Political Economy of Reform in China (Kobe University Social Science Research Series) 9811902011, 9789811902017

Citation preview

SPRINGER BRIEFS IN ECONOMICS KOBE UNIVERSIT Y SOCIAL SCIENCE RESEARCH SERIES

Kai Kajitani Tomoki Kamo

Political Economy of Reform in China

SpringerBriefs in Economics

Kobe University Social Science Research Series Series Editors Yunfang Hu, Kobe University Graduate School of Economics, Kobe, Japan Shigeyuki Hamori, Kobe University Graduate School of Economics, Kobe, Japan Editorial Board Masahiro Enomoto, Kobe University RIEB, Kobe, Japan Yoshihide Fujioka, Kobe University Graduate School of Economics, Kobe, Japan Yuka Kaneko, Kobe University Center for Social Systems Innovation, Kobe, Hyogo, Japan Kazumi Suzuki, Kobe University Graduate School of Business Administration, Kobe, Japan Kenji Yamamoto, Kobe University Graduate School of Law, Kobe, Japan

The Kobe University Social Science Research Series has been established as a subseries of the SpringerBrief in Economics Series, but in fact this exciting interdisciplinary collection encompasses scholarly research not only in the economics but also in law, political science, business and management, accounting, international relations, and other subdisciplines within the social sciences. As a national university with a special strength in the social sciences, Kobe University actively promotes interdisciplinary research. This series is not limited only to research emerging from Kobe University’s faculties of social sciences but also welcomes cross-disciplinary research that integrates studies in the arts and sciences. Kobe University, founded in 1902, is the second oldest national higher education institution for commerce in Japan and is now a preeminent institution for social science research and education in the country. Currently, the social sciences section includes four faculties — Law, Economics, Business Administration, and International Cooperation Studies — and the Research Institute for Economics and Business Administration (RIEB). There are some 230-plus researchers who belong to these faculties and conduct joint research through the Center for Social Systems Innovation and the Organization for Advanced and Integrated Research, Kobe University. This book series comprises academic works by researchers in the social sciences at Kobe University as well as their collaborators at affiliated institutions, Kobe University alumni and their colleagues, and renowned scholars from around the world who have worked with academic staff at Kobe University. Although traditionally the research of Japanese scholars has been publicized mainly in the Japanese language, Kobe University strives to promote publication and dissemination of works in English in order to further contribute to the global academic community.

More information about this subseries at https://link.springer.com/bookseries/15423

Kai Kajitani · Tomoki Kamo

Political Economy of Reform in China

Kai Kajitani Graduate School of Economics Kobe University Kobe, Hyogo, Japan

Tomoki Kamo Keio University Fujisawa-shi, Kanagawa, Japan

ISSN 2191-5504 ISSN 2191-5512 (electronic) SpringerBriefs in Economics ISSN 2520-1697 ISSN 2520-1700 (electronic) Kobe University Social Science Research Series ISBN 978-981-19-0201-7 ISBN 978-981-19-0202-4 (eBook) https://doi.org/10.1007/978-981-19-0202-4 © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

This book contains four research papers that attempt to clarify the issues behind China’s rapid economic growth, using empirical studies. The book makes three distinctive points. First, it elucidates the unique economic development of China from a different perspective than the “state capitalism” theory, based on empirical research on the Chinese economy and politics with the cooperation of leading scholars. Second, the book paints a total picture of China through an interdisciplinary analysis of economics, politics, and history. Third, this book focuses on the various political–economic context of China’s rