibet-specific interpretation, see Fischer (2002, 23).   2.  See Winkler (2008a; 2008b).   3.  See Sulek (2011) for an interesting discussion of more recent dynamics with regard to the sale of sheep herds by pastoralists in Golok, which she attributes to the caterpillar fungus boom.  4. Many cases of selling state-owned trading companies to Muslims were recounted to me by Tibetan, Muslim, and Chinese scholars and officials during fieldwork in Qinghai, Gansu, Sichuan, and the TAR in 2004, as well as by several Western scholars and INGO staff.   5.  As analyzed further in this chapter, the local officials (Tibetan and Chinese) and Muslim scholars or teachers whom I interviewed tended to differentiate between local Muslims settled in Tibetan areas from those who are not. Cheating and communal tensions were blamed on the latter.   6.  For an example of a conflict over questions of pricing and forced procurement with a local Han-owned abattoir in Derge County in Sichuan in 2006, see RFA (2006). Although this case was with a Han Chinese owner, the similarity to Tibetan-Muslim conflicts is apparent, as described later in this chapter.   7.  See Zhang (2012) for an interesting study of Tibetan carpet production in Lhasa.



Boycotts and Religious Networks 367

  8.  This information derives from several interviews from June to August 2004 with Dr. Ma Cheng Jun, a Salar professor in the Qinghai Minorities University in Xining, and from a series of interviews conducted during a visit to Xunhua County in June 2004, which included the manager of the main factory and several county officials. Goodman (2004, 396) also notes that the other richest man in Qinghai is a Hui Muslim from Hualong County.   9.  Interview, Dr. Ma, Xining, June 2004. 10.  Interview, Dr. Ma; this is not to say that Muslims in other parts of China were not involved in manufacturing. For instance, nearby Linxia appears to have a longer tradition in textile manufacturing. 11.  Based on observations and interviews made during a visit to the factory in June 2004 and accounts from a group of ten master’s students (including three Muslims) who joined me for the visit. The factory workers earned twenty yuan a day in 2004, and worked eight hours a day for six days a week with no bonus (only the managers received bonuses). 12.  For an excellent ethnography of both policies in an undisclosed pastoral location of Qinghai, see Cencetti (2012). 13.  This is based on unpublished GDP data from the prefecture statistical department, obtained from an official. Interestingly, as in the TAR, the overall GDP of the prefecture was recessionary up to 1995 and then grew rapidly from 1996 onward. 14.  The senior Han official, interviewed in July 2004, said that they were watching the price of aluminum to decide whether to keep operating. The factory had reopened when I next visited in November 2004, although at a much reduced output. 15.  A longer version of this section was published as a part of Fischer (2005b; 2008b). 16.  I have found that Tibetan elites all over Tibet, including those in Beijing, to be well versed in current political and religious matters in India. 17.  This video footage was discussed by Holly Gayley during a conference presentation at the International Association of Tibetan Studies, Bonn, August 2006. 18.  For instance, in November 2004, a Tibetan scholar in Lhasa reported to me one well-publicized case of Muslims trafficking young rural Tibetan girls to older or handicapped Muslim men in Gansu. However, on further questioning, she admitted that Tibetans had also been involved in trafficking. 19. Such prejudices were common in conversations with Tibetans and Han, particularly in Qinghai, to an extent that many of the Tibetans I interviewed simply assumed that Muslim businesspeople cheat, even if they personally had no experience (or awareness) of having been cheated by a Muslim. 20.  The locations of various localized boycotts were reported to me by one Tibetan scholar in Xining during an interview in July 2004. 21.  Many Tibetans had heard about this lama, but all of those whom I talked to either did not know his name or else were not willing to divulge it, probably because of the sensitivity of these issues in local politics. I believe that they were referring to Khenpo Jigme Phuntsok, as mentioned above with respect to the video promoting vegetarianism, in which case, they would have been wrongly referring to Serthar as Ngawa. Serthar is located in Kardze (Ch. Ganzi) Prefecture in Sichuan, not Ngawa Prefecture, although it is technically part of Amdo. Hence, “Ngawa” might be a general term used by Qinghai Tibetans to refer to the parts of Amdo located in Sichuan.

368

Chapter 8

22. The fact that Ngawa Tibetans seem to break the mold with respect to Tibetan stigmatization of business and trade probably owes to their strategic location as a trade center between Kham and Amdo. Further research on this issue would be very fruitful. 23.  From discussions with other Western Tibet scholars, it seems that these sentiments in Lhasa can be traced even further back into the mid-1980s, if not before, including restaurant conflicts and commercial conflicts similar to the ones in Amdo described below. 24. This perception was expressed to me on several separate occasions during fieldwork in 2004 by two high school students from Chentsa who were studying in Rongwo, three high school teachers in Rongwo, one unemployed high school graduate in Chentsa, and three Tibetan masters students in Xining. 25.  I discussed these cases with a senior Chinese official from Malho Prefecture who insisted that the government had tried to deal with each crime without discrimination, treating all equally under the law. 26.  At the time, I had rarely met a Tibetan in Tibet who was not an ardent supporter of Bush Junior. There were two reasons for this, one that he was perceived to act strongly against China, and the other that he attacked a Muslim tyrant (Saddam Hussein). The detonation of the two Buddha statues in Afghanistan in 2001 was well known to educated Tibetans and did not endear them to the Taliban either. Following 9/11, rumors abounded in Dharamsala, India, where I was living at the time, that certain lamas were claiming that these events had been prophesied in the Kalachakra Tantra and that Bush was one of the prophesied Shambhala Warriors who had come to defeat the Muslim hordes. 27.  The events surrounding the poisoning, including the newspaper articles, were recounted to me during a visit to Golok in May 2004 by a Western scholar who had been living there during the events (I did not consult the newspaper articles myself). 28.  I first noted this during a stopover in Bema County, Golok, in May 2004, where a popular Sichuanese-owned restaurant had been bought from a Muslim owner during the boycott. 29.  For instance, Yi (2005a, 9) notes that Tibetan-owned businesses are much more numerous in Rongwo Town than they were in the 1990s. 30.  The exact numbers are impossible to record given that tracking this myth became one of the foci of my mobile fieldwork (see introduction). Nonetheless, versions of the story were told to me by at least thirty high school students, ten teachers, ten scholars, five Tibetan officials, two lamas, and over fifty farmers, nomads, or common lesseducated workers. Much more significant is the fact that, of all those interviewed, only two scholars and one journalist admitted outright that the myths were false. Several other scholars and teachers, and one lama, caved only after I pointed out several of the inconsistencies in the myths, as noted in this section. Most other people insisted on the validity of the myths, even after I pointed out the inconsistencies. 31.  Sources included local and foreign scholars, Tibetan and Chinese officials, lamas, and Tibetan journalists, teachers, and students. 32.  Xining and Lanzhou are important intellectual nerve centers for Tibetans given that they are the exclusive centers of provincial tertiary education in each province. Chengdu plays a similar role for both Kham and Amdo Tibetans in Sichuan. 33.  For instance, many (although not all) of the high school and university students, and teachers whom I interviewed in Qinghai had studied with lamas.



Boycotts and Religious Networks 369

34.  I came across many examples of such interaction during my fieldwork. For instance, to name a few, one wealthy businessman in Ngawa was supporting a local monastery and primary school. In Rongwo, I met an official who was related to a reincarnate lama, and they were mutually supporting each other’s education projects. Also see Costello (2002). 35.  A meeting in Golok was reported to me by a local resident there. A similar meeting in Chentsa following the clash in January 2003 was reported to me by a senior official from Malho Prefecture. 36.  I am indebted to David Keen for pointing out to me how this more subtle interpretation of instrumentalism could describe some aspects of my analysis of contemporary Tibetan-Muslim ethnic conflict. 37.  For instance, see Yi (2005a, 9) on the marginalization of Muslims from local county politics in Rebgong. 38.  For a more detailed discussion of these interviews, see Fischer (2008b). 39.  The report is not clear whether one or two mosques were involved. If two, then one was first destroyed and then another was under siege. 40.  For some of the most rigorous scholarly sources on these issues, see Schwartz (1994), Goldstein (1997), Rabgey and Sharlho (2004), Sperling (2004), and Barnett (2006). 41. For official statements on the ethnic minority laws in China, see PRC (2000a; 2005). Specifically on the TAR, see PRC (1998; 2000b; 2004). For some scholarly work on China, see Postiglione (1992), MacKerras (1994), Sautman (1999), and Iredale et al. (2001). With specific reference to Tibet, see Goldstein (1997; 1998). 42.  With respect to Tibet, see the edited collection by Goldstein and Kapstein (1998). See Makley (2007) for a focus on Amdo. 43.  Obviously, cultural cordiality is conditioned by the fact that the Chinese effectively maintain dominance in the state and thereby profit from this deterrence. However, besides the usual cases of hooliganism or bar brawls that are typical in Qinghai between Tibetans and Chinese (and among Tibetans themselves), the two groups generally share many points of cultural affinity, particularly with respect to religion, but also in their ability to bond in similar social practices, such as eating the same food, drinking heavily, singing karaoke, and dancing. 44. I observed this frequently during my time living with (and serving) a reincarnate lama in India from 1998 to 2001. This lama, from Central Tibet, had followings of disciples from Amdo, Mongolia, Taiwan, Russia, and North America, and groups from all these locations would regularly visit. Generally, Chinese and Mongolian disciples forthrightly fulfilled their role as patrons, whereas Westerners tended to be much more aware of quality control, which can become embarrassingly complicated when donor sympathy has been garnered through idealized stereotypes or cultural misunderstandings. 45.  For instance, see the contributions by Zablocki (2009a; 2009b) on transnational Tibetan Buddhism. 46.  This is often referred to as the “priest-patron” relationship, although the term “priest” is a poor translation for the term “lama” (see the glossary at the beginning of the book). 47.  Also see various contributions in the edited volume by Kapstein (2009).

370

Chapter 8

48.  For instance, see Sperling (2004, x). He argues that “[t]he priest-patron relationship was simply not a barometer of Tibet’s [political] status, in spite of current Tibetan use of it as such.” 49.  Some local Tibetans claimed that the Jonang are among the most organized Tibetan order in their dealings with the Chinese. During fieldwork in Golok, I was told by the ex-monk mentioned above that the head lama of this order was very popular in China and his death a few years earlier, along with accompanying miraculous signs, was reported in the Chinese press. 50. References to donation values are taken from interviews with the Golok ex-monk mentioned above and with one Western researcher who was also closely connected to these networks. I also observed similar amounts being disbursed by Chinese disciples during my time living with a reincarnate lama in India from 1998 to 2001, as noted earlier. 51.  This estimate is derived entirely from my own informal qualitative judgments based on the known sources of funding during the time I was living in the exile community in India from 1995 to 2001. 52.  The Karmapa trulku lineage is one of the highest in Tibet and the highest of the Karma Kagyu order. The current reincarnation escaped into exile in 2000 at the age of fifteen, much to the annoyance of the PRC given that it had officially sanctioned his recognition. During a Chinese New Year gathering in a Tibetan town near Chengdu in January 2004, my Chinese and Tibetan hosts showed me a somewhat hagiographic video of the Karmapa that had been produced in Chinese by a Taiwanese scholar and was apparently a popular underground video in China, among both Tibetans and Chinese. 53.  Notably, Sopa Trulku, a reincarnate lama and head lama of Tongkyap Monastery in Darlak, Golok, who self-immolated on 6 January 2012, was apparently involved in such types of social welfare activities. See VOA (2012). 54. For instance, one disciple of a Golok lama interviewed in May 2004 explained to me that local governments often use these reasons to restrict the size of monasteries or the activities of the lamas. 55.  This insight is difficult to source given that it derives from innumerable discussions with lamas and monks, many recently arrived from Tibet, along with many public teachings by the Dalai Lama and other lamas, all during my time in the exile community in India from 1995 to 2001, as well as from discussion with lamas and monks in Tibet from 2003 to 2007. 56.  See Zenz (2013) for a similar and very interesting discussion of how Tibetan educators in Qinghai (in this case, educators in the modern secular education system) use discourses of development and “backwardness” to promote an alternative civilizing project of “modern Tibetanness” that is based on the study of the “Tibetan language-culture.” 57.  See Dreyfus (2002) on Tibetan religious nationalism. 58. For instance, see the excellent histories by Goldstein (1989b) and Shakya (1999), both of which make this point.

9 Conclusion: From Polarization to Protest in Contemporary Tibet

The research underlying this book has been motivated by a concern that the strategy of the central government of the PRC to resolve the contested political status of Tibet through the force of massive subsidies is accelerating the marginalization of Tibetans within their own homeland. However, such marginalization is difficult to portray through poverty statistics. To speak in terms of absolute poverty or living standards is, to a certain degree, to fall prey to the politics of statistical representation by the government. Certainly, Tibetans on average have been gradually improving their economic situation through the force of “trickle down” (and, in many cases, caterpillar fungus). Year by year, they consume more and more meat, butter, instant noodles, and bottled drinks; they have more and more mobile phones, motorbikes, televisions, and other durable goods; and they spend more and more money in Internet cafés and karaoke bars. Indeed, a Western manager of an INGO working on HIV and sex work in the TAR reported to me in 2006 that the fastest-growing element of sex work in Lhasa appears to be Tibetan men purchasing sexual services from Tibetan women (usually migrants from rural areas or small towns). General levels of health also seem to have improved since the beginning of the reform period, at least according to official accounts of mortality rates and life expectancy (levels of tuberculosis and hepatitis B in the population are nonetheless huge contemporary worries). Illiteracy and levels of schooling have also gradually improved. And so the officially sanctioned Chinese press and even much of the Western press cite cases of individual Tibetans who seem satisfied with the opportunities that have opened in the wake of the complete overhaul of their economy. However, this tells us little about the ability of Tibetans 371

372

Chapter 9

to self-determine the paths that their lives and livelihoods take within the context of the rapid social and economic transformations causing these marginal increments of wealth. It also tells us little about why people might become increasingly discontent amid rising prosperity. It would be surprising if economic improvements had not occurred for the average Tibetan given the sheer torrent of subsidies that the central government has been spending and investing in the TAR and, increasingly, in other Tibetan areas. Moreover, the creeping pace of change in illiteracy rates and post-primary schooling rates would appear anomalous with the sheer speed of economic changes. In this respect, one obvious adverse aspect of economic growth in Tibet since the mid-1990s that might indicate cause for grievance has been high and, at times, rapidly increasing inequality, whether urban-rural or intra-urban. If official statistics are to be relied on, urban-rural household income inequality in the TAR reached 5.5 in 2001, a height that was never before witnessed in the PRC at a provincial level of aggregation. While this peak was subsequently corrected, at least down to an average (rising) level of inequality in the rest of western China, intra-urban household inequality appeared to have taken over as the driving force of polarization, as represented by the difference between average wages of employed people in formal urban units (including many migrants) and average urban (permanent resident) household incomes. The ratio of these two in the TAR again reached a peak in 2007, at levels never before observed in the PRC (according to this measure, at a provincial level of aggregation). The sharp increase in urban inequality is particularly significant given that it occurred at a key moment in the transition of local Tibetan labor out of agriculture, although the above measures do not reflect these migrants and thus the increase cannot be explained by the entry of rural migrants into the urban income distribution, as might be implied by the arguments of Sautman (2008) or Hu and Salazar (2008). More relevant is the fact that the sharp increase in urban inequality occurred in the midst of substantial net in-migration of non-Tibetans (mostly Han) from outside the TAR and was more or less instituted by the government through its subsidy, wage, and state-sector employment policies. These drivers of urban inequality have thereby had a powerful effect on establishing the foundations of future social stratification in the urban areas of the TAR. Similar dynamics appear to have occurred in other Tibetan areas outside the TAR, with local variations but common themes of subordination in urban employment. Such indications of inequality would therefore seem to offer important clues of the marginalization and discontent that have accompanied growth. Rising inequality would reflect a heightening of the hurdles faced by the poorer strata of society to access the average level of norms operating within an economy, and in particular the norms operating where



Conclusion: From Polarization to Protest in Contemporary Tibet 373

most of the opportunities of growth are being generated. Marginalization understood this way is more or less akin to a relative notion of poverty or deprivation, in much the same way that the concept of social exclusion is commonly understood in the development studies, economics, or social policy literatures. In my earlier work (e.g., Fischer 2005a), I focused mostly on this understanding of relative deprivation by referring to the dilemma of development in Tibet as one of “ethnically exclusionary growth.” I argued that the instituted structures of growth tend toward an exclusion of large sections of the population from the dynamic sectors of the local economy, regardless of whether or not such exclusion is actually intended. In other words, effective discrimination can arise due to disjunctures between the existing skill sets in a labor force and the skill sets that are demanded by patterns of growth (whether these are in terms of formal schooling qualifications or in terms of informal cultural or linguistic qualifications such as Chinese fluency). Such disjunctures might be iteratively related to intentional discrimination, whether at the level of personal interactions (e.g., see ICT 2012a) or else through impersonal policy making that feeds into the structural disjunctures and various institutional norms and practices in the first place. Both aspects of this iteration can be observed in Tibet and are reinforced by the increasingly externalized and polarized flows of wealth in the economy that bear little relation to the indigeneity of the Tibetan areas. Those excluded from privileged access to these flows, whether the majority of Tibetans or even many poor Han Chinese migrants, are in turn progressively marginalized from these dominant processes of wealth circulation driving growth, even though their material consumption might have improved in the process given the sheer amount of subsidies. The combination of these factors provides a forceful argument for an understanding of exclusion as structural, institutional, and agentive processes operating within inequality and relative deprivation. However, marginalization and exclusion cannot be restricted to this understanding of relative deprivation. For one, it was not clear to me in my fieldwork whether Tibetans were necessarily the poorest in China, even though the income statistics show them to be and they definitely possess among the worst average levels of formal schooling and among the worst mortality rates and longevity in China (albeit all improving over time). Indeed, much of the livelihood strategies taken by rural Tibetan households and their related social responses to development are arguably better interpreted through an understanding that rural Tibetans are not relatively poor in China or do not perceive themselves as such, at least not from the perspective of what I have called “subsistence capacity.” In addition, we can observe important exclusionary processes operating within middle and upper strata of the regional labor hierarchy. The fact

374

Chapter 9

that the key arenas of employment exclusion in Tibet are urban, in a region where almost three-quarters of the labor force was working in farming or herding in 2000, implies that some of the strongest exclusionary pressures have been faced by a minority within this labor force, much of which has been relatively privileged rather than deprived. This includes upwardly mobile rural migrants, high school and university graduates, state-sector staff and workers, and others who in fact possess the greatest potential to overcome the skills disjunctures or culturally defined axes of discrimination that obstruct the majority of the Tibetan population. These strata of the labor force are also under considerable competitive pressure, albeit of a different nature than the competition faced by those lower in the hierarchy. In particular, the sharp retrenchment of state-sector employment as a share of total urban employment, which occurred much more sharply in the TAR than in other provinces in China, has limited the government’s ability to use such employment as a means to alleviate some of these competitive pressures and, instead, has put downward pressure on elite and/or upwardly mobile Tibetans within the local urban labor hierarchy. Exclusion in this sense needs to be understood and evaluated relative to comparable cohorts of students and employees, with similar levels of education and aspiring to enter similar sectors of employment. Rapidly inflating institutional norms within such sectors have accompanied the massive increases in subsidization and investment from elsewhere in China, with the result that the bar is constantly being raised for the more skilled strata of labor. This becomes the basis for exclusionary processes at the upper end of the labor hierarchy. While rising inequality might provide an important clue of these processes, it does not identify them or describe them given that exclusions in upper labor strata do not show up at the bottom end of the regional income distribution. Hence, notions of relative deprivation are very ill suited to analyze such types of exclusion, unless we tautologically define exclusion itself as deprivation. Similarly, in relation to contestation and conflict, many of those facing exclusionary pressures and who express the most discontent are not necessarily among the poorest. Rather, many appear to be from what could be considered the middle or upper strata of Tibetan society, such as rich farmers and herders who had sent their children to high school and university and who are frustrated that public employment is no longer guaranteed by the local government, or else graduates who prefer and have the family resources to remain unemployed rather than degrade themselves in jobs that they consider beneath them, such as in trade or commerce. Similarly, some reports from resettlements of herders in Qinghai have suggested that those facing the most loss, and hence who have shown the most resistance and/or grievance, are those who were relatively wealthy prior to resettlement, in terms of having sufficient herd sizes to maintain



Conclusion: From Polarization to Protest in Contemporary Tibet 375

subsistence and also afford bouts of conspicuous consumption. Poorer herders, in contrast, have been generally more at ease with leaving their herds, which were insufficient for subsistence in any case, and moving to welfare dependence and low-skilled employment in the new settlements.1 It is the relatively wealthy—with the leisure that wealth can afford—who can spend hours sitting in tea houses engaged in discussions over nationalist politics, not the average toiling farmer or herder disgruntled over stagnant or falling barley or wool prices and seeking out low-wage offfarm employment options to make ends meet. Indeed, as discussed in the previous chapter, some of the leading “hooligans” involved in fights with Muslim restaurant owners in Golok were the sons of county or prefecture officials and thus were well endowed with both impunity and leisure. And yet, when we wish to suggest that rising inequality might induce conflict, the implication is that conflict is driven by the relatively poor, that is, by those who bear the burden of inequality and thus have the greatest cause for grievance. The dissonance in the logic is evident. Given that we can observe similar dissonance between the representations and realities of other conflicts around the world,2 the Tibetan case possibly offers some more general insights into this issue.

Marginalizing Out of Poverty: Lessons from the “Tibet Model” Fundamentally, our conceptions of marginalization, exclusion, and disempowerment need to be extracted from metrics of poverty and even inequality, insofar as these metrics only end out reflecting absolute or relative poverty. As noted above, when inequality is rising, this might aptly reflect certain aspects of marginalization, albeit mostly those at the lower strata of a social and economic hierarchy. However, marginalization in effect needs to be understood as potentially occurring throughout social hierarchies, from all positions, relative to the social strata from which one is being marginalized and/or into which one aspires to enter. This is not to suggest that inequality is unimportant, but that its importance needs to be conceived in ways that conventional inequality metrics are ill suited to address, in the sense that rising inequality can be seen to produce exclusions, dislocations, and important social responses across social hierarchies, not just at the bottom. To take a simple example, when income inequality rises, there is effectively a thinning out of the income distribution, in the sense that there are fewer people in the middle of the distribution and more on the bottom and/or top of the distribution (usually bottom). Thinning occurs irrespective of growth, that is, it takes place even in a context of (disequalizing)

376

Chapter 9

growth given that it is relative. Hence, in the process of thinning, middle strata potentially face the greatest relative downward displacements. Lower strata will experience more competitive pressures due to the downward displacements of those from above, or reduced opportunities for upward mobility, and this will induce considerable churning within the lower strata.3 However, many poor might be insulated from churning at higher levels, such as rural dwellers with little or no integration into urban labor markets. Those in the lower strata who do experience exclusion will more likely experience it as obstruction of upward mobility, while their relative position within a social hierarchy might remain unchanged, insofar as they remain in farming or herding, for instance. Rather, the greatest insecurity in terms of loss of relative position (rather than poverty) is usually faced by various middle strata, such as those competing within urban areas or within growth-inducing sectors of an economy. It is in this sense that we often see intensifying exclusionary pressures among middle classes in contexts of unequal growth. Accordingly, disequalizing growth fuels impulses for social protection across a social hierarchy (per Polanyi’s double movement) precisely because dislocation, expressed as the perception or experience of a loss of relative position, is exacerbated across the social hierarchy by rising inequality. Given that relative dislocation potentially occurs most strongly at the middle strata, the impulse would similarly occur most strongly at these strata, particularly because people in these strata would also possess the resources to be able to resist. Indeed, this resonates with the classic observation made by both Karl Polanyi (1944) and Hannah Arendt (1951) that economic insecurity among the middle classes in Germany was a crucial factor in generating strong social support for fascism in the 1930s (which was one of the manifestations of double movement that Polanyi identified). These insights are all the more important within a transitional and disempowered peripheral context, in which locals are progressively marginalized from the changing sources of value in the local economy and from the main levers of power controlling decision making with respect to these changes. In these settings, inequality combined with economic and political subordination tends to reinforce the already entrenched tributary role that peripheries play within capitalist systems, accentuating the net outflows of financial, human, and physical resources from these regions even in the presence of subsidized inflows, as in the case of the Tibetan areas. In the event that such peripheral economies grow through external stimuli (such as subsidies), growth essentially reinforces the external dependence of the dominant processes of wealth circulation and accumulation in these economies, such as those located in government administration, largescale infrastructure investments, or tourism in the case of the TAR, while simultaneously marginalizing local populations in the event that they do



Conclusion: From Polarization to Protest in Contemporary Tibet 377

not comply with the strategized needs of such processes. The inefficiencies associated with such forms of growth are not necessarily due to any entrepreneurial disability of locals. Rather, they are mostly related to the incentive structures that guide wealth circulation, particularly among elites who have privileged access to this circulation. The polarizing tendencies of this model in the case of Tibet have tended to exacerbate the dislocating and subordinating pressures faced by Tibetans occupying the middle and upper strata of the local labor hierarchy. While only a minor portion of subsidies and subsidized investments have trickled down into Tibetan rural areas, this portion has nonetheless been large enough, given the overall deluge of centrally channeled subsidies, to stimulate a rapid degree of socio-economic transformation and (absolute) poverty reduction in these rural areas. However, the manner by which subsidies have been channeled creates the conditions for important ethnically discriminatory institutional biases in most aspects of the economy outside the rural areas. First, because subsidies and the policies guiding subsidies are instituted by members of the dominant Han Chinese nationality, they structure advantage along axes that include important Han Chinese attributes, such as Chinese fluency, Chinese work cultures, and connections to government or business networks in China proper. Indeed, these attributes are some of the important cultural dimensions through which schooling confers advantage within a setting of disempowered development and/ or cultural hegemony, in addition to its role in generating the appropriate skills required by changing structures of employment. Second, the increasing reliance on external forms of finance and labor in the upper layers of the economy (i.e., those that are predominantly driving growth) compels a demand for rationalization and standardization within the bureaucratized channels of wealth circulation. This demand effectively discriminates against Tibetans and other minority groups given that rationalization and standardization are guided by Han Chinese norms, thereby reinforcing the modes of linguistic and cultural disadvantage in public employment, contracting, and other aspects related to the administered and subsidydependent parts of the local economy. Finally, the increasing predominance of subsidies in the local economy and the outflows that they facilitate, such as massive trade deficits with the rest of China, have increasingly decentered the institutional incentive structures for local elites, including both Tibetans and Han Chinese. Incentive structures increasingly encourage such elites to position themselves with respect to the externalized wealth flows rather than to locally oriented forms of accumulation given that, among other considerations, much more money can be made with much greater ease through privileged access to state patronage and para-state networks than through painstaking and relatively petty profit-making activities based in local productive activities. As a

378

Chapter 9

result, there is little dispersion of the externalized flows of wealth into the local economy, and outflows of wealth are privileged at the expense of local reinvestment, which further biases the system toward a dependence on externally supplied managerial and technical labor. This further intensifies the demand for standardization and rationalization even while reinforcing the structural disjunctures and institutional disincentives that rationalization was supposed to address in the first place. In this context, elite or upwardly mobile Tibetans have been increasingly left with a convoluted choice between disadvantaged assimilation, which in many cases also implies a perceived downward displacement in the local social hierarchy, or else an erstwhile form of segregation that has proven increasingly difficult to maintain. This convoluted choice for relatively elite or upwardly mobile Tibetans has been especially manifest in the changing nexus of minority education and public employment systems in the Tibetan areas, even though both systems were originally devised to offer preferential treatment to such minority elites. On one hand, Tibetans have been disadvantaged in the mainstream schooling system due to linguistic or cultural biases, or even economically disadvantaged due to the higher costs of such education compared to the minority system. On the other hand, the segregated option of minority education has ultimately led to later disadvantages in urban employment in the context of the employment reforms, given that the reforms have curtailed guaranteed or preferential employment in favor of more meritocratic principles (defined according to the dominant group). In turn, this places greater emphasis on the qualitative aspects of education demanded by urban employers, such as language ability or status of degree. As a result minority education has been devalued by the dominant (Han Chinese) group that most controls the urban employment opportunities that are accordant with a tertiary level of education. Increased in-migration accentuates the imbroglio by short-circuiting the possibility for local educational development to meet the emerging labor market demands or else for local political action to adapt these demands to local capabilities in Tibetan areas. Hence, improved schooling outcomes at the primary and junior secondary levels have done little to improve this institutional disjuncture between education and employment systems; if anything, they exacerbate it by increasing the supply and aspirations of young Tibetans entering into senior secondary and tertiary education. Indeed, the end of preferential employment practices in the Tibetan areas risks accelerating the redundancy of the minority schooling system precisely at a time when education policy has been pushing for expanded enrollments within this same minority system. The impasse highlights the very valid need to design educational strategies in conjunction with urban employment strategies, along with the need for effective political representation and decision making by minor-



Conclusion: From Polarization to Protest in Contemporary Tibet 379

ity groups within these designs. Indeed, the perceived dead-end scenario reinforced the belief among many quarters of both common and elite Tibetans whom I interviewed that the only solution is found in nationalist (although not necessarily secessionist) politics, that is, increased local control over education and employment in order to create the conditions whereby Tibetan-prioritized education can be supported by Tibetan-prioritized employment. This is the logical and positive political response to the impasse. Less positive responses include impulsive conflicts between Tibetans and Muslims. Disequalizing and disempowered growth in Tibet has thereby set off exclusionary dynamics across the social hierarchy through these various institutional mechanisms such that conflict has been aggravated within these sustained bursts of subsidized prosperity. Minority nationalism and other discourses of contestation evolve out of an intuitive understanding that experiences of exclusion are rooted in disempowerment. For instance, local perceptions of population swamping, while probably misconceived, can be seen as a reactive lens through which locals interpret their concerns with urban employment exclusions within a context of stark political disempowerment. In particular, the view voices a legitimate concern that in-migration might exacerbate economic exclusion among Tibetan locals in their towns and cities, particularly at a crucial stage of their population and economic transitions when such exclusions no doubt heavily influence the long-term trajectories of ethnic stratification. Hence, the symbiotic conflation of population swamping and exclusion seems to provide the discursive superstructure through which Tibetans understand their contemporary impasse within disempowered development. Similarly, conflicts with Muslims and religious cooperation with Han Chinese poignantly demonstrate how the patterns of exclusion and the possible avenues for alternative forms of inclusion have conditioned the impulse for social protection in ways that contradict the ideological assumption implicit or explicit in state development policies that economic growth and prosperity will necessarily reduce the manifold apparitions of ethnic nationalism and religiosity. The political dimension of double movement is to be emphasized in these examples: rather than rejecting integration into China, elites have aimed at redefining the modes of integration in a manner that preserves an indigenous sense of hierarchy and dignity, at least on terms defined by the dominant sources of authority within Tibetan society. In this sense, our understanding of current tensions in Tibet needs to be grounded not only in the historical context of political domination and subordination but also in the heightened degree of economic and social polarization generated by the latest phase of intense subsidization and rapid growth since the mid-1990s. This argument is an important distinction from the contention that ethnic inequalities in the TAR and

380

Chapter 9

other Tibetan areas are the result of spatial inequalities, as argued by Sautman (2008) and Hu and Salazar (2008). Rather, inequalities differ from elsewhere in China in that they are instituted and structured through the perverse special treatment accorded to the Tibetan areas by the PRC, which results in a progressive appropriation of ownership in the local economy by outsiders, placing local Tibetans at a disadvantage despite the profusion of subsidies. The degree to which discrimination is specifically or systematically intended within these strategies is perhaps less important than the fact that the result is effectively discriminatory. In any case, the strong perception of discrimination in the eyes of local Tibetans offers important insights into the explosion of tensions in March 2008 and beyond. Ultimately, local initiatives and locally generated investment and accumulation have played a very limited role in the overall processes of economic change in the Tibetan areas since growth accelerated in the 1990s. In the tense political environment, they might have even been discouraged. In the TAR in particular, where subsidization has reached its zenith, the local Tibetan population has been rendered more or less irrelevant as an agent causing growth, at least in terms of causes that are independent of subsidies. Meaningful de-centralization has simply not taken place in Tibet the way it has in most other regions of China during the reform period. This particular lack of agency exacerbates a feeling of alienation despite all the monumental change and pockets of affluence. In particular, additional state responses following the protests in 2008 have involved the intensification of pressure tactics directed toward the dominant sources of authority within Tibetan society, such as patriotic education campaigns in many monasteries. Such tactics have been accompanied by efforts, such as in Qinghai, to accelerate resettlements, consolidate primary schools (implying an almost obligatory boarding of rural primary school–aged children), and deepen the use of Chinese-medium in such schools. The confluence of these various factors possibly helps to explain why nationalist protests have taken the disturbing turns they have with the wave of self-immolations from 2011 onward.

Final Reflections Language, as a tool of both oppression and resistance, has stood out in this research as a central factor conditioning exclusion and marginalization, but not because of linguistic degeneration as implied by the idea of “cultural genocide.” Rather, despite enormous efforts and much success by Tibetans to preserve and modernize their linguistic traditions (including the idea of widespread literacy in Tibetan, which is an entirely modern affair), language plays a crucial role as a tool of political and economic domination



Conclusion: From Polarization to Protest in Contemporary Tibet 381

and subordination by the Han Chinese in Tibet, serving as a key mode of bias operating in most aspects of political and economic life outside the rural areas and small towns. From this understanding, I have generally advocated that two dimensions of current Tibet policy need to be seriously addressed in order to lessen both economic marginalization and social instability. On one hand, urban employment for disadvantaged Tibetan locals requires protection and promotion at both lower and upper strata through preferential policies. On the other hand, the linguistic and cultural disadvantages faced by Tibetans in higher levels of education and urban employment need to be lessened, albeit in ways that do not undermine Tibetan initiatives to develop their language and culture, and that are not predicated on eliminating differences with the Chinese altogether through assimilation or “acculturation.” The importance (and realities) of Tibetanization have been amply explored by other scholars, such as with respect to bilingual education or else in response to the arguments of Ma (2007) and other scholars that the eradication of the preferential minzu system is necessary in order to deal with the politicization of minorities in China, as noted in the introduction and chapter 5.4 My closing argument here is to support the case for Tibetanization from a socio-economic angle, particularly with respect to integration into urban employment during a decisive stage of labor transition in Tibet. To a certain degree, the existing national minority laws in China already provide ways to resolve both dimensions of current Tibet policy if their full implications are put into practice. These implications are exemplified by the recommendation by the Tenth Panchen Lama in the 1980s that publicsector employees working in Tibetan areas should have at least some working knowledge of the Tibetan language, supported by a strong promotion of bilingual education (i.e., Tibetan medium in all subjects) in these areas. Similar approaches could be taken to public contracting as a means to apply preferentiality at middle and lower levels of the urban hierarchy, such as in publicly contracted construction work targeted to support the development of locally (Tibetan) owned and run businesses. Indeed, the potential for preferentiality in contracting is particularly great given the weight of statefunded investment in the Tibetan areas. Other applications of the minority nationality laws allow for policies such as minority language requirements on street signage (which are already in implementation, although the Tibetan usually ends out being symbolic) or even construction codes that new buildings must be built in ethnic minority style. These implications do not require re-imposition of past systems of guaranteed employment; rather, they call for the maintenance, expansion, and/or innovation of practices of preferentiality within local governance and within the existing reformed system of competitive public employment. In other words, there is substantial “wiggle room” to explore policies that could allow for the

382

Chapter 9

coordination of education and employment strategies in ways that would partially correct the disadvantages faced by Tibetans, or even give a strong competitive advantage to Tibetans over non-Tibetan migrants, such as with the requirement of working knowledge of Tibetan for public employment in Tibetan areas. Such strategies would also bolster support for a Tibetanmedium education system, and the positive synergies could go a long way to address many of the underlying grievances driving the current protests, contrary to common government suspicions that they would in fact fuel the rise in ethnic nationalism. Moreover, the approach of applying linguistic requirements in public employment and preferentiality in local contracting avoids the political quagmires often associated with the more typical approaches to preferentiality or affirmative action, such as quotas and reservations (as are commonly and controversially used in countries such as India, the United States, and elsewhere) or else the system of lower entry requirements for university as practiced in China.5 Rather, it creates preferentiality through changing the institutional norms of competition within public employment (or public contracting), according to terms that are quite justifiable and legitimate, and in conformity with the formally established laws. It is also justifiable and legitimate from a local governance perspective, in the sense that government officials should be able to communicate with the people they serve. In practice, much would obviously need to change for the application of such an approach to become effective, as analyzed for instance by Zenz (2013) with respect to corruption and informalization at the local (county) level of government in the Tibetan areas of Qinghai. Nonetheless, a movement in this direction would help to tilt the scales of advantage toward Tibetans within the reformed system of competitive employment, at least for educated and relatively elite Tibetans and within a broad regional scale. Preferentiality and Elites On this last point, preferential policy approaches are undoubtedly oriented toward a minority of Tibetans, that is, those educated up to a tertiary level and who are relatively elite in this and other respects. Indeed, this criticism of affirmative action policies is often made in India, with reference to the fact that university and public employment reservations benefit mostly a “creamy layer” of the disadvantaged groups targeted by such policies, thereby resulting in entrenched within-group inequalities. However, despite this contention, the importance of cultivating inclusion at the middle and upper strata of the labor hierarchy should not be underestimated. As I have discussed in general terms in Fischer (2011a) without reference to Tibet, there are several important reasons for paying attention to



Conclusion: From Polarization to Protest in Contemporary Tibet 383

exclusions occurring at the middle strata of a social hierarchy. One is that exclusions at these levels can lead to longer-term stratifying and potentially impoverishing life trajectories without any obvious short-term poverty or inequality outcomes. Such considerations are especially important in contexts of structural change such as urbanization and migration, rising education levels, or changing livelihood patterns, during which the exact distributional implications of such changes might not be obvious, but where powerful stratifying and subordinating social processes might nonetheless be at work. The second is that exclusions among the non-poor have an important bearing on obstacles to upward mobility faced by the poor attempting to enter these non-poor social strata. This is especially important in situations where poverty reduction strategies are predicated on upward mobility, such as education or entrepreneurship, versus improving the terms of labor. Third, as argued in the two previous chapters, inequalityinduced conflict is arguably driven by the non-poor much more than by the poor. This perspective is important because it implicates middle classes and elites into an understanding of how inequality might induce conflict, as discussed above, and helps to correct the common tendency in much of the literature to implicitly blame inequality-induced conflict on the poor. Moreover, exclusion among the relatively elite can have important symbolic influences among the non-elite and population at large. This point is made by Jackson (1996) with reference to her critique of the idea of a “feminization of poverty.” She argues that rescuing gender from poverty analysis is necessary because the subordination of women is not caused by poverty and thus anti-poverty policies are not necessarily appropriate to tackle gender issues. Moreover, she suggests that the position of non-poor women is relevant to poor women through their role in influencing gender bargaining, ideologies, and opportunities for poor women, or through changing societal perceptions of women’s options. This effect can be both positive and negative. For instance, Jackson offers the example of “sankritization” in India, which has carried with it particular gender ideologies centered on the dependency of women, or else dowry practices among middle income groups, both of which have had regressive influences on gender issues in lower income groups. Thus, a “poverty focus misses the range of interconnected gender issues across classes and socioeconomic strata” (501). Reframing this important point to apply to the issue of minority subordination within the Tibetan context, an exclusion that does not necessarily lead to any particular deprivation may still have a very powerful effect on various social processes of disadvantage and grievance. This is especially the case if exclusions happen at the upper end of a social hierarchy, precisely because of their powerful symbolic effect on public perceptions within the disadvantaged group. To borrow the terminology of Stewart (2002), preferentiality or affirmative actions might not be well suited to address poverty or “vertical” inequalities

384

Chapter 9

(i.e., inequalities between individuals and households), although if used skillfully, they can be effective in addressing “horizontal” inequalities (i.e., inequalities between groups). The criticism that such policies tend to be elitist within disadvantaged groups arguably misses the point because their purpose should not be conceived as one of poverty or (vertical) inequality reduction. Rather, their purpose, for which they need to be judged, is to increase the representation of disadvantaged groups in positions of social, political, and economic influence, and these positions are generally only accessed by elites of all groups in any case. The potency of increasing effective representation at these levels cannot and arguably should not be judged in terms of poverty or inequality reduction, but in terms of other, more strategic political economy dynamics of group empowerment. In this respect, an increase in preferentially supplied public employment would probably not come close to matching the increasing supply of tertiary-educated Tibetans from across the region (particularly from Qinghai). However, there is little doubt that a lot more public employment could be made available to locals if greater preferentiality were applied, enough to make a substantial dent in the exclusionary trends and to support more positive public perceptions among Tibetans regarding the possibilities for upward mobility through the combined education and employment systems. For instance, as analyzed in chapter 5, staff and worker employment in state-owned units in the TAR increased from 136,646 people in 2003 (which appears to be the trough after a round of austerity in public employment in the early 2000s) to 170,748 in 2008, and by 2010 there were 207,267 employees in state-owned urban units (out of a total of 222,116 employees in all urban units). Even though this represents a rapidly declining share of staff and worker employment (or employment in urban units) in total formally registered urban employment in the TAR, from 72 percent of the total in 2000 to 42 percent in 2010, the nominal increase of about 70,000 state-sector employees is nonetheless considerable with respect to the supply of tertiary educated Tibetans over this same period. As discussed in chapter 6, about 165,000 people in the TAR would have had a tertiary level of education in 2010 according to the latest census, although this included non-Tibetan migrants. According to the 2009 population survey, which included only permanent residents and thus would have had a much higher representation of Tibetans, only about 50,000 permanent residents in the TAR had a tertiary level of education. From this perspective, the scope for practicing preferentiality within increasing state-sector employment is substantial, particularly considering that such employment accounts for almost the totality of formal “urban unit” employment in the TAR. Even if a fraction of the increase in state-sector employment were reserved for Tibetans, this would be enough to make a major impact both on the realities and perceptions of inclusion among more-educated Tibetans, as well



Conclusion: From Polarization to Protest in Contemporary Tibet 385

as on the perceptions of the value of schooling among the population in general (and hence the worthiness of pursing such schooling). Moreover, as I already argued in the conclusion of Fischer (2005a), the government need not worry about the inefficiencies associated with hiring such less experienced, able, or fluent employees (that is, experienced, able, and fluent in the ways of China proper) given that, in macroeconomic terms, its subsidization strategies are extremely inefficient in any case (at least, from the local perspective of the TAR). The role of preferentiality in this sense must be understood as supporting the dynamic synergies between Tibetanization in the schooling system and inclusion into the parts of the employment system that conform to rising levels of schooling. This is particularly important at the current juncture when the cohorts of students impacted by the education drives in the late 1990s and 2000s aimed at increasing enrollments at the primary and junior secondary levels are increasingly starting to enter and exit senior secondary and tertiary levels of schooling over the course of the next ten years or so. At the very least, there will be a continued need for the expansion of the Tibetan schooling system over the next ten years or so, and the labor needs of such an expansion would be ideally filled by Tibetan graduates.6 The current contradictions within the education-employment nexus and the trends of current education policy making might prevent an effective expansion in this sense, thereby exacerbating the exclusionary trends as analyzed in this book. However, it is vital to try to influence a change in this direction. The role of preferential policies would need to play a central role within such a change, in combination with a variety of other initiatives, such as reformed and improved financing for lower (i.e., county) levels of government. Preferentiality and Bilingualism in the Education-Employment Nexus of Tibet As is perfunctorily noted in most official government pronouncements on the matter, China already practices various forms of preferentiality for minorities. Currently, these are mostly practiced in the form of easier entry requirements for university, whereas most other preferential policies have been substantially watered down since the late 1980s, particularly with respect to public employment and contracting. The political trend has been against reviving and strengthening such policies, even at the level of public opinion in China. Many Chinese remain convinced that the government has already delivered mostly benefit to Tibetans through spending enormous amounts of aid in the Tibetan areas and that Tibetans are essentially complaining with their mouths full or asking for special privileges above and beyond what other minorities receive, let alone the majority Han Chi-

386

Chapter 9

nese. Thus, while the legal principle might exist, the trend of political sentiments might be against such forms of positive discrimination. Moreover, recent education reforms in Tibetan areas appear to be following this political trend. For instance, there has been pressure in Qinghai, where Tibetan-medium schooling is probably the most developed out of all the Tibetan areas, to increase Chinese-medium in the Tibetan areas of this province. This pressure started with a primary school consolidation policy in parts of Qinghai in 2009 (and possibly before this time in Sichuan) and was complemented with further schooling reforms in 2010 that explicitly aimed to increase the use of Chinese language in Tibetan primary schools.7 While the consolidation reforms have been happening throughout China, they have implied the boarding of primary school-aged children in Tibetan areas given the topographic and demographic characteristics of these areas, whereas boarding in the more densely populated areas of most of China proper would not necessarily be required for primary school consolidation. Moreover, these schooling reforms are being implemented according to a uniform national logic despite the fact that these Tibetan areas have not yet caught up in their massive educational lag with other parts of China, including closely neighboring regions and sources of migration into these Tibetan areas (such as the Han Chinese parts of Sichuan). In this manner, the new primary schooling model in the Tibetan areas risks coming to quickly resemble some of the worst past practices of, say, Canada, in terms of the experiences of the residential schools and their devastating impact on several generations of aboriginal children and communities in Canada. However, despite these lessons, there is a strong possibility in the zeal of mega-development planning currently enveloping Tibet that, as during the Cultural Revolution in China proper, these policies will severely debilitate a generation of young students now being forced through these schools, often in complete violation of parental rights and in a manner that Han Chinese people would probably never accept. The reforms in the Tibetan areas are motivated by an effort to overcome some of the apparent obstacles to improving rural education in these areas and also to enhance the integration of young Tibetans into mainstream Han society. However, the reforms are possibly one of the more tragic aspects of current policy shifts in Tibetan areas, particularly in terms of the separation of young children from their families for long periods of time, in conditions that might be described as factory schools with sub-standard care. While the care and quality of teaching in the consolidated schools might well be better than the very poor standards in many rural schools, it is questionable whether this compensates for the far more important loss of daily parental care and knowledge transmission, as well as the loss of parents’ ability to participate in and monitor such schooling. This intimate aspect of radical dislocation and alienation should not be underestimated as a background



Conclusion: From Polarization to Protest in Contemporary Tibet 387

to current unrest and protest in much of these Tibetan areas, particularly given that some of the Tibetan self-immolators in 2011 and 2012 explicitly made a connection between their act and the new education policies. The policy is certainly not necessary, even if the government wishes to otherwise spend its money on a variety of large-scale high-tech projects. It is also not dictated by an imagined abstract principle of the demands of the so-called Socialist market economy. Given the sheer scale and inefficiency of the subsidy deluge in the TAR, for instance, the government could well afford any variety of other schooling models, even if these proved to be more costly. Indeed, they should prove more costly given that one fundamental problem with rural schooling has been chronic under-funding. Nonetheless, or precisely because of this context, some have argued that an agenda of enhancing the practice of positive discrimination via the use of linguistic requirements in public employment is unrealistic in the current context and given the assimilationist (or acculturationist) agenda of the central and provincial governments. Rather, a common argument is that the best way to address the educational disadvantages faced by Tibetans and their consequent marginalization from urban employment opportunities is through improving their fluency in Chinese and their competency in Chinese-medium education, so as to improve their ability to integrate and compete in a regional economy and society in which Chinese is effectively the hegemonic lingua franca. Along these lines, many Chinese (and some foreign) scholars rightly recognize the linguistic biases and disadvantages faced by Tibetans but then conceive of the solution in terms of stronger integration, especially by way of better Chinese-medium education, perhaps starting as early as primary school, so as to prepare young Tibetans for the de facto domination of the Chinese language in the wider context into which they must integrate. The option of strengthening the use of Tibetan in Tibetan areas, in the education system and also in employment practices, is rarely considered by such scholars as even remotely realistic. Advocates of Tibetan-medium education (or mother-tongue medium) would retort that the best way of assuring linguistic competency in Chinese would be through Tibetan-medium, that is, teaching Chinese as a second language to unilingual Tibetan students. Similarly, the best of way of achieving competency in other subjects would be through Tibetan medium so that unilingual Tibetan students would not have to struggle with trying to learn Chinese while as trying to learn another subject, such as mathematics. Despite these well-accepted insights from national and international education specialists, proponents of the so-called realist position would reply that Tibetan-medium might be a reasonable strategy up to the primary or perhaps junior secondary levels of schooling, but not above, if only because of a lack of capacity to conduct comprehensive and relatively good-quality senior secondary level instruction in Tibetan

388

Chapter 9

medium in all the required subjects and across all the Tibetan areas. Moreover, Tibetan, medium even at lower levels of schooling usually results in Tibetan students who require considerable linguistic catch-up (in Chinese language competency) once they enter the senior secondary level, and who are subsequently disadvantaged if and when they enter the inevitably Chinese-dominated tertiary levels of education or else in urban labor markets after graduation. Admittedly, the evidence of this book could be used to support such an argument, as I have come to increasingly realize over the years of presenting this research in various venues, particularly in China. For instance, my emphasis of the severe educational lags in literacy and levels of schooling faced by Tibetans in comparison to other regions of China and their resulting disadvantages in labor markets has been generally well received by audiences of Chinese scholars and policy makers to whom I have presented in China or abroad on various occasions. In particular, the idea of a skills disjuncture between the employment demanded by the structural and institutional patterns of economic growth versus the local supplies of labor in Tibet supports the common understanding among Chinese scholars and policy makers that the scarcity of qualified labor in Tibet is a major impediment to carrying out development strategies there. Indeed, this understanding is a core rationale for why the government encourages migration from other parts of China, even if only as a temporary fix and with no malicious intentions regarding assimilation. In this sense, my own arguments could be seen to reinforce the common view in China that the problem and the solution of Tibet is fundamentally one of education. However, the more important question is: what kind of education? The idea that education is at the root of development problems in Tibet fits into the Marxist modernization theory mind-set of many Chinese scholars, as epitomized by Wang and Bai (1991 [1986]). As discussed in chapters 4 and 7, Wang and Bai regarded the backwardness of Tibet and other minority areas to be based on a poor quality of human resources. The idea of “quality” in this sense bears more relation to the culturalist meaning implied by the Chinese word suzhi than to an objective mastery of technical subjects or proficiencies. For them, poverty and backwardness could be ultimately resolved only through improving such quality, which would amount to an attitudinal shift, “reversing the attitude of the local inhabitants towards social wealth and changing their traditional ways of exploiting natural resources” (92). While seemingly progressive at the time, insofar as they argued against fast-track assimilation or simply treating the Tibetan (and other minority) areas as if they were Han areas, the underlying implication of their thesis was not only assimilationist (in a gradual sense), but also bordered on very derogatory forms of prejudice, even while being based on observations that might have some empirical



Conclusion: From Polarization to Protest in Contemporary Tibet 389

validity (such as the subsistence basis of rural livelihoods in the regions they visited in the early 1980s). Hence, I have come to be very cautious in the way that I advocate the idea of education as an important part of resolving exclusionary dynamics in Tibet, particularly in the current context where enrollment rates have been increasing substantially, yet within an increasingly assimilationist environment. More subaltern-sensitive perspectives might nonetheless emphasize the need for Chinese-medium education to address the problems of urban employment exclusion in Tibetan areas as a matter of practical necessity of modern life within China, even if this reality is seen as deplorable to a certain degree. For instance, Ben Hillman (2008) argues that “[d]espite accusations to the contrary, the Chinese government has made increasing efforts to sponsor bilingual education, but this too is a double-edged sword” because many Tibetans subsequently never catch up in terms of a sufficient level of Chinese literacy to be able to compete with other Chinese students in higher-level entrance exams. In what might have been an implied criticism of my advocacy for a greater application of linguistic requirements for public employment in minority areas,8 and despite otherwise repeating many of the empirical points and arguments that I had previously made in (Fischer 2005a), he noted that “levels of Tibetan literacy among Tibetans can be as low as, or lower than, levels of Chinese [illiteracy],” implying that a policy of insisting on Tibetan literacy for public employment could disadvantage Tibetans as much as it would disadvantage non-Tibetans. He also asserted that there is “no systematic discrimination of Tibetans by employers—in fact Tibetans are accorded preferential treatment in state jobs.” It is not clear on what basis he makes this claim beyond several anecdotal examples and despite ample evidence to the contrary, as discussed in chapter 6. Rather, he argued that employment exclusion is simply a matter of labor market principles in which the most skilled people are getting the jobs regardless of ethnicity, and hence Tibetans are not getting the jobs because employers cannot find Tibetans with skills that meet their needs. Hence, the problem is one of matching labor supply and demand, not discrimination. As noted above, this argument echoes my own regarding the skills disjuncture that has emerged with rapid economic growth. However, it is somewhat simplistic in its analysis of discrimination, particularly with regard to the endogeneity of the skills disjuncture to the instituted structures of economic growth, which in turn might be understood as discriminatory in a broader sense. Moreover, much like in the work of Sautman (2008), the defense of the government’s existing preferential policies ignores a serious undermining of these policies since the late 1990s. Alluding to poor levels of Tibetan literacy as a means to undermine the case for linguistic requirements also conflates the general conditions of illiteracy and low

390

Chapter 9

schooling levels among the Tibetan population with the intended targets of preferential policies, who are the 5 to 10 percent of Tibetans who have achieved a senior secondary or tertiary level of education and hence are not obstructed by illiteracy, particularly in Amdo, where many such graduates have a competent level of Tibetan literacy. Hillman’s argument in this sense appears to be influenced by the very particular conditions prevailing in Dechen (Ch. Diqing) Prefecture in Yunnan, where he apparently did most of his fieldwork and about which he wrote most of his Tibet-focused academic work. Notably, there has been no official bilingual education in Dechen and it is arguably the most sinicized of Tibetan areas. The more general problem with this line of argument is that it serves as a Trojan horse for sympathetically undermining the case for Tibetan-medium education at the senior secondary and tertiary levels, and for the importance of supporting such education with preferential employment strategies adapted to the graduates from such education. It also typically ignores the strong cases of Tibetan-medium education, such as in Qinghai, where Tibetan medium has expanded vigorously at secondary and tertiary levels, in many cases with much success, as analyzed in depth by Zenz (2013). The education-employment nexus in this sense is a package deal: the two parts cannot be separated from each other. Undermining the rationale for bilingual education coordinated with stronger forms of preferentiality in employment is effectively a capitulation to an assimilationist agenda, whether this be explicit or implicit in the argumentation. Moreover, assimilation (or acculturation) simply does not solve marginalization. In the absence of any significant shifts in the structural and institutional bases of power relations between dominant and subordinated groups, it usually merely transmutes subordination into a variety of new trajectories, as evidenced for instance by minorities in the United States or Europe,9 or increasingly by the experience of assimilated or acculturated Tibetans in the urban areas of their region. That such transmutation might be reflected as increasing living standards is precisely why we need to separate our understanding of subordination and marginalization from the metrics of poverty and inequality, particularly when the former are occurring among middle and upper strata of the minority labor force. Admittedly, the alternative strategy proposed here is perhaps an increasingly elusive possibility given current trends in Tibet policy, as noted above. It is also predicated on the faith that there is a sincere intent on the side of Tibet policy makers in Beijing and provincial capitals to devise forms of “harmonious” development that are non-marginalizing for Tibetans and hence less prone to induce unrest. As some exile Tibetans often point out to me, this might not be the case. It might be the case, as the most pessimistic fear, that current Tibet policy is guided in a more Machiavellian manner, intent precisely on pushing assimilation regardless of the short- or



Conclusion: From Polarization to Protest in Contemporary Tibet 391

medium-term unrest or the marginalization that this might generate, as an ultimate long-term “total solution” to the Tibet problem. However, I leave these more political questions to others more suited to the task. From my own experience, I believe that this is not the case, even though a genuine practice of minority autonomy, founded on Tibetan political representation and coordinated Tibetan control over minority education and local employment, appears to be an increasingly unlikely and elusive prospect in the context of current political dynamics in China, both national and local. In this regard, the linguistic divide and the associated gulf in worldviews for the large part clarify the fundamental tension in China’s Tibet policy between autonomy and assimilation. As noted in the preface, there is even a rift in the understanding of “bilingual education.” Tibetans usually use the term to refer to strengthening the use of Tibetan-medium in schooling, ideally across all subjects, so that Tibetans can be guaranteed full competence in their own language as the basis for learning other subjects and languages. In contrast, Chinese people usually (although not always—I have met some notable exceptions) use the term to refer to bringing in more Chinesemedium earlier in the minority education system in order to improve the Chinese language skills of Tibetan students, even if this means further eroding the quality or the coverage of the Tibetan-medium or even the Tibetan literacy of the students (which is generally quite poor in the current system, as noted by Hillman, unless students specifically specialize in linguistic or cultural subjects). Even if there is an appreciation for the heritage of the Tibetan language, this latter option is usually advocated as a modern and inevitable necessity to help Tibetans to function and perform in the cultural mainstream of China, much as all US citizens must seek fluency in English as an implicit credential of their (functional) citizenship. The problem, of course, is that this common Chinese perspective simply does not grapple with the indigeneity of Tibetans in Tibet and with the very real historical reality that, prior to 1950, most Tibetan areas and definitely the area now incorporated as the TAR were self-governed by Tibetans and in Tibetan language, regardless of the ebbs and flows of decaying empires and emerging nation-states. The sudden and radical shift in the 1950s, to rule by outsiders who did not speak Tibetan and understood little about the people they had subjugated, was forced and was quickly embedded within an ideological animosity toward multiculturalism that lasted until the 1980s. It is also wishful thinking that assimilation or acculturation—linguistic or otherwise—can be the solution to deeply entrenched systems of subordination, discrimination, and disadvantage. Rather, it usually involves the reproduction of subordination, discrimination, and disadvantage on a new assimilated footing, given that the former are more fundamentally rooted on asymmetrical power relations, vastly unequal in the case of Tibet. As my critics from the Tibetan exile community rightly point out, it will only be

392

Chapter 9

when these power relations are seriously and sincerely addressed that we will begin to be able to conceive of a more just path of development for Tibetans in Tibet. It is my hope that this book might encourage steps in that direction. In the meantime, development in Tibet continues to proceed in a disempowered manner, in the macro-policy sense used in this book. We can only hope that the marginalizing tendencies will be somehow mediated, whether through benevolence, oversight, or even neglect.

Notes   1.  As already noted in chapter 7, I am grateful to Jampel Dell’Angelo for letting me use this example from his MSc research at the London School of Economics.  2.  See Fischer (2008a; 2011a) for further discussion of this point. Also, see an interesting recent argument along these lines with reference to Pakistan in Blair et al. (2012).   3.  With respect to churning, Hills (1998) makes the important insight that high mobility is not necessarily contradictory with high levels of inequality.  4. As noted elsewhere, see Zenz (2013) for an excellent recent discussion of these debates. On bilingual education policy and practice, see an excellent gathering of expertise on this issue at the Tibet Governance and Practice Forum, “Language Policy and Practice in Tibet: Concept, Experience and Challenges for the Future,” Montreal, Quebec, Canada, 19-23 May 2012 (http://tgapforum.org/node/42). Also see the seminal work by Postiglione (e.g., 2008; 2009).  5. Indeed, approaches to affirmative action such as quotas and reservations are typical in highly segmented social policy systems, such as in the United States or India, in which the targeting of disadvantages in a comprehensive way at lower levels of the education system is rendered very difficult and fragmented, in contrast to more universalistic systems of provisioning that internalize such practices within an integrated and unified education system servicing most of the population (besides the very rich, who will always find ways to buy themselves out). At the very least, affirmative action policies must be evaluated from this broader institutionalist perspective. For further discussion, see Fischer (2012b).   6.  See a fairly precise quantitative analysis of this for the Tibetan areas of Qinghai in Zenz (2013).  7. See Cencetti (2012, chapter 10) for an excellent ethnography of these schooling policies in one Tibetan area in Qinghai. Also see some discussion of this in Zenz (2013).  8. Hillman (2008) published his article in the Far Eastern Economic Review only a few weeks after I had written one in the same venue following the March 2008 protests: see Andrew M. Fischer, “Reaping Tibet’s Whirlwind,” Far Eastern Economic Review, March 2008. http://www.feer.com/features/2008/march/chinastibet-problem   9.  Indeed, one of the bizarre weaknesses in the argument by Ma (2007) is his uncritical positive appraisal of the “culturisation” strategy toward minority groups in the United States.

Bibliography

Abu-Lughod, Janet. 1989. Before European Hegemony: The World System A.D. 12501350. New York: Oxford University Press. Adams, Vincanne. 1998. “Suffering the Winds of Lhasa.” Medical Anthropological Quarterly 12(1): 74-102. (ADB) Asian Development Bank. 2003. The 2020 Project: Policy Support in the People’s Republic of China, Final Report and Policy Directions. Manila: Asian Development Bank. Andersen, Anders Hojmark, Sarah Cooke, and Michael Wills. 1995. The New Majority. London: Tibet Support Group. (AP) Associated Press. 2012. “Chinese Website: Dalai Lama Has ‘Nazi’ Policies.” New York Times, March 24. Last accessed 29 March 2012. http://www.eastasiaforum .org/2012/03/21/tibet-s-suicidal-politics/ Arendt, Hannah. 1951. The Origins of Totalitarianism. London: Allen and Unwin. Arrighi, Giovanni. 1994. The Long Twentieth Century: Money, Power and the Origins of Our Times. London: Verso. Attané, Isabelle, and Youssef Courbage. 2000. “Transitional Stages and Identity Boundaries: The Case of Ethnic Minorities in China.” Population and Environment 21(3): 257-80. Bangsbo, Ellen. 2008. “Schooling for Knowledge and Cultural Survival: Rural Schools in the Tibetan Borderland.” Educational Review 60(1): 69–84. Barnett, Robert. 1998. “Preface.” In A Poisoned Arrow: The Secret Petition of the 10th Panchen Lama, edited by Robert Barnett. London: Tibet Information Network, xi-xxiii. ———. 2006. “Beyond the Collaborator-Martyr Model: Strategies of Compliance, Opportunism and Opposition within Tibet.” In The Tibet Question, edited by Barry Sautman and June Dreyer, 25-66. Armonk, NY: M. E. Sharpe. ———. 2009. “The Tibet Protests of Spring, 2008: Conflict between the Nation and the State.” China Perspectives 2009(3): 6-23. 393

394

Bibliography

———. 2011. “Viewpoint: Are Tibet Burnings Plot or Policy Failure?” BBC News (16 November). http://www.bbc.co.uk/news/world-asia-15738522. Bass, Catriona. 1998. Education in Tibet: Policy and Practice since 1950. London: TIN/ Zed Books. ———. 2005. “Learning to Love the Motherland: Educating Tibetans in China.” Journal of Moral Education 34(4): 433-49. Bauer, Ken. 2003. “State and Range: Pastoral Development Policies in the TAR since 1951.” Paper presented at the 10th International Conference of the International Association for Tibetan Studies, Oxford, September 2003. Bauer, Ken, and Yonten Nyima. 2011. “Laws and Regulations Impacting the Enclosure Movement on the Tibetan Plateau of China.” Himalaya 30(1 & 2): 23-37. Beckwith, Christopher. 1987. The Tibetan Empire in Central Asia: A History of the Struggle for Great Power among Tibetans, Turks, Arabs and Chinese during the Middle Ages. Princeton, NJ: Princeton University Press. Blair, Graeme, C. Christine Fair, Neil Malhotra, and Jacob N. Shapiro. 2012. “Poverty and Support for Militant Politics: Evidence from Pakistan.” American Journal of Political Science 57(1): 30-48. Blecher, Marc J. 1991. “Developmental State, Entrepreneurial State: The Political Economy of Socialist Reform in Xinji Municipality and Guanghan County.” In The Road to Crisis: The Chinese State in the Era of Economic Reform, edited by Gordon White, 265-91. London: Macmillan. Blecher, Marc J., and Vivienne Shue. 1996. Tethered Deer: Government and Economy in a Chinese County. Stanford, CA: Stanford University Press. Bloom, Gerald, and Jing Fang. 2003. “China’s Rural Health System in a Changing Institutional Context.” IDS Working Paper 194, Institute of Development Studies. Borjigidai, Uyunbiliq. 2002. “The Hoshuud Polity in Khökhnuur (Kokonor).” Inner Asia 4(2): 181-196. Brandt, Loren, Huang Jikun, Li Guo, and Scott Rozelle. 2002. “Land Right in Rural China: Facts, Fictions and Issues.” China Journal 47: 67–97. Brenner, Robert. 1976. “Agrarian Class Structure and Economic Development in Pre-Industrial Europe.” Past and Present 70 (February): 30-75. ———. 1977. “The Origins of Capitalist Development: A Critique of Neo-Smithian Marxism.” New Left Review 104: 27-92. Burjgin, Jirgal, and Naran Bilik. 2003. “Contemporary Mongolian Population Distribution, Migration, Cultural Change, and Identity.” In China’s Minorities on the Move: Selected Case Studies, edited by Robyn Iredale, Naran Bilik, and Fei Guo, 53-68. Armonk, NY: M. E. Sharpe. Cabezon, José Ignacio. 1997). “Islam in the Tibetan Cultural Sphere.” In Islam in Tibet and the Illustrated Narrative Tibetan Caravans, edited by Gray Henry, 13-34. Louisville, KY: Fons Vitae. Cannon, Terry. 2000. “Introduction: The Economic Reforms, Demographic Processes and Environmental Problems.” In China’s Economic Growth: The Impact on Regions, Migration, and the Environment, edited by Terry Cannon, 1-29. London: Macmillan. Cannon, Terry, and Alan Jenkins, eds. 1990. The Geography of Contemporary China: The Impact of Deng Xiaoping’s Decade. London: Routledge.



Bibliography 395

Cardoso, Fernando Henrique. 1977. “The Consumption of Dependency Theory in the United States.” Latin American Research Review 12(3): 7-25. Cencetti, Elisa. 2012. “Des bergers sans troupeau: Les nouveaux nomads de l’économie socialiste de marché en Amdo-Qinghai (RPC).” PhD diss., École des Hautes Études en Sciences Sociales, Paris. Chang, Ha-Joon. 2002. Kicking Away the Ladder: Development Strategy in Historical Perspective. London: Anthem Press. Cheng, Joseph Y. S. 2003. “Qinghai’s Economic Development Strategy.” Issues and Studies 39(2): 189–218. Childs, Geoff. 2000. “The 1958 sKyid grong Census: Implications for the Study of Tibetan Historical Demography.” Tibet Journal 25(2): 29–41. ———. 2001a. “A Brief History of Nub-ri: Ethnic Interface, Sacred Geography, and Historical Migrations in a Himalayan Locality.” Zentralasiatische Studien 31: 7–29. ———. 2001b. “Old-Age Security, Religious Celibacy, and Aggregate Fertility in a Tibetan Population.” Journal of Population Research 18(1): 52-66. ———. 2003. “Polyandry and Population Growth in a Historical Tibetan Society.” History of the Family 8(3): 423–44. ———. 2004. “Ethnographic and Demographic Analysis of Small Populations Using the Own-Children Method.” Field Methods 16(4): 379–99. ———. 2006. “Namas (mna’ ma) and Nyelus (nyal bu): Marriage, Fertility, and Illegitimacy in Tibetan Societies.” In Tibetan Borderlands, edited by P. C. Klieger, 89-113. Leiden: Brill. ———. 2008. Tibetan Transitions: Historical and Contemporary Perspectives on Fertility, Family Planning and Demographic Change. Leiden: Brill. Childs, Geoff, Melvyn C. Goldstein, Ben Jiao, and Cynthia M. Beall. 2005. “Tibetan Fertility Transitions in China and South Asia.” Population and Development Review 31(2): 337-49. Childs, Geoff, Melvyn C. Goldstein, and Puchung Wangdui. 2011. “An Entrepreneurial Transition? Development and Economic Mobility in Rural Tibet.” Himalaya 30(1-2): 51-62. CNN. 2003. “Tibetans, Muslim Huis clash in China.” CNN, 23 February. Coleman, William M. 2002. “The Uprising at Batang: Khams and its significance in Chinese and Tibetan history.” In Khams pa Histories, edited by Lawrence Epstein, 31-55. Leiden: Brill. Costello, Susan E. 2002. “The Economics of Cultural Production in Contemporary Amdo.” In Amdo Tibetans in Transition: Society and Culture in the Post-Mao Era, edited by Toni Huber, 221-240. Leiden: Brill. ———. 2003. “The Flow of Wealth in Golok Pastoralist Society: Toward an assessment of Local Financial Resources for Economic development.” Paper presented at the 10th International Conference of the International Association for Tibetan Studies, Oxford, September 2003. (CPSY) Department of Population, Social, Science and Technology Statistics, National Bureau of Statistics. Various years up to 2011. China Population Statistical Yearbook [various years up to 2011]. Beijing: China Statistical Press. (CSY) National Bureau of Statistics. Various years up to 2011. China Statistical Yearbook [various years up to 2011]. Beijing: China Statistical Press.

396

Bibliography

(CTIC) China Tibet Information Center. 2010a. “More Tibetan Graduates Work in Inland Areas.” China Tibet Online, 26 May 2010. http://chinatibet.people.com. cn/7000128.html. (CTIC) China Tibet Information Center. 2010b. “Inland High Schools to Enroll More Tibetan students.” China Tibet Online, 22 June 2010. http://chinatibet. people.com.cn/7035313.html. Diaz-Alejandro, Carlos F. 1984. “Latin American debt: I Don’t Think We Are in Kansas Anymore.” Brookings Papers on Economic Activity 2: 335-90. Davis, Kingsley. 1963. “The Theory of Change and Response in Modern Demographic History.” Population Index 29(4): 345-66. De Janvry, Alain. 1975. “The Political Economy of Rural Development in Latin America: An Interpretation.” American Journal of Agricultural Economics 57(3): 490-99. ———. 1981. The Agrarian Question and Reformism in Latin America. Baltimore: Johns Hopkins University Press. Dillon, Michael. 1999. China’s Muslim Hui Community: Migration, Settlement and Sects. Surrey: Curzon Press. Dreyer, June Teufel. 2003. “Economic Development in Tibet under the People’s Republic of China.” Journal of Contemporary China 12(36): 411–30. Dreyfus, Georges. 1998. “The Shuk-den Affair: History and Nature of a Quarrel.” Journal of the International Association of Buddhist Studies 21(2): 227-70. ———. 2002. “Tibetan Religious Nationalism: Western Fantasy or Empowering Vision?” In Voices of Difference, edited by P. Christiaan Klieger, 37-56. Leiden: Brill. Du, Fachun. 2012. “Ecological Resettlement of Tibetan Herders in the Sanjiangyuan: A Case Study in Madoi County of Qinghai.” Nomadic Peoples 16(1): 116-33. Dyson, Tim. 2001. “A Partial Theory of World Development: The Neglected Role of the Demographic Transition in the Shaping of Modern Society.” International Journal of Population Geography 7(2): 67-90. ———. 2011. “The Role of the Demographic Transition in the Process of Urbanization.” Population and Development Review 37(supplement): 34-54. Fan, C. Cindy. 2005. “Interprovincial Migration, Population Redistribution, and Regional Development in China: 1990 and 2000 Census Comparisons.” Professional Geographer 57(2): 295-311. Fischer, Andrew Martin. 2002. Poverty by Design: The Economics of Discrimination in Tibet. Montreal: Canada Tibet Committee. ———. 2004. “Urban Fault Lines in Shangri-La: Population and Economic Foundations of Interethnic Conflict in the Tibetan Areas of Western China.” Crisis States Working Paper no. 42. London: Crisis States Programme, London School of Economics. ———. 2005a. State Growth and Social Exclusion in Tibet: Challenges of Recent Economic Growth. Copenhagen: Nordic Institute of Asian Studies Press. ———. 2005b. “Close Encounters of an Inner Asian Kind: Tibetan-Muslim Co-existence and Conflict Past and Present.” Crisis States Working Paper no 68. London: Crisis States Programme, London School of Economics. ———. 2006. “Subsistence Capacity: The Commodification of Rural Labour Reexamined through the Case of Tibet.” DESTIN Working Paper no. 06-75. London: Development Studies Institute, London School of Economics.



Bibliography 397

———. 2007. “A Theory of Polarisation, Exclusion and Conflict within Disempowered Development: The Case of contemporary Tibet in China.” PhD diss., London School of Economics. ———. 2008a. “Resolving the Theoretical Ambiguities of Social Exclusion with reference to Polarisation and Conflict.” DESTIN Working Paper no. 90, London School of Economics. ———. 2008b. “The Muslim Cook, the Tibetan Client, his Lama and Their Boycott: Modern Religious Discourses of Anti-Muslim Economic Activism in Amdo.” In Conflict and Social Order in Tibet and Inner Asia, edited by Toni Huber and Fernanda Pirie, 159-92. Leiden: Brill. ———. 2008c. “Tibet.” In Guerres d’aujourd’hui: Pourquoi ces conflits? Peut-on les résoudre? Edited by Sara Daniel, 407-53. Paris: Éditions de la Villa. ———. 2008d. “‘Population Invasion’ versus Urban Exclusion in the Tibetan Areas of Western China.” Population and Development Review 34(4): 631-62. ———. 2008e. “Subsistence and Rural Tibetan Household Livelihood Strategies in the Context of Rapid Economic and Social Transition.” Journal of the International Association of Tibet Studies 4: 1-49. ———. 2009a. “Educating for Exclusion in Western China: Institutional Foundations of Conflict in the Tibetan Areas of Qinghai.” CRISE Working Paper no. 69, Oxford: Centre for Research on Inequality, Security and Ethnicity, Queen Elizabeth House, Oxford University. ——— 2009b. “The Political Economy of Boomerang Aid in China’s Tibet.” China Perspectives 2009(3): 38-55. ——— 2010. “Is China Turning Latin? China’s Balancing Act between Power and Dependence in the Lead up to Global Crisis.” Journal of International Development 22(6): 739–57. ——— 2011a. “Reconceiving Social Exclusion.” BWPI Working Paper 146, Brooks World Poverty Institute, University of Manchester. ——— 2011b. “Beware the Fallacy of Productivity Reductionism.” European Journal of Development Research 23: 521-26. ——— 2011c. “The Great Transformation of Tibet? Rapid Labour Transitions in Times of Rapid Growth.” Himalaya 30(1-2): 63-77. ———. 2012a. “The Geopolitics of Politico-Religious Protest in East Tibet.” Cultural Anthropology Hot Spots Forum, April. http://www.culanth.org/?q=node/530. ———. 2012b. “Inequality and the Universalistic Principle in the Post-2015 Development Agenda.” Official Background Paper prepared for the global thematic consultation on Addressing Inequalities in the Post-2015 Development Agenda, UNICEF and UN Women. Frank, André Gunder. 1971. Capitalism and Underdevelopment in Latin America. London: Penguin. Furtado, Celso. 1956. Uma Economia Dependente. Rio de Janeiro: Ministério da Educação e Cultura. ———. 1967. “Intra-Country Discontinuities: Towards a Theory of Spatial Structures.” Social Science Information 6(6): 7-16. ———. 1973. “The Brazilian ‘Model.’” Social and Economic Studies March: 122-31. ———. 1983. Accumulation and Development: the Logic of Industrial Civilization. Oxford: Martin Robertson (translated from 1978 Portguese version).

398

Bibliography

Gaerrang. 2011. “The Alternative to Development on the Tibetan Plateau: Preliminary Research on the Anti-Slaughter Movement.” Revue d’Etudes Tibetaines 21: 31-43. ———. 2012. “Alternative Development on the Tibetan Plateau: The Case of the Slaughter Renunciation Movement.” PhD diss., University of Colorado, Boulder. Germano, David F. 1998. “Re-membering the Dismembered Body of Tibet: The Contemporary Ter Movement in the PRC.” In Buddhism in Contemporary Tibet: Religious Revival and Cultural Identity, edited by Melvyn Goldstein and Matthew Kapstein, 53-94. Berkeley: University of California Press. Gladney, Dru C. 1991. Muslim Chinese: Ethnic Nationalism in the People’s Republic. Cambridge, MA: Harvard University Press. Goldstein, Mely C. 1971. “Serfdom and Mobility: An Examination of the Institution of Human Lease in Traditional Tibetan Society.” Journal of Asian Studies 30(3): 521-34. ———. 1986. “Re-examining Choice, Dependency and Command in the Tibetan Social System: Tax Appendages and Other Landless Serfs.” Journal of Tibet Studies 11(4): 79-112. ———. 1987. “On The Nature of the Tibetan Peasantry: a rejoinder.” Tibet Journal 12: 61-65. ———. 1989a. “Freedom, Servitude and the ‘Servant-Serf’ Nyima: A Re-rejoinder to Miller.” Tibet Journal 14(2): 56-60. ———. 1989b. A History of Modern Tibet, 1913-1951: The Demise of the Lamaist state. Berkeley: University of California Press. ———. 1997. The Snow Lion and the Dragon: China, Tibet, and the Dalai Lama. Berkeley: University of California Press. ———. 1998. “Introduction” and “The Revival of Monastic Life in Drepung Monastery.” In Buddhism in Contemporary Tibet: Religious Revival and Ethnic Identity, edited by Melvyn Goldstein and Matthew Kapstein, 1-45. Berkeley: University of California Press. ——— 2007. A History of Modern Tibet, volume 2: The Calm before the Storm: 19511955. Berkeley: University of California Press. Goldstein, Melvyn C., Geoff Childs, and Puchung Wangdui. 2008. “‘Going for Income’: A Longitudinal Analysis of Change in Farming Tibet, 1997-98 to 200607.” Asian Survey 48(3): 514-34. ———. 2010. “Beijing’s ‘People First’ Development Initiative for the Tibet Autonomous Region’s Rural Sector: A Case Study from the Shigatse Area.” China Journal 63(1): 57-75. Goldstein, Melvyn C., Ben Jiao, Cynthia M. Beall, and Phuntsog Tsering. 2002. “Fertility and Family Planning in Rural Tibet.” China Journal 47: 25-39. ———. 2003. “Development and Change in Rural Tibet: Problems and Adaptations.” Asian Survey 43(5): 758-79. Goldstein, Melvyn C., and Matthew Kapstein, eds. 1998. Buddhism in Contemporary Tibet: Religious Revival and Ethnic Identity. Berkeley: University of California Press. Goldstein, Melvyn C., William Siebenschuh, and Tashi Tsering. 1997. The Struggle for Modern Tibet: The Autobiography of Tashi Tsering. Armonk, NY: M. E. Sharpe. Goodkind, Daniel M. 2004. “China’s Missing Children: The 2000 Census Underreporting Surprise.” Population Studies 58(3): 281-95.



Bibliography 399

Goodman, David S. G. 2004a. “The Campaign to ‘Open Up the West’: National, Provincial-Level, and Local Perspectives.” China Quarterly 178: 317-34. ———. 2004b. “Qinghai and the Emergence of The West: Nationalities, Communal Interaction and National Integration.” China Quarterly 178: 379-99. Greatrex, Roger. 1994. “A Brief Introduction to the First Jinchuan War (17471749).” In Tibetan Studies: Proceedings of the 6th Seminar of the International Association for Tibetan Studies, Fagernes 1992, Volume 1, edited by Per Kvaerne, 247–63. Oslo: Institute for Comparative Research in Human Culture. Greenhalgh, Susan. 2003. “Science, Modernity, and the Making of China’s Onechild Policy.” Population and Development Review 29(2): 163-96. (GSY) Gansu Bureau of Statistics. Various years up to 2011. Gansu Statistical Yearbook [various years up to 2011]. Beijing: China Statistical Press. Guo, Daping. 1996. “China’s Tibetan Population and Population in Tibet.” China Population Today 13(2): 7-8. Guo, Yuhua, and Aishu Chang. 2006. “Life Cycle and Social Security: A Sociological Exploration of the Life Course of Laid-Off Workers.” Social Sciences in China 26(1): 19-33. Hall, Peter A., and Rosemary C. R. Taylor. 1996. “Political Science and the Three New Institutionalisms.” Political Studies 44(5): 936-57. Hansen, Peter. 2003. “Why Is There No Subaltern Studies for Tibet?” Tibet Journal 28(4): 7-22. Harrell, Stevan. 2001. Ways of Being Ethnic in Southwest China. Seattle: University of Washington Press. Harris, Nancy S., Patricia B. Crawford, Yeshe Yangzom, Lobsang Pinzo, Palden Gyaltsen, and Mark Hudes. 2001. “Nutritional and Health Status of Tibetan Children Living at High Altitudes.” New England Journal of Medicine 344: 341-47. Hechter, Michael. 1975. Internal Colonialism: The Celtic Fringe in British National Development, 1596-1966. Berkeley: University of California Press. Hill, Polly. 1986. Development Economics on Trial: The Anthropological Case for a Prosecution. Cambridge: Cambridge University Press. Hillman, Ben. 2008. “Money Can’t Buy Tibetans’ Love.” Far Eastern Economic Review 171(3): 8-16. Hills, John. 1998. “Does Income Mobility Mean That We Do Not Need to Worry About Poverty?” In ‘Exclusion, Employment and Opportunity’, edited by Anthony B. Atkinson and John Hills, 34-55. London: CASE paper 4, London School of Economics. Hilton, Isabel. 1999. The Search for the Panchen Lama. London: Viking. Hirschman, Albert O. 1970. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. Cambridge, MA: Harvard University Press. Horlemann, Bianca. 2002a. “A Re-evaluation of the Tibetan Conquest of EighthCentury Shaxhou/Dunhuang.” In Tibet, Past and Present: Tibet Studies I, edited by Henk Blezer, 49-66. Leiden: Brill. ———. 2002b. “Modernisation Efforts in Mgo Log: A Chronicle, 1970-2000.” In Amdo Tibetans in Transition: Society and Culture in the Post-Mao era, edited by Toni Huber, 241-69. Leiden: Brill. Hu, Xiaojiang. 2003. “The Little Shops of Lhasa, Tibet: Migrant Businesses and the Formation of Markets in a Transitional Economy.” PhD diss., Harvard University.

400

Bibliography

Hu, Xiaojiang, and Miguel A. Salazar. 2008. “Ethnicity, Rurality and Status: Hukou and the Institutional and Cultural Determinants of Social Status in Tibet.” China Journal 60: 1-21. Huang, Jikun, Scott Rozelle, and Honglin Wang. 2006. “Fostering or Stripping Rural China: Modernizing Agriculture and Rural to Urban Capital Flows.” Developing Economies 44(1): 1-26. Huang, Yasheng.1995. “China’s Cadre Transfer Policy toward Tibet in the 1980s.” Modern China 21(2): 184-204. Hussain, Athar. 2002a. “Demographic Transition in China and its Implications.” World Development 30(10): 1823-34. ———. 2002b. “Second Lag: Inter-Provincial Inequality and Provincial Growth Patterns.” Unpublished manuscript supplied by the author. ———. 2003. “Urban Poverty in China: Measurement, Patterns and Policies.” InFocus Programme on Socio-Economic Security. Geneva: International Labour Office. Hussain, Athar, Robert Cassen, and Tim Dyson. 2006. “Demographic Transition in Asia and Its consequences.” IDS Bulletin 37(3): 79-87. ICT. 2012a. “Evidence of Overt Chinese Discrimination against Tibetans in the Job Market.” Washington, DC: International Campaign for Tibet. ———. 2012b. “60 Years of Chinese Misrule: Arguing Cultural Genocide in Tibet.” Washington, DC: International Campaign for Tibet. Iredale, Robyn, Naran Bilik, and Wang Su. 2001. Contemporary Minority Migration, Education, and Ethnicity in China. Cheltenham, UK: Edward Elgar. Iredale, Robyn, Naran Bilik, and Fei Guo. 2003. China’s Minorities on the Move: Selected Case Studies. Armonk, NY: M. E. Sharpe Iselin, Lilian. 2011. “Modern Education and Changing Identity Constructions in Amdo.” Himalaya 30(1 & 2): 91-100. Jackson, Cecile. 1996. “Rescuing Gender from the Poverty Trap.” World Development 24(3): 489-504. Jagou, Fabienne. 2002. “The Sixth Panchen Lama’s Chinese Titles.” In Khams Pa Histories, edited by Lawrence Epstein, 85-102. Leiden: Brill. Jin, Wei. 2010. Xizang yuan zhu yu fa zhan (Tibet: Assistance & Development). Lhasa: Xizang ren min chu ban she (Tibet People’s Publishing House). Kapstein, Matthew T. 2000. The Tibetan Assimilation of Buddhism: Conversion, Contestation, and Memory. Oxford: Oxford University Press. ———, ed. 2009. Buddhism between Tibet and China. Somerville, MA: Wisdom Publications. Kay, Cristóbal. 1989. Latin American Theories of Development and Underdevelopment. New York: Routledge. ———. 2005. “Celso Furtado: Pioneer of Structuralist Development Theory.” Development and Change 36(6): 1201-7. Khan, Azizur Rahman, and Carl Riskin. 2001. Inequality and Poverty in China in the Age of Globalization. Oxford: Oxford University Press. Knight, John, and Lina Song. 1999. The Rural-Urban Divide: Economic Disparities and Interactions in China. Oxford: Oxford University Press. Kwok, K. 2008. “Monks Need Education, Say Mainland Scholars.” South China Morning Post, 27 March.



Bibliography 401

Lan, Fang. 2012. “Aid Programs in Tibet Lack Efficiency, Says Scholar.” CaixinOnline, 18 December. http://english.caixin.com/2012-12-18/100473750.html Lee, Grace O. M., and Malcolm Warner. 2005a. “Unemployment in the People’s Republic of China.” In Unemployment in Asia, edited by John Benson and Ying Zhu, 97-114. New York: Routledge. ———. 2005b. “The Management of Human Resources in Shanghai.” In China’s Business Reforms: Institutional Challenges in a Globalized Economy, edited by Russell Smyth, On Kit Tam, Malcolm Warner, and Cherrie Jiuhua Zhu, 124-138. New York: Routledge Curzon. Lee, James Z., and Feng Wang. 1999. One Quarter of Humanity: Malthusian Mythology and Chinese Realities, 1700-2000. Cambridge, MA: Harvard University Press. Lewis, W. Arthur. 1954. “Economic Development with Unlimited Supplies of Labor.” Manchester School of Economic and Social Studies 22: 139-91. ———. 1978. The Evolution of the International Economic Order. Princeton NJ: Princeton University Press. Li, Lulu. 2008. “Transition et stratification sociale dans les villes chinoises.” In La nouvelle sociologie chinoise, edited by Laurence Roulleau-Berger, Yuhua Guo, Peilin Li, and Shiding Liu, 119-144. Paris: Editions du CNRS. Li, Qiang. 2002. “Stratification in China’s Urban Society during the Transition Period.” Social Sciences in China 23(1). Li, Shi. 2012. “Making a More Effective Income Distribution Policy for Inclusive Development: China’s Experiences.” Paper prepared for the Global Poverty Reduction and Development Forum on Inclusive Development and Poverty Reduction, International Poverty Reduction Center in China, UNDP China, October 17, 2012, Beijing, 41-64. Liu, Shiding. 2008. “Structure du marché du travail et changement institutionnel dans la Chine contemporaine.” In Actes du Colloque La Chine et l’internationalisation de la sociologie, edited by Laurence Roulleau-Berger, June 2008. Lipman, Jonathon N. 1997. Familiar Strangers: A history of Muslims in Northwest China. Seattle: University of Washington Press. Liu, Yi, and Chusheng Lin. 2001. “The Tax Division Scheme: Central-Local Financial Relations in the 1990s.” In China’s Regional Disparities: Issues and Policies, edited by V. F. S. Sit and D. Lu, 85-104. Huntington, NY: Nova Science Publishers. Liu, Yunguo, and Gerald Bloom. 2002. “Designing a Rural Health Reform Project: The Negotiation of Change in China.” IDS Working Paper no.150, Brighton: Institute of Development Studies. Livi-Bacci, Massimo. 2001. A Concise History of World Population. Cambridge, MA: Blackwell. Lu, Feng. 2000. “China’s Grain Economy and Trade Policy.” In The Chinese Economy under Transition, edited by Sarah Cook, Shujie Yao and Juzhong Zhuang, 180– 197. London: Macmillan. Lustgarten, Abrahm. 2006. “Next Stop, Lhasa.” Fortune Magazine, 8 June, accessed on 12 June 2006 at http://money.cnn.com/magazines/fortune/fortune_archive/2006/06/12/8379261/index.htm. Ma, Rong. 1996. Xizang de renkou yu shehui (Population and Society of Tibet). Beijing: Tongxin Publishing House.

402

Bibliography

———. 1998. “Economic Patterns of the Tibet Autonomous Region.” In Development, Society, and Environment in Tibet: Papers presented at a panel of the 7th Seminar of the International Association for Tibetan Studies, Graz 1995, edited by Graham E. Clarke, 167-86. Vienna: Verlag der Osterreichischen Akademie der Wissenschaften.  ———. 2007. “A New Perspective in Guiding Ethnic Relations in the 21st Century: ‘De-politicization’ of Ethnicity in China.” Asian Ethnicity 8(3): 199-217. Ma, Rong, and Pan Nagu. 1992. “Changes and Geographic Distribution of the Tibetan Population in China.” In Tibetan Population in China Today (in Chinese, translated by Chen Wei), 78-94. Beijing: China Tibetology Publishing House. Ma, Rong, and Tanzen Lhundup. 2008. “Temporary Migrants in Lhasa in 2005.” Journal of the International Association of Tibet Studies 4: 1-42. MacKerras, Colin. 1994. China’s Minorities: Integration and Modernization in the Twentieth Century. New York: Oxford University Press. Mahalanobis, Surojit. 2003. “Tibetans Are a Marginalised Lot: The Samdhong Rimpoche Interview.” Economic Times, 31 May. Makley, Charlene E. 2007. The Violence of Liberation: Gender and Tibetan Buddhist Revival in Post-Mao China. Berkeley: University of California Press. ———. 2009. “Ballooning Unrest: Tibet, State Violence and the Incredible Lightness of Knowledge.” In China in 2008: A Year of Great Signifance, edited by Kate Merkel-Hess, Kenneth L. Pomeranz, and Jeffrey N. Wasserstrom, 44-55. Lanham, MD: Rowman & Littlefield. Maconi, Lara. 2008. “Au-delà du débat linguistique: Comment définir la littérature tibétaine d’expression chinoise? Autour des ‘spécificités nationales’ et des ‘spécificités régionales.’” Revue d’études tibétaines 14: 117-56. Marshall, Steven D., and Susette Ternent Cooke. 1997. Tibet outside the TAR: Control, Exploitation and Assimilation, Development with Chinese Characteristics. Washington, DC: Alliance for Research in Tibet. McGranahan, Carole, and Ralph Litzinger, eds. 2012. “Self-Immolation as Protest in Tibet.” Cultural Anthropology Hot Spots Forum, April. http://www.culanth. org/?q=node%2F526. Meinert, Carmen. 2002. “Chinese Chan and Tibetan Rdzogs Chen: Preliminary Remarks on Two Tibetan Dunhuang Manuscripts.” In Religion and Secular Culture in Tibet: Tibetan Studies II, edited by Henk Blezer, 289-307. Leiden: Brill. Miller, Beatrice D. 1987. “A Response to Goldstein’s ‘Reexamining Choice, Dependency and Command in the Tibetan Social System.’” The Tibet Journal 12(2): 65-67. ———. 1988. “Last Rejoinder to Goldstein on Tibetan Social System.” Tibet Journal 13(3): 64-67. Morcom, Anna. 2011. “History, Traditions, Identities and Nationalism: Drawing and Redrawing the Musical Cultural Map of Tibet.” In Tibetan Studies: An Anthology, edited by S. Arslan and P. Schweiger. PIATS 2006: Proceedings of the Eleventh Seminar of the International Association for Tibetan Studies, Koenigswinter, International Institute for Tibetan and Buddhist Studies. Myrdal, Gunnar. 1957. Economic Theory and the Underdeveloped Regions. London: Duckworth.



Bibliography 403

Naughton, Barry. 2002. “Provincial Economic Growth in China: Causes and Consequences of Regional Differentiation.” In China and its Regions: Economic Growth and Reform in Chinese Provinces, edited by Mary-Françoise Renard, 57-86. Cheltenham, UK: Edward Elgar. Nee, Victor. 1992. “Organizational Dynamics of Market Transition: Hybrid Forms, Property Rights and Mixed Economy in China.” Administrative Science Quarterly 37: 1-27. Niamir, Maryam. 1990. Herder’s Decision-making in Natural Resources Management in Arid and Semi-arid Africa. Rome: Food and Agriculture Organization of the United Nations (FAO). Nietupski, Paul Kocot. 1999. Labrang: A Tibetan Buddhist Monastery at the Crossroads of Four Civilizations. Ithaca: Snow Lion Publications. ———. 2002. “Sino-Tibetan Relations in Eighteenth-Century Labrang.” In Territory and Identity in Tibet and the Himalayas, edited by Katia Buffetrille and Hildegard Diemberger, 121-33. Leiden: Brill. Oi, Jean C. 1992. “Fiscal Reform and the Economic Foundations of Local State Corporatism in China.” World Politics 45(1): 99-126. (PD) People’s Daily. 2002. “Tibetans See Fastest Population Growth in History.” People’s Daily, 5 September, last accessed on 27 September 2005 at http://english. people.com.cn/200209/05/eng20020905_102675.shtml. (PD) People’s Daily. 2008. “Official Figures: Tibetan Tourism Booming before Riots.” People’s Daily, 25 April. Peng, Wenbin. 2002. “Frontier Process, Provincial Politics and Movements for Khampa Autonomy during the Republican Period.” In Khams Pa Histories: Visions of People, Place and Authority, edited by Lawrence Epstein, 57-84. Leiden, Brill. Peng, Xizhe. 1987. “Demographic Consequences of the Great Leap Forward.” Population and Development Review 13(4): 639-70. Pirie, Fernanda. 2005. “Feuding, Mediation and the Negotiation of Authority among the Nomads of Eastern Tibet.” Max Planck Institute for Social Anthropology Working Paper 72. Halle: Max Planck Institute for Social Anthropology. Polanyi, Karl. 2001 [1944]. The Great Transformation: The Political and Economic Origins of Our Time. Boston: Beacon Press. Polanyi Levitt, Kari. 2005. “Raul Prebisch and Arthur Lewis: The Basic Dualities of Development Economics.” In The Pioneers of Development Economics, edited by K. S. Jomo, 193-208. Zed: London. ———. 2006. “Keynes and Polanyi: the 1920s and The 1990s.” Review of International Political Economy 13(1): 152–77. Postiglione, Gerard. 1992. “China’s National Minorities and Educational Change.” Journal of Contemporary Asia 22(1): 20-44. ———. (2008). “Making Tibetans in China: the Educational Challenges of Harmonious Multiculturalism.” Educational Review 60(1): 1-20. ———. (2009). ”Dislocated Education: The Case of Tibet.” Comparative Education Review 53(4): 483-512. Postiglione, Gerard, Ben Jiao, and Sonam Gyatso. 2006a. ”Household Perspectives on School Attendance in Rural Tibet.” Educational Review 58(3): 317-37. ———. 2006b. “Education in Rural Tibet: Development, Problems and Adaptations.” In Education and Social Change in China: Inequality in a Market Economy, edited by Gerard A. Postiglione, 136-62. Armonk, NY: M. E. Sharpe.

404

Bibliography

Postiglione, Gerard, Ben Jiao, and Li Xiaoliang. 2012. “Education Change and Development in Nomadic Communities of the Tibetan Autonomous Region (TAR).” International Journal of Chinese Education 1: 89-105. Postiglione, Gerard, Zhiyong Zhu, and Ben Jiao. 2004. “From Ethnic Segregation to Impact Integration: State Schooling and Identity Construction for Rural Tibetans.” Asian Ethnicity 5(2): 195-217. PRC. 1998. New Progress in Human Rights in the Tibet Autonomous Region. White Paper, February. Beijing: Information Office of the State Council of the People’s Republic of China. ———. 2000a. National Minorities Policy and its Practice in China. White Paper, June. Beijing: Information Office of the State Council of the People’s Republic of China. ———. 2000b. The Development of Tibetan Culture. White Paper, June. Beijing: Information Office of the State Council of the People’s Republic of China. ———. 2001a. One Hundred Questions and Answers about Tibet. Beijing: China Tibet Information Center. ———. 2001b. Tibet’s March toward Modernization. White Paper, 8 November. Beijing: Information Office of the State Council. ———. 2004. Regional Ethnic Autonomy in Tibet. White Paper, May. Beijing: Information Office of the State Council of the People’s Republic of China. ———. 2005. Regional Autonomy for Ethnic Minorities in China. White Paper, February. Beijing: Information Office of the State Council of the People’s Republic of China. (Qinghai Tabulation) Department of Population, Social, Science and Technology Statistics, National Bureau of Statistics. 2002. Tabulation on the 2000 Population Census of Qinghai. Beijing: China Statistical Press. (QSY) Qinghai Statistical Bureau. Various dates up to 2011. Qinghai Statistical Yearbook [various dates up to 2011]. Beijing: China Statistical Press. Rabgey, Tashi, and Tseten Wangchuk Sharlho. 2004. “Sino-Tibetan Dialogue in the Post-Mao Era: Lessons and Prospects.” Policy Studies no. 12. Washington, DC: East-West Center. Rabinovitch, Simon. 2012. “Chinese GDP: Doesn’t Add Up.” Financial Times, 15 February. Rangasami, Amrita. 1985. “‘Failure of Exchange Entitlements’ Theory of Famine: A Response.” Economic and Political Weekly 20(41 & 42): 1747-51 and 1797-1800. RFA. 2004. “Hundreds of Jobless Tibetans Protest in Qinghai Province.” Radio Free Asia, 12 October. http://www.savetibet.org/News/News.cfm?ID=2764&c=6 2004. ———. 2006. “Tibetan, Blinded, Freed from Jail, Five Still Held for Slaughterhouse Fire.” Radio Free Asia, 2 February. http://www.rfa.org/english/news/social/2006/02/01/tibet_fire/. ———. 2007. “Chinese Authorities Detain Monks, Others after Tibetan-Muslim Clashes.” Radio Free Asia, 22 August. http://www.rfa.org/english/tibet/tibet_muslim-20070822.html. Robin, Françoise. 2011. “L’histoire au miroir de l’art: Décrire la catastrophe au Tibet.” Atelier D 01: Minorités Tibétaines de la République Populaire de Chine Créations artistiques et littéraires contemporaines. http://www.reseau-asie.com/ userfiles/file/D01_robin_creation_litteraire_histoire_tibet.pdf. ———. 2012. “Ceci n’est pas un navire. Interprétations et lectures du poème de Jangbu (Ljang bu) Ce navire peut-il nous mener sur l’autre rive? (Gru gzings chen po



Bibliography 405

‘dis nga tsho pha rol tu sgrol thub bam, 2000).” Impressions d’Extrême-Orient 2011(2). http://ideo.revues.org/219 ———. Forthcoming. “Occupés à créer. Les artistes et écrivains tibétains et l’État chinois.” In Vivre ensemble au Tibet: Vers une solution au conflit sino-tibétain? edited by S. Corcuff. Lyon: Éditions de l’École Normale Supérieure de Lyon. Roth, Kenneth. 2004. “Defending Economic, Social and Cultural Rights: Practical Issues Faced by an International Human Rights Organization.” Human Rights Quarterly 26: 63–73. Roulleau-Berger, Laurence. 2009. “‘New Chinese Society,’ inequalities and Fragmentation.” DevISSues 11(1): 15-17. Saich, Tony. 2004. Governance and Politics of China. New York: Palgrave Macmillan. Samuel, Geoffrey. 1982. “Tibet as a Stateless Society and Some Islamic Parallels.” Journal of Asian Studies 41(2): 215-29. ———. 1993. Civilized Shamans: Buddhism in Tibetan Societies. Washington, DC: Smithsonian Institution Press. ———. 2002. “Buddhism and the State in Eighth Century Tibet.” In Religion and Secular Culture in Tibet: Tibetan Studies II, edited by Henk Blezer, 1-19. Leiden: Brill. Sautman, Brian. 1999. “Ethnic Law and Minority Rights in China: Progress and Constraints.” Law and Policy 21(3): 283-314. ———. 2008. “Protests in Tibet and Separatism: The Olympics and Beyond,” China Left Review 1. ———. 2012. “Tibet’s Suicidal Politics.” East Asia Forum, March 21. Accessed 29 March 2012, http://www.eastasiaforum.org/2012/03/21/tibet-s-suicidal-politics/. Sautman, Brian, and Eng, Irene. 2001. “Tibet: Development for Whom?” China Information 15(2): 20-74. Scott, James. 1976. The Moral Economy of the Peasant: Rebellion and Subsistence in Southeast Asia. New Haven, CT: Yale University Press. ———. 1985. Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven, CT: Yale University Press. ———. 1990. Domination and the Arts of Resistance: Hidden Transcripts. New Haven, CT: Yale University Press. ———. 2009. The Art of Not Being Governed: An Anarchist History of Southeast Asia. New Haven, CT: Yale University Press. Schwartz, Ronald D. 1994. Circle of Protest: Political Ritual in the Tibetan Uprising. London: Hurst. Shakya, Tsering. 1999. The Dragon in the Land of Snows. London: Pimlico. ———. 2002. “Blood in the Snows: Reply to Wang Lixiong,” New Left Review 15 (May/June): 39–60. ———. 2008. “Tsering Shakya: Tibetan Questions—Interview.” New Left Review 51 (May/June): 5-27. Shaoguang, Wang. 2004. “China’s Health System: From Crisis to Opportunity.” Yale-China Health Journal 3: 5-49. Shen, Minggao. 2009. “Macroreview: Deconstructing China’s Investment.” Caijing. 16 July. Siebenschuh, William R., and Tashi Tsering. 2003. The Struggle for Education in Modern Tibet: The Three Thousand Children of Tashi Tsering. Lewiston, NY: Edwin Mellen Press.

406

Bibliography

Skinner, George William. 1994. “Differential Development in Lingnan.” In The Economic Transformation of South China: Reform and Development in the Post-Mao Era, edited by Thomas P. Lyons and Victor Nee, 17-56. Ithaca, NY: Cornell University Press. Smith, Oliver. 2012. “China to Build £3bn Theme Park in Tibet.” Telegraph, 9 July 2012. Smith, Warren. 2009. Tibet’s Last Stand: The Tibetan Uprising of 2008. Lanham, MD: Rowman & Littlefield. Smyer Yu, Dan. 2011. The Spread of Tibetan Buddhism in China: Charisma, Money, Enlightenment. London: Routledge. Solinger, Dorothy J. 1993. China’s transition from socialism: statist legacies and market reforms, 1980-1990, Armonk, NY: M. E. Sharpe. ———. 1999. Contesting Citizenship in Urban China: Peasant Migrants, the State, and the Logic of the Market. Berkeley: University of California Press. ———. 2002. “Labour Market Reform and the Plight of the Laid-Off Proletariat.” China Quarterly 170: 304-26. Sperling, Elliot. 2004. “The Tibet-China Conflict: History and Polemics.” Policy Studies no. 7. Washington, DC: East-West Center. (SSY) Sichuan Bureau of Statistics. Various years up to 2011. Sichuan Statistical Yearbook [various years up to 2011]. Beijing: China Statistical Press. Stewart, Frances. 2002. “Horizontal Inequalities: A Neglected Dimension of Development.” CRISE Working Paper no.1. Oxford: Queen Elizabeth House, Oxford University. Sulek, Emilia. 2011. “Disappearing Sheep: The Unexpected Consequences of the Emergence of the Caterpillar Fungus Economy in Golok, Qinghai, China.” Himalaya 30(1-2): 9-22. Sun, Huaiyang, and Xiru Li. 1996. “The Evolution and Current Status of China’s Tibetan Population.” Chinese Journal of Population Science 8(2): 221-29. Sunkel, Osvaldo. 1973. “Transnational Capitalism and National Disintegration in Latin America.” Social and Economic Studies 22(1): 132-76. Swift, Jeremy J. 1977. “Sahelian Pastoralists: Underdevelopment, Desertification and Famine.” Annual Review of Anthropology 6: 457-78. ———. 1989. “Why Are Rural People Vulnerable to Famine?” IDS Bulletin 20(2): 8-15. (Tabulation) Department of Population, Social, Science and Technology Statistics, National Bureau of Statistics. 2002. Tabulation on the 2000 Population Census of the People’s Republic of China. Beijing: China Statistical Press. Tao, Ran, Fubing Su, Mingxing Liu, and Guangzhong Cao. 2010. “Land Leasing and Local Public Finance in China’s Regional Development: Evidence from Prefecture-Level Cities.” Urban Studies 47(10): 2217-36. Tashi, Nyima, Liu Yanhua, and Tej Partap. 2002. Making Tibet Food Secure: Assessment of Scenarios. Kathmandu: International Centre for Integrated Mountain Development. (TGIE) The Government of Tibet in Exile. 1993. “Forced Abortion and Sterilization in Tibet.” Dharamsala, India: Department of Information & International Relations, 23 December. ———. 1996. Tibet: Proving Truth from Facts. London: Office of Tibet. Tibetinfonet. 2011. “Tourism Booming in Tibet.” Tibetinfonet Update, 28 January. TIN. 1994. Survey of Birth Control Policies in Tibet. London: Tibet Information Network.



Bibliography 407

———. 1999a. A Sea of Bitterness; Patriotic Education in Qinghai Monasteries. London: Tibet Information Network. ———. 1999b. Relative Freedom? Tibetan Buddhism and Religious Policy in Kardze, Sichuan, 1987-1999. London: Tibet Information Network. ———. 2001. “Expulsion of Nuns and Students Threaten Survival of Important Tibetan Buddhist Institute.” Tibet Information Network News Update, 19 August. ———. 2004c. “Death of Respected Chabcha Leader.” Tibet Information Network News Update, 26 August. TIN/Fischer. 2003a. “Despite Economic Boom, Rural Standards of Living in the Tibet Autonomous Region Still below 1992 Levels.” Tibet Information Network News Update, 6 February. ———. 2004. “Central Lhasa Gets Facelift with ‘Tibetan Characteristics.’” Tibet Information Network News Update, 30 December. ———. 2005. “Tibetans Lose ground in Public Sector Employment in the TAR: Streamlining Effectively Discriminates against Tibetans.” Tibet Information Network News Update, 20 January 2005. Trungpa, Chogyam. 1973. Cutting through Spiritual Materialism. Boston: Shambhala. TSGUK. No date. “Birth Control Policies in Tibet.” London: Tibet Support Group UK. Tsui, Kai-yuen, and Youqiang Wang. 2004. “Between Separate Stoves and a Single Menu: Fiscal Decentralization in China.” China Quarterly 177: 71-90. (TSY) Tibet Bureau of Statistics. Various years up to 2011. Tibet Statistical Yearbook [various years up to 2011]. Beijing: China Statistical Press. Tuttle, Gray. 2005. Tibetan Buddhists in the Making of Modern China. New York: Columbia University Press. (UN) Population Division of the United Nations Secretariat. 2002. World Population Prospects: The 2002 Revision, Volume I: Comprehensive Tables. Geneva: United Nations. UN. 2004. “UNICEF Head Wraps Up Tibet Visit Aimed at Slashing Disparities with Rest of China.” Geneva: UN News Centre. (UN/Prebisch) United Nations Economic Commission for Latin America 1950. Economic Development of Latin America and Its Principal Problems. Lake Success NY: United Nations Department of Economic Affairs. UNDP. 1999. China Human Development Report: Transition and the State. Oxford: Oxford University Press. UNPO. 1997. “‘China’s Tibet’: The World’s Largest Remaining Colony. Report of a Fact-Finding Mission and Analyses of Colonialism and Chinese Rule in Tibet.” The Hague: Unrepresented Nations and Peoples Organization. UNRISD. 2010. Combating Poverty and Inequality: Structural Change, Social Policy and Politics. Geneva: United Nations Research Institute on Social Development.  van Beek, Martijn. 2000. “Beyond Identity Fetishism: Communal Conflict in Ladakh and the Limits of Autonomy.” Cultural Anthropology 15(4): 525-69. van Spengen, Wim. 2002. “Frontier History of Southern Kham: Banditry and war in the Multi-ethnic Fringe Lands of Chatring, Mili, and Gyethang, 1890-1940.” In Khams Pa Histories, edited by Lawrence Epstein, 7-29. Leiden: Brill. VOA. 2008. “Dalai Lama: China’s Demographic Policy Threatens Tibetan Culture.” Voice of America News, 29 March. VOA. 2012. “A Tibetan Lama Self-Immolates in Golok Darlak.” Voice of America News, 8 January.

408

Bibliography

Wang, Lixiong. 1998. Tian Zang: Xizang de Mingyun (trans. Sky Burial: The Fate of Tibet). Brampton, ON: Mirror Books. Wang, Xiaoqiang, and Nanfeng Bai. 1991 [1986]. The Poverty of Plenty. Basingstoke: Macmillan (translated from the original Chinese version published in 1986). Warren, Bill. 1973. “Imperialism and Capitalist Industrialization.” New Left Review 81: 3-46. ———. 1980. Imperialism: Pioneer of Capitalism. London: Verso. Washbrook, David A. 1988. “Progress and Problems: South Asian Economic and Social History c1720-1860.” Modern Asian Studies 22(1): 72-96. Wei, Yehua Dennis. 2000. Regional Development in China: States, Globalization and Inequality. London: Routledge. White, Ben. 2011. “Who Will Own the Countryside? Dispossession, Rural Youth and the Future of Farming.” Valedictory address delivered on 13 October 2011, International Institute of Social Studies, The Hague. Whiting, Susan H. 2001. Power and Wealth in Rural China: The Political Economy of Institutional Change. Cambridge: Cambridge University Press. Williams, Dee Mack. 1997. “The Desert Discourse of Modern China.” Modern China 23(3): 328-55. Winkler, Daniel. 2008a. “Tracking Yartsa Gunbu (Cordyceps sinensis)—Collection, Trade and Tradition in Tibet.” Economic Botany 62(3): 291-305. ———. 2008b. “The Mushrooming Fungi Market in Tibet Exemplified by Cordyceps Sinensis and Tricholoma Matsutake.” Journal of the International Association of Tibetan Studies 4: http://www.thlib.org/?tid=T5571. Wong, Christine P., ed. 1997. Financing Local Government in the PRC. Hong Kong: Oxford University Press. Wong, Christine P., and Richard M. Bird. 2005. “China’s Fiscal System: A Work in Progress.” Working Paper no. 0515, International Tax Program, Institute for International Business, Joseph L. Rotman School of Management, University of Toronto. Woo, Wing Thye. 2002. “Some Observations on the Ownership and Regional Aspects in Financing the Growth of China’s Rural Enterprises.” In China and Its Regions: Economic Growth and Reform in Chinese Provinces, edited by Mary-Françoise Renard, 244-277. Cheltenham, UK: Edward Elgar. (WB) World Bank. 2000. Helping Countries to Combat Corruption: Progress at the World Bank since 1997. Washington, DC: World Bank. ———. 2002a. HNP Stats. Washington: World Bank, see http://devdata.worldbank. org/hnpstats/cd2.asp (accessed in January 2004). ———. 2002b. China: National Development and Sub-national Finance. A Review of Provincial Expenditures. Washington, DC: World Bank. ———. 2003. World Development Report 2004: Making Services Work for Poor People. Washington, DC: World Bank. Xi, Jinsheng, Tianlu Zhang, and Juan Liu. 1992. “Changes and Current Status of the Population in Tibet.” In Tibetan Population in China Today (in Chinese, translated by Chen Wei). Beijing: China Tibetology Publishing House. Xinhua. 2006. “Tertiary Sector Bolsters Tibet Economy,” People’s Daily Online, 18 January. ———. 2011a. “Tibet’s population Tops 3 Million; 90% are Tibetans.” China Forum 4 May. http://bbs.chinadaily.com.cn/thread-699715-1-1.html.



Bibliography 409

———. 2011b. “Tibetan Students Go to Inland Key High Schools.” China Tibet Online 27 September. http://chinatibet.people.com.cn/96069/7605296.html. Xinhuanet. 2006. “Tibet to Eliminate Illiteracy by 2010.” Chinaview.cn 15 September. http://news.xinhuanet.com/english/2006-09/15/content_5096651.htm. Yan, Hairong. 2000. “The Tibet Question and the Conundrums of Modernisation.” Asian Ethnicity 1(2): 156-63. Yang, Dali L. 1996. Calamity and Reform in China. Stanford, CA: Stanford University Press. ———. 1997. Beyond Beijing: Liberalization and the Regions in China. London: Routledge. Yang, Shuzhang, and Weijiang Liu. 1992. “A Macro Analysis of Population Growth of China’s Ethnic Groups in the 1980s.” China Population Today 9(3): 9-11. Yang, Shuzhang, and Hezhong Zhu. 1993. “China’s Ethnic Groups: Fertility Change in Five Autonomous Regions.” China Population Today 10(4): 8-14. Yeh, Emily T. 2003a. “Taming the Tibetan Landscape: Chinese Development and the Transformation of Agriculture.” PhD diss., University of California Berkeley. ———. 2003b. “Tibetan Range Wars: Spatial Politics and Authority on the Grasslands of Amdo.” Development and Change 34(3): 499-523. ———. 2007. ”Tropes of Indolence and the Cultural Politics of Development in Lhasa, Tibet.” Annals of the Association of American Geographers 97(3): 593-612. ———. 2009a. “Tibet and the Problem of Radical Reductionism.” Antipode 41(5): 983-1010. ———. 2009b. “Living Together in Lhasa: Ethnic Relations, Coercive Amity and Subaltern Cosmopolitanism.” In The Other Global City, edited by Shail Mayaram, 54-85. London: Routledge. ———. 2013. Development as Gift: Landscape Transformation and State Incorporation in Tibet. Ithaca, NY: Cornell University Press. Yeh, Emily T., and Mark Henderson. 2008. “Interpreting Tibet’s Urbanization: Administrative Scales and Discourses of Modernization.” Journal of the International Association of Tibetan Studies 4: 1-44. http://www.thlib.org/collections/texts/ jiats/#!jiats=/04/yeh/. Yi, Lin. 2005a. “Muslim Narratives of Schooling, Social Mobility and Cultural Difference: A Case Study in Multi-ethnic Northwest China.” Japanese Journal of Political Science 6(1): 1-28. ———. 2005b. “Choosing between Ethnic and Chinese Citizenships: The Educational Trajectories of Tibetan Minority Children in Northwestern China.” In Chinese Citizenship: Views from the Margins, edited by Vanessa L. Fong and Rachel Murphy, 41-67. London: Routledge. ———. 2008. Cultural Exclusion in China: State Education, Social Mobility and Cultural Difference. Oxford: Taylor & Francis. Yixing, Zhou, and Laurence Ma. 2003. “China’s Urbanization Levels: Reconstructing a Baseline from the Fifth Population Census.” The China Quarterly 173: 176-96. Yundannima. 2012. “From ‘Retire Livestock, Restore Rangeland’ to the Compensation for Ecological Services: State Interventions into Rangeland Ecosystems and Pastoralism in Tibet.” PhD diss., University of Colorado, Boulder. Zablocki, Abraham. 2009a. “The Taiwanese Connection: Politics, Piety, and Patronage in Transnational Tibetan Buddhism.” In Buddhism between Tibet and China, edited by Matthew Kapstein, 379-414. Somerville, MA: Wisdom Publications.

410

Bibliography

———. 2009b. “Transnational Tulkus: The Globalization of Tibetan Buddhist Reincarnation.” In TransBuddhism: Transmission, Translation and Transformation, edited by Nalini Bhushan, Jay Garfield, and Abraham Zablocki, 43-53. Amherst: University of Massachusetts Press. Zenz, Adrian. 2010. “Beyond Assimilation: The Tibetanisation of Tibetan Education in Qinghai.” Inner Asia 12(2): 293-315. ———. 2013. “Tibetanness” under Threat? Neo-Integrationism, Career and Market Reforms in Qinghai, P.R. China. Leiden: Brill. Zhang, Tianlu. 1990. “Growth of China’s Ethnic Population Must Be Brought under Control.” China Population Newsletter 7(5): 8-9. Zhang, Tianlu, and Mei Zhang. 1994. “The Present Population of the Tibetan Nationality in China.” Social Sciences in China 15(1): 46-65. Zhang, Tracy. 2012. “The Making of ‘Valuable’ Carpets in Lhasa.” Journal of Material Culture 17(1): 83-101. Zhang, Yi. 2001. Strangers in the City: Reconfigurations of Space, Power, and Social Networks within China’s Floating Population. Stanford, CA: Stanford University Press. Zou D. 1991. “China’s Ethnic Population and Relevant Policies.” China Population Today 8(1): 8-9.

Index

abattoir, 339-40, 347, 366n6 Aboriginal People, Canada, 21, 119, 127-28, 326, 386 acculturation, 252. See also assimilation affirmative action (or positive discrimination), 280, 382, 383-87, 392n5 Africa, 17, 24-25, 48, 77, 122n16 age structure, population, 88-91, 105-6, 159-60, 242-3n5 agency, 250, 282-83, 363-65 agrarian transition, 34, 196-203, 208, 217, 242n1, 329n2, 372 agricultural systems, 60, 292, 303 agriculture, 15, 318 aid, boomerang, 30, 129 aid, international, 22, 163 aid programs or projects, 2, 320, 385 alienation, 386-87 Amdo, xx, 50, 52-56 animal husbandry, 297-98 appropriation of local resources (e.g. land, mining), 21, 34, 130 Arendt, Hannah, 376 assets, productive fixed (rural households), 292-93, 299-304, 318, 326-27

assimilation, xxxvi, xxxviii, 11, 130, 184, 251, 276, 284, 378, 381, 38791 autonomous areas, minority, 123n25 autonomy, xxxviii, 11, 75, 326-27, 357, 364, 391; law, national minority, 35, 251, 280, 357, 381; loss of, 16, 22; national minority (or minority nationality), 23, 35, 280, 353, 391. baby boom, 89 backwardness, 10, 81n58, 85, 92, 118, 178, 181, 293-94, 319, 322, 327, 370n56, 388 barley, 69 Beijing, 109-10, 139, 221, 231-33, 264 Bhutan, 16, 59, 87 biases, cultural, linguistic and political, 12, 30, 119, 129, 195, 237, 239, 249-51, 267, 270, 272, 281-82, 387; in education provisioning, 161, 253. birth rates, crude, 90-93, 105, 123n22 boarding schools, 160, 162, 193, 256, 286n11 boycott (against Muslim businesses), 76, 116, 279, 336, 346-57, 366 brick making, 205 411

412

Index

British invasion of Tibet, 51, 56 Buddhism, 364-65 Buddhism, Tibetan, 35, 50, 355, 357-58 Buddhist discourse, 340 business, 239, 279, 284, 310, 312, 318; networks, Tibetan-Muslim, 354; Tibetan, 311, 314, 318, 326, 346, 361. cadre employment. See employment, cadre. See also cadres, Tibetan cadres: Han Chinese, 271, 274, 359; Tibetan, 23, 179, 233-35, 239, 271, 274, 278, 283, 289n44, 346. capital (or capitalist) accumulation, 14, 19, 318-19 capitalism, xxxi, 14-15, 324, 376 carpet making, 314, 341 carpentry, 332n35 cash income and expenditure, 305, 307, 322-23 caterpillar fungus, 12, 69, 156, 176, 194, 296, 303, 311-12, 316, 321, 324, 332n37, 339, 342, 345, 371 cemeteries, 352 census, population, 87, 101-113, 118, 121n4, 125n39, 240-41, 254-55, 257, 260-64, 285n9 Central Tibet, 51-52 cereal grains, 297-98, 303, 306 Chabcha, 55, 316, 343-45 Chengdu, 4, 105, 359 Chentsa (Ch. Jianza) County, 114, 347, 349, 352 Chinese Buddhists, 56, 364 Chinese Muslims, xx, 35, 53, 70, 83, 107, 112-13, 131, 311, 339-40; competition with, 236, 252, 279, 284. Chongqing, 4, 109-111 circular payments system, 66; See also regional development strategy, Maoist. civil war, China, 16, 55 collective approaches to land management, 61, 324 collectivization, 56-57, 72, 90, 165, 179, 295, 320, 324, 339 colonialism, 179, 322

Comfortable Housing Project, 204-6, 267, 325 commerce, 50, 52, 76, 240-41, 253, 267, 279, 310-11, 316-17, 326, 339, 345, 374 commodification (or commoditization), 14, 192, 194, 293, 308, 32224, 327-28 commodity wars, 338 communism, xxxv, 364 Communist Party of China, xxix, 11, 359 competition, with migrants, 1, 130, 237, 266-69, 277, 284, 345 conflict, interethnic, 35, 283, 354 conflict, Tibetan-Muslim, 13, 76, 335, 337, 347-57, 363, 365 conspicuous consumption, 292, 318 construction, 73, 139-43, 147-48, 176, 180-81, 204-6, 211-14, 266-67; companies, 311, 313-14; work, 313, 316 consumer price index (CPI), 71, 295 consumer price inflation. See inflation, price contraception, 93 corruption, 1, 382 cost of living, 296, 329n8 counterinsurgency, 57-58, 88, 320 CPC. See Communist Party of China cultural genocide, xxxv-xxxvi, 84, 380 cultural networks, 272 cultural norms and preferences, 308-9, 321 Cultural Revolution, xxxiv, 63-65, 99, 165, 320, 357, 386 culturalist, 318-9 Dalai Lama, Fourteenth, xxii, 5-7, 23, 43n7, 51, 57, 75, 83, 94, 279-80, 289n50, 316, 360, 364 Dalai Lama, Thirteenth, 16, 56 death rates, crude, 88-91, 118, 122n16 Dechen (in Yunnan), 95, 390 de-collectivization, 67-68, 72, 179, 322 deficit, local government fiscal, 148-56 deficits, sustainability of, 154-55



Index 413

de-industrialization, 336, 337, 341-45 delineation of Tibet, 93-95 Democratic Reforms, 57, 320 demographic perspective, 33, 193 demographic transition (or population transition), xxxi, 14, 48, 83, 85-86, 88-94, 118-20, 242n1 demonstrations. See protests dependence: on China, 16, 22, 30; on external skilled labor, 378, 388; on subsidies, 24, 34, 129, 162, 267, 376; on welfare, 328, 375. dependency theory, xxxii, xln2, 18-19, 44nn12-15 desertification, 99, 123n27 developing countries, 9, 87-88, 178, 269 development: economic, 14-15; economics, structuralism. See structuralism meaning, xxxi, 13-18; peripheral economic, 17-22, 376; strategies, western, 2-4, 127, 141, 168, 171, 181-84, 218, 282, 388 studies, xxxii, 25, 28, 44n12, 248, 293, 317, 373 dignity, 283, 293, 309, 311-12, 316, 326, 346, 355, 363-64, 366 disadvantages, linguistic, cultural and educational, 35, 239, 249-52, 268, 269-70, 272, 277-78, 281-82, 377, 381, 387, 391 discipline, xxxi, 209, 312-314, 328, 332n38, 341-42 discrimination, 25-29, 130, 247, 250, 253, 279, 282-83, 311, 315, 348, 377, 389-91; effective, 182-84, 271, 283, 373-74, 380; positive. See affirmative action; in state-sector (public) employment or contracting, 271, 281, 314, 389; and urban bias, 86, 120. disembedding, 31-33, 192, 328, 335 dislocation, 35, 283, 386-87 dispossession, 192, 308-9, 325 double movement, 32, 283, 294, 32729, 335-37, 365-66, 376, 379 Drepung Monastery, 61, 361 economic analysis, macro-structural, 36-37

economic growth, rapid, 5, 132-34, 180, 191, 252, 336, 389 economic lagging, 3, 133-38 economic rebalancing, 4, 154 education, 193, 317, 322, 388-89; bilingual, xxxviii-xxxix, 251, 276, 381, 386-91, 392n4; campaigns, 254-55, 264, 286n14; Chinesemedium, xxxviii, 251, 266, 278, 380, 386-91; costs of providing, 161-62; infrastructure and inputs, 158-61; investment in, 171-72, 318; levels. See schooling levels; provisioning, 253; reforms. See schooling reforms; system, minority, 186n18, 276-78, 279, 282, 288-89n39-42, 289n51, 378-79, 385, 391; tertiary, 253-54, 262-64, 270, 277, 378, 388; as tertiary national accounting category, 144; Tibetan-medium, xxx, xxxviiixxxix, 266, 276, 278, 381, 387-91; vocational, 253, 276, 332n37. See also schools, vocational. 8-7 Poverty Reduction Plan, 3, 133 elasticity of growth to subsidies, 162-65 empire, Mongolian, 50 employment: cadre, 7, 207, 233-35, 271; construction, 204-6, 267, 269; in education, 207; informal, 201, 207, 225-26, 245n30; low wage, 292, 309, 316, 323-24, 327; manufacturing, 204-5, 243n14; mining and quarrying, 204-5, 243n14; offfarm, 192, 197, 198-203, 292, 306, 315, 343; primary sector, 196-98, 242n4; public sector, 7, 207, 239, 250, 289n44, 337, 345, 378-79, 382, 384-85, 389; registered urban, 221, 225, 233; rural, 198-202, 241, 243n14, 267-68, 298; secondary sector, 203-6; in state-owned units, 226-28, 245n32, 271; tertiary sector, 203, 206-7, 211; urban, 34, 201, 225-28, 241, 251, 282, 381, 387; urban unit, 186n20, 224-28, 232-35, 244-45n25, 384; wage, 309, 322.

414

Index

employment behavior, 292-93, 312, 317-18 employment expectations, 292, 308 employment policy and strategies, 279, 372 employment reforms, 30-31, 125n42, 192, 249, 270, 272-81, 288n36, 336, 345, 382 employment registration status, 197, 201 employment shares, 117, 242n4 employment system, 42, 125n42, 249, 282, 378, 384-85 employment transition. See labor transitions enrolments. See school enrolments enterprise reforms, 67 entrepreneurship, 24, 34, 178, 198, 267, 321, 326, 339, 348, 377, 383 environmental concerns, degradation, and protection, 4, 17, 90, 99, 303, 320, 325-26, 328 ethnic bias, 85, 86, 119, 183 ethnic niches, 269, 278, 337, 339, 365 ethnic status, 28, 56 ethnic status switching, 106-07, 1245n37 ethnic stratification (or segregation), 120, 195, 342, 372, 379 exams (in public employment competitions), 272, 287n28 exclusion (socio-economic), 25, 130, 247-9, 270, 283, 329, 373-6, 380, 382-5, 389; ethnic, 30, 120, 247. exclusionary: growth, 29, 373; processes, public employment or upper labor strata, 250, 269-81, 283, 291, 373-74, 382-85; processes, urban, 34, 86, 98, 193-94, 235, 248, 251-52, 266-68, 281-84, 355, 379, 389. expenditure, government, 2, 128, 148-54, 157-62, 174, 186n17; for capital construction, 152-53, 157-58; on education, 157-62, 177; on government administration, 157-58. See also government

administration; per person, 158-59; on security, 185n8. Falun Gong, 364 family planning policies, 84, 90-93, 115, 118, 123n22 famine of the Great Leap Forward. See Great Leap Forward farmers and herders, 308-9, 316, 328, 374 farming and herding, 12, 197, 283, 291, 337, 339, 345 farming areas, 94, 259, 277, 299, 301-2, 306, 309-11, 313, 320, 339 farming, commercial forms of, 99, 318 fertility rates. See total fertility rates fertility transition, 88-93, 122n17 feudal society, 59, 85 feuding, 317 fieldwork methods and locations, 40-41 finance (as tertiary national accounting category), 147 financial (or economic) crisis, global, 149, 154, 168 financial sector, in old Tibet. See monastic financial system fiscal decentralization, 4, 67-68 Five-Year Plan, 16 Five-Year Plan, Eleventh, 3, 62, 100, 151, 164, 204, 238 Five-Year Plan, Ninth, 3, 133, 151, 164 Five-Year Plan, Seventh, 133 Five-Year Plan, Tenth, 3, 133, 151, 164 Five-Year Plan, Twelfth, 3-4, 148, 164 fluency, Chinese, 30, 129, 251, 253, 266-69, 373, 377, 387 food security, 305 forced abortions and sterilizations, 84, 91-93, 123n22-23 foreign-funded urban units, 171 forestry, 1998 moratorium on, 105 freedom, 292, 303, 307, 317-18, 321, 323 Furtado, Celso, 14-15, 18-20, 25, 44n15 Ganden Phodrang government in Lhasa, 52, 61, 64



Index 415

Gansu, 69, 109-110, 130-31 Gansu Province, historic, 53, 78n13, GDP growth rates, 175 GDP per capita, real trends, 69-71, 133-138 gender ideologies, 383 gendered work expectations or division of labor, 312, 316-17 geopolitics, xxix, 179 ghettoization, 326 Golok, 52, 55, 57, 93, 277, 302, 31011, 349-50, 353, 356-57, 360 governance, minority area, 26, 28-29, 35, 274, 321, 381-82 government administration, 144-48, 180-81, 212, 376; employment in, 207, 228; investment in, 171-72; See also expenditure, government, on government administration. government and party agencies. See government administration graduates from high school or university, 275, 277-79, 326, 345, 374 grasslands. See rangelands Great Leap Forward, 58, 79n34, 88, 99, 320, 333n56 grievances, 9, 31, 250, 291, 343, 372, 374, 382 gross capital formation, 166-68, 174, 187-88n26 growth, exclusionary. See exclusionary growth Guangdong, 102, 109-110 guaranteed public employment for graduates, 270, 274-76, 326, 378, 381 Guizhou, 70, 102, 109-11 Haidong District, 53, 55, 94, 96, 114 Han chauvinism, 26, 48, 178 Han Chinese: headcount in the TAR, 112; patrons, 357; in Qing Empire, 51. hardship wage allowances, 12, 230-31, 246n34 harmonious society, 5, 390 health care, 88, 118, 143, 322; investment and/or spending in, 171.

hegemony, 74-75, 336, 377 herding. See farming and herding herds, 302-3, 328 hierarchy, social and ethnic, 316-17, 335, 346, 355, 363, 365-66, 375-76 high school educated (Tibetan or other), 249, 270, 277, 285n8 high school, inland, 112, 125n42 higher education, 157, 161-62 Hill, Polly, 294, 304, 323 hotels and catering (as national accounting or employment category), 144, 147, 207, 214, 229, 236-37 household production (rural), 306, 322 housing construction projects, 147, 318 Hualong, 55, 79n29, Hu Jintao, 23, 274 Hu Yaobang, 274 Hui, xx, 106, 116, 342 human resources, 103, 319, 388 human rights, xxxv, 45n22 hydroelectric power generation (or dams), 114, 116, 187n25, 214, 244n16, 343-44 illiteracy, 86, 119, 159-60, 252, 253-60, 273, 286n12, 292, 321, 371-72, 390 illiteracy rates, 258-59 incentive structures, institutional, 12, 24, 38, 99-100, 182-83, 209, 236, 250, 306, 354, 377 income concentration, 233 income distribution, household, 194, 215-17, 371, 375 income, non-farm household, 305 incomes: average household, 5, 248, 292; per capita rural net household, 129, 156, 215-16, 294-96, 298-99, 318, 327, 339; per capita urban disposable household, 218-21, 228-29, 235, 245-46n33, 372. income survey data, household, 194, 225 incorporation into PRC, 9, 48, 77 India, 87, 93, 123n23, 279, 382 individual household responsibility system, 72, 305, 320, 322

416

Index

industrial (or economic) restructuring, 141-44, 213-14, 326-27, 342, 345-46 industrialization, 14-19, 138, 140, 322 industries, processing, 336, 338-42, 344 industry (as subsector of secondary sector), 138-42 inefficiency, economic, 129, 173, 385 inequality, 1, 11, 28, 194-5, 215, 238, 244n17, 247, 270, 274, 372, 375, 383-4, 390; and conflict, 383; educational, 247-48, 252, 266, 273, 277-78, 281, 386, 388; horizontal, 249, 384; interethnic, 6, 183, 37980; intra-urban income, 119, 195, 215-17, 217-38, 372; land assets, 305-6; richest and poorest deciles and quintiles, 217, 222-24, 244n23; rural, 296, 306, 331n27; spatial, 118-19, 130, 183, 380; urbanrural, 6, 85, 195, 215-22, 230, 232, 244n19-21, 252, 372; urban-rural and discrimination, 26; urban wage, 194, 217, 228-37, 239, 274. inflation, price, 80-81n50, 132, 155, 184n5, 295 infrastructure, 48, 311, 345, 376 Inner Mongolia, 56, 84, 86, 99, 120, 136, 299 insecurity of livelihoods or employment strategies, 275, 283, 346 institutionalist approaches in social sciences, 38, 250, 284-85n1 instrumentalism, 355, 363 insurgency. See uprising integration into China through religious networks, 357 interdisciplinary methodology, 36, 249 interethnic relations, 50-51, 335-66 intermarriage, 53 investment: foreign, 20; subsidized, 163-64, 172. investment in fixed assets, 128-29, 15253, 165-72, 187-8n26; composition, 168-69; ownership, 170-71, 188n27;

rural-urban distribution, 169-70; sectors, 171-72; source of funds, 166; trends, 167-68. Islam, 355 Japanese invasion of China, 55 Jigme Phuntsog, Khenpo, 347, 362, 367n21 job assignment system. See guaranteed public employment Jonang order of Tibetan Buddhism, 360, 370n49 Kathmandu, 314 Kardze Tibetan Autonomous Prefecture, 96, 302 Karmapa, seventeenth, 361, 370n52 Kashmir, 51, 355 Kham, xxi, 50, 52 Kumbum Monastery, 53 labor: low skilled, 267, 269, 292, 315, 327; manual wage, 28, 309, 314, 322; skilled and/or educated, 267, 388. labor hierarchies, 35, 268, 279, 291, 294, 321-22, 324, 335, 345, 373-74, 377 labor market reforms. See employment reforms labor market regulation (socialist legacy), 281 labor migration, 34, 192-93, 200, 322 labor productivity, xxxi, 14-18, 208-10, 243n15, 318, 321-22 labor supply, 307 labor (or employment) transitions, 1718, 30, 33, 191-93, 195, 196-207, 206, 215, 326, 371, 381 Labrang (Xiahe), 53 Ladakh, 51, 63, 347 lama, xxi, 311, 352-53, 358, 361, 364 lama-patron relationship and networks, 357, 359-62, 364, 369n46 land colonization, 99, 123n27-28 land tenure, 306, 320 landholding sizes, 301-3, 305-6



Index 417

language, as tool of repression and resistance, xxxvi-xxxvii, 380 language (or linguistic) requirements in public employment, minority, 38182, 387, 389-90 Lanzhou, 4, 105, 368n32 layoffs, 273, 279, 345 Lewis, Arthur, 19, 45n16, 294, 307-8 Lhasa, 96-97, 104, 117, 145, 197, 236, 242, 257, 259, 269, 278, 313-14, 349; British invasion of, 51. liberalization: economic, 338-40; religious practice, 357. life expectancy at birth, 16, 47, 84, 8788, 122n16, 371 Linxia (Hezhou), 53-54, 313 literacy, Chinese, 253, 266, 389 literacy, Tibetan, 273, 389-90 livelihood strategies, 198, 292-94, 31517, 321, 323, 326-27, 373 livelihoods, 21, 181, 322, 383, 389 livestock, 299-300 livestock limits, 304 living standards, 8, 21, 239-40, 283, 291, 336, 371, 390 loans, 61-63, 147, 166, 169, 187n25, 333n44, 354 Ma Bufang, 55, 58, 79n31, 94, 350, 356 Malho (Ch, Huangnan) Prefecture, 114-17 Manchu, 48, 51, 54 manufacturing, 138, 171, 204-5, 344; artisanal, 214. Mao Zedung, 43n7 Maoism, 11, 32, 99-100, 242n1, 357, 363; radicalization of, 63-64, 74, 90, 359. Maoist industrialization strategy, 65-67, 141, 165, 178, 205 marginalization, xxxv-xxxvi, 9-11, 25-26, 35, 247, 322, 324, 353-54, 380-81; and assimilation, 39091; through integration, 291-92; intersectoral, 211-13; in peripheral economies, 18-22, 25, 376; and poverty and inequality, 10, 25, 28-29,

127-28, 215-17, 371-73, 375; from urban employment, 30-31, 85-86, 119-20, 240, 251, 387. market incentives, 317 market integration, 293 martial law, 11, 23, 274 Marxist theories of ethnicity, 59 meat and skins, 69, 297-98, 339 method, comparative quantitative, 13031, 194 middle-class, 6, 289n46, 376, 383; Tibetan urban, 31, 230, 234, 271. middle strata of labor hierarchy, 283, 345, 374, 376, 382 migrant labor, 209, 212, 217 migrants: Chinese Muslim, 12, 114, 315; Han Chinese, 7, 12, 183, 207, 263, 284, 311-12, 313-15, 348, 350, 373; out-of-province, 197, 249, 268, 285n9; rural, 197-98, 283, 374; skilled, 102, 126n44; spontaneous, 235-6, 246n39, 252; temporary, 101-2, 105, 112, 125n41, 184n1, 223-24, 236, 241-42, 263, 310. migration, 98-113, 192-93, 238-39, 252, 282-83, 322, 383; Han Chinese, xxxvi, 104; and inequality, 235-37; and livelihood strategies, 292; into Tibet from other parts of China, 83, 97, 112, 211, 214, 252, 267-68, 378-79, 388; out (of Tibet or other western regions), 102, 105, 112, 116, 126n44-45, 202, 279; return, 115; rural-rural, 198; seasonal, 104, 112, 197-98; temporary, 104, 197-98. military personnel, 104, 117, 124n31 military and security presence, 145-47, 321 milk, 297-98, 304 Ming Empire, 54 mining, 113, 129, 138-42, 144, 171, 177, 181-82, 204-5 minority education (or schooling) system. See education system, minority minzu. See nationality

418

Index

mobility, social, 284, 291, 345, 376, 378, 383-84 mobility, spatial, 200 modernity, 48-49 modernization, 59 modernization theory, 60, 179, 317-19, 388 monastery construction, 311 monastic financial system (in old Tibet), 61-63 monastic networks, 52, 61 monetization, 32, 194 Mongolians, 53, 99, 106, 116 Mongols, 48, 51 Mongours, 53, 116 mortality rates, 58, 88, 93, 122n16, 371 multiplier effect of subsidies on growth, 162-65, 177 Muslim cook myth, 351-54 Muslim militancy, 54 Muslim warlords, 52, 56 Muslims. See Chinese Muslims mutton, 304 Myrdal, Gunnar, 22 nation-state consolidation, PRC, 9 nation-state, emergence of modern Chinese, 55 nationalism: Chinese, 363; minority, 365, 379, 382; religious. See religious nationalism; Tibetan, 9, 25051, 284, 289n49, 354, 357, 365-65, 375, 379. Nationalists (Chinese Republican), 52, 359, 363 nationality (Ch. minzu, as in ethnicity), 101, 381 natural population increase rate, 89-91, 101-2, 105-13, 197 natural resource extraction, 132 neoclassical economics, 44n10 neoliberalism, 25, 179 Nepal, 51 Ngawa Tibetan Qiang Autonomous Prefecture, 96, 311-12, 348-49, 351 Ngawa traders and businesses, 310, 348, 354

niche markets, 341 Ningxia, 69, 109-11, 299 non-governmental organization (NGO), 312, 341 nuclear weapons, 66 occupation, xxix, xxx, xxxiv-xxxv, 9, 16, 24, 35, 49, 51, 57, 61, 63-64, 74, 83, 281 ODA. See overseas development assistance one child demographic profile, 105 one child policy, 90-92 Open the West Campaign, 3, 43n4, 116, 133, 136, 141, 145, 151, 201, 203-4, 221, 230, 238, 271, 274, 281 opportunity cost of labor, 307 overseas development assistance. See aid, international OWC. See Open the West Campaign ownership, 45n18, 128, 173, 179, 181 ownership, external, 20, 130, 173, 180, 267, 380 Panchen Lama, xxii Panchen Lama, Eleventh, 75, 81n56, 365 Panchen Lama, Ninth, 56-57 Panchen Lama, Tenth, 36, 53, 251, 280, 289n49, 381 Party Secretary of the TAR, 23 pastoralism and pastoralists, 292, 299, 301-4, 306, 318, 320, 330n14, 330n22 pastoral areas, 309-11, 317, 325-26, 339-40, 349 paternalism, 179, 326 patriotic education campaigns, 6, 10, 13, 32, 75, 380 peasant, 293-94, 317, 322, 329n2 “people-first” development approach, 3, 5 People’s Liberation Army, 16, 51, 65, 356 PLA. See People’s Liberation Army Polanyi, Karl, 21, 32, 192, 283, 294, 309, 321, 324, 327, 376



Index 419

polarization, 18-22, 195, 222, 237-40, 251, 283, 343; class, 284; of education outcomes, 262; intersectoral, 208-15. policy analysis, 128 political disempowerment, 120, 129, 273, 320, 322, 375-77, 379 political representation, 282, 378, 391 political repression, 5, 10, 250 politics of statistical representation, 59, 371 population: Han, 84, 111, 124n41; Tibetan, 87, 121n11. population density, 301 population distribution, urban-rural, 96-98 population growth: rapid, 69, 84, 8788, 100, 118, 343; rural, 105-6, 197, 200, 298. population growth rates, 106-113 population momentum, 89-90, 98, 105, 197 population shares, 85-86, 98, 106-7, 114-17, 125n44 population surveys, 101-3, 107, 240-42, 254-55, 260-62 population swamping (or invasion or transfer), 83-85, 98-100, 114, 117, 120, 315, 379 post-colonialism, 9 Post Washington Consensus, 45n18, 178 poverty, 6-7, 10, 28, 283, 292-94, 310, 319-20, 323, 326, 328, 330n24, 375, 383, 388, 390; absolute income, 43n5, 60, 194, 215, 239, 244n17, 247-48, 250, 296, 371, 375, 377; feminization of, 383; relative, 28, 248, 373, 375. poverty alleviation strategies, 3, 17, 325-26, 328, 383 poverty rates: rural, 5, 71, 194, 215, 296, 330-31n27, 371; urban, 221, 248. power relations, 391-92 preferential policies for minorities, 106, 251, 282, 378, 381-85, 389-90;

employment (or contracting) policies, 227, 239-40, 251, 282, 314, 378, 381-85, 389-90. prejudice, 26-28, 347, 355 price-scissor strategy of rural-urban development, 66-67, 320 prices, grain and wool, 60, 66, 68-69, 197, 298, 321, 337-38, 340 primary sector, 138, 140; definition of, 45-46n24. primordialism, 76-77, 354-55, 368n26 private enterprise: rural, 198; urban, 170-71, 207, 224-27, 279. privatization, 339, 341 productivity, agricultural, 296-98, 3068, 331-32n31-3 protectionism, local, 338 protests: 1950s uprisings in Kham and Amdo. See uprisings, 1950s Kham and Amdo; 1987-89 in Lhasa, 7374, 274, 329n6; in 2000s regarding employment policy, 280-81; 2008, 1, 8, 30, 34, 35, 75, 83, 104, 112, 129-30, 193-95, 235, 239, 250, 275, 357, 380; self-immolations. See selfimmolations. public contracting, 381 public management and social organizations. See government administration Public Security Department, 101, 124n33, 125n41, 312 purchasing power, 17, 340 Qing Empire, 22, 51-52, 54-56 Qinghai, 130-31 Qinghai Province, creation of, 55 Qinghai Province, net migration rates of, 111-13 quality (as a cultural concept; Ch. suzhi), 273, 333n49, 388 racism. See prejudice railway, Qinghai-Tibet or Golmud-Lhasa, 104, 139-40, 143, 145-48, 153, 171, 174-75, 180, 188n28, 204, 206, 236, 271, 310

420

Index

ranching, 99, 318, 321 rangeland, 297, 302, rangeland conflicts, 301, 366 rangeland degradation, 325 rangeland management, 320 rapeseed, 69, 297 rationality, 317-18, 321 rationalization of economic and employment policy, 233-34, 272, 281-82, 283, 344, 377-78 Rebgong, 53, 114, 343-44 recession, 3, 71 reform period, 32, 363, 365, 371, 380 refugee flows to India, 79n33, 90, 100 regional development strategy: Maoist, 66, 129, 148, 152, 164-65, 178, 320, 339; reform period, 67-68, 252, 295. regional redistribution, 2, 21, 22 religious elites (or leadership), 354-56, 357-58, 362, 364-66 religious nationalism, 6, 76, 364-66 religious practice, 5, 362-63 religious revival, 75, 357-59, 365-66 resettlements, 32, 34, 192-94, 197, 294, 325, 328, 333n57, 343, 374-75, 380 residency status, 101-2, 107, 124n3233, 200 residents: permanent, 101, 112, 121n4, 125n41, 222, 229, 233, 241-42; temporary, 101, 125n41, 223-24, 226, 241-42, 264, 310; urban, 268, 273. resistance against state control or subordination, 291, 294, 309, 328 restaurant, 332n36; Muslim, 349-51, 353. retrenchment from state-sector employment, 223, 234, 239, 280, 374, 384 revenue, local government, 146-49 revolution, 320 rimpoche, xxi riot, 347, 356 riots, Lhasa, 1, 7-8 risk insurance (role of rural assets), 303 Rongwo (Town and Monastery), 53, 115, 354, 360

rural economic organization (Maoist and reform), 306 rural economy, 24 Salar Muslims, xx, 53, 58, 106, 114-17, 302, 306-7, 313, 341-42 salaries, 23, 221, 229, 245-46n33; public sector, 146; See also wages. savings, rural, 303 school enrolments, 39, 249, 256, 282, 286n11, 286n14, 289n51, 378, 389 schooling: quality of, 254, 256; resistance toward, 286n11, 384-85. schooling levels (or rates), 93, 183, 195, 247, 252, 253-54, 260-67, 273, 281, 285n10, 286n12, 292, 310, 343, 345, 371-72, 378, 385, 390; city-rural differences, 265-66; and exclusion, 383; among nationalities in China, 286-87n17. schooling reforms (consolidations or regarding language of instruction), 32, 158, 192, 256, 276, 343, 380, 386 schooling system, 249, 385 schools: mainstream Chinese, 278, 288n40; per person, 158-62; primary, 253, 386; rural, 256, 286n14, 386; secondary, 159-61, 253-54, 261-62; vocational, 159-61. See also education, vocational. Scott, James, 291, 293-94, 323 secondary sector, 139-43 secularism, 76 self-employed individuals: rural, 198; urban, 171, 224-27, 279, 284, 346. self-immolations, 6-7, 13, 32, 75, 185n8, 328, 361, 380, 387 serfdom, 47, 59 Serthar, 75, 347, 362 17-Point Agreement, 22 sex workers, 105-6, 236, 312, 371 Sichuan, 109-10, 130-31, 218 Shanghai, 109-10, 139, 221, 222, 23133, 264, 360 Shigatse, 40, 51, 69, 95, 177, 195-96, 257, 259



Index 421

Sino-Indian War of 1962, 63 skills formation, 268 social instability, 251, 281 social services (as amalgamated tertiary national accounting category), 14446 social status, 309 socio-economic structure, 29 SOE. See state-owned enterprise Soviet Union, 24, 242n1 spatiality, 119 staff and workers (urban), 147, 161, 186n20, 221-22, 224-28, 230-35, 244n24, 271, 274, 345, 374, 384. See also employment, urban units standardization of public sector employment, 272, 281-82, 377-78 state-owned enterprise (or units), 14, 170, 188n27, 205, 226-27, 271 state-sector employees (or employment), 12, 195, 218-22, 264, 270, 272, 274, 315, 381, 384 steel and aluminum production, 113, 213-14, 243-44n16, 344 structural disjunctures, 266-68, 281, 388-89 structural economic change (or transformations), 128, 138-41, 180, 383 structuralism (in development economics), xxxii, 18-22, 25, 250 structuralism (in sociology), 38, 250 structuralist development economics method, 37-39 subordination, 10-11, 25, 249, 269, 293, 377-79, 383, 390-91 subsidies: direct budgetary, 2-3, 12829, 148-55, 172, 178, 180; efficiency of, 162-65, 171; equity of, 155-62; for state-owned enterprises, removal of, 340-41; indirect, 148, 152-53; per person, 2, 155-56, 159; total (combined direct and indirect), 172-77. subsidization, 4, 29, 48, 64, 66, 100, 103, 111, 129, 138, 156, 180-81, 236, 380; and polarization, 214, 379-80.

subsidization policy (or strategies), 12, 30, 141, 152, 237, 239-40, 372, 385 subsistence (modes of production or livelihoods), 34, 60, 293, 298, 3045, 323, 327, 389 subsistence capacity, 293, 305-9, 31718, 321, 323-24, 326-28, 373 subsistence needs, 306-7, 324, 33031n27, 375 subsistence strategies, 291, 323, 326-27 Sufism, 54 Sunkel, Osvaldo, 19-22, 44n15 sustainability (of economic model and related transformations), 152 tax reform, 4, 133, 150, 152 taxation, xxxixn1, 150-51 taxi business, 269 teachers, 270; per person, 158-62. technology transfer, 269 terms of trade for agricultural goods, 17, 68-69, 133, 308, 325, 337-38 tertiary (service) sector, 15, 132, 138, 140, 143-48, 184n6, 212, 236-37, 266-67 textile factories, 115, 314, 341-42, 344 TGIE. See Tibet Government in Exile Third Front Strategy. See Maoist industrialization strategy Three Rivers Source area (Ch. sanjiangyuan), 325 Tibet Autonomous Region, net migration rates of, 109-13. Tibet Government in Exile, xxxiv, 95, 122n11, 346 Tibet Support Groups, xxxiv, 83, 95 Tibet Work Forum, Fifth, 4, 25, 129, 131, 148 Tibet Work Forum, Fourth, 3, 144 Tibet Work Forum, Second, 72-73 Tibet Work Forum, Third, 3, 73, 133 Tibetan Buddhism. See Buddhism, Tibetan Tibetan elites (and intellectual community), 352-57, 378 Tibetan Empire, 50-52

422

Index

Tibetan exile community in India and Nepal, 52, 314, 346, 361 Tibetan medicine industry, 182 Tibetan Muslims, 51, 63 Tibetan representation in public (or cadre) employment, 233-35, 238-39, 271, 274, 384 Tibetan uprising of 1959, 57, 88 Tibetanization, 251, 274, 276, 381, 385 total fertility rates, 88-93, 105, 118, 122-23n17, 195 tourism, 104, 113, 129, 144, 146-47, 176-77, 181-82, 185n9, 267, 345, 361, 376 township and village enterprises, 68, 198-203, 241, 243n13 trade (as national accounting and employment category), 144, 147, 214, 229, 236-37 trade deficit: domestic (with rest of China), 173, 175-77, 178, 180, 377; international, 168, 173-75, 180, 188n29. trade, Muslim, 54, 311 trade, routes with India, 63 transfer payments: regional, 21, 148-51; welfare (to households), 215, 279, 298-99. transport (as national accounting and employment category), 144, 147, 212, 214 trickle down, 129 truck driving, 332n35 trulku, xxi, xxiii, 364 trulku networks and lineages, 52, 357, 360-61 Tsaidam Basin, 95, 115 See also World Bank “Qinghai Project” tsampa (roasted barley flour), 333n51 Tsering, Tashi, 314-15 TSG. See Tibet Support Group tuition fees, 277-78, 286n14, 322 TVEs. See township and village enterprises unemployment, 193, 235, 249, 269, 278-79, 284, 292, 327, 374 university, 277

university educated (Tibetans or other), 249, 270, 277, 384 uprising: 1959, Lhasa and Central Tibet. See Tibetan uprising of 1959; 1950s, Kham and Amdo, 8, 64, 88, 250, 320. urban areas, 117 urban bias, 6-7, 119 urban employment exclusion. See exclusionary processes, urban urban labor markets, 249, 260, 276, 388 urbanization, xxxi, 48, 83, 93, 98, 119, 140, 173, 176, 238-39, 252-53; and employment transition, 193, 198203, 268, 343; and exclusion, 383; and illiteracy, 258; and inequality, 217, 230; and migration, 322, 326, 345; as poverty reduction strategy, 325; and protests, 250 urbanization rates, 96-97, 105, 117, 243n11 US foreign policy, xxxv Utsang, xxiii Uyghur, 191 value-added, 43-44n9, 208-10 vegetable farmers, 197 vocational education. See education, vocational wage concentration, 233 wage (and related work) expectations, 252, 292, 307-9, 313, 322, 342. See also employment expectations wage policy, 12, 230-32, 239, 372 wage rates, 305, 307, 332n37, 341 wages, average (of staff and workers or of urban unit employment), 221, 224-25, 228-35, 245n27, 372; in education sector, 158-59, 161-62. wages and salaries as source of rural incomes, 296, 298, 305, 318 wealth flows, 377-78 Wenchuan earthquake in Sichuan in 2008, 96, 151, 168, 183 western development strategies. See development strategies, western or regional



Index 423

Western Development Strategy. See Open the West Campaign wheat, 69 wool, 60, 68-69, 80n44, 337-42 wool wars. See commodity wars working class, 273, 309, 316 World Bank “Qinghai Project”. See Tsaidam Basin

Xi’an, 4 Xining, 53-54, 94, 105, 368n32 Xinjiang, 10, 23, 25, 28, 75, 84, 86, 92, 102-03, 109-12, 120, 191; Han migration, 98-99; rural assets, 299; urban-rural inequality, 218. Xunhua Salar Autonomous County, 5355, 114-17, 302, 306, 313, 341-42

xibu da kaifa. See Open the West Campaign

yak, 303 Yushu, 93, 277, 311, 349

About the Author

Andrew M. Fischer is associate professor of social policy and development studies at the Institute of Social Studies (ISS) in The Hague, Netherlands (part of Erasmus University Rotterdam since 2009). He is also convener of the MA major in Social Policy for Development at ISS, in which he leads teaching on poverty studies, population, social policy, inclusive development, and development economics. A development economist by training and an interdisciplinary social scientist by conviction, he has been involved in researching or working in development studies or in developing countries for over twenty-five years and on China’s regional development strategies in the Tibetan areas of Western China for more than twelve years. Beyond his work on Tibet, he deals more generally with inclusive development strategies to combat poverty, inequality, and social exclusion. A major focus of his current research is centred on the imperative and challenges of redistribution in contemporary development. He has published widely on subjects related to China’s development and on the international development agenda more generally. He earned a PhD in Development Studies from the London School of Economics and an MA in Economics from McGill University in Montreal, Canada.

425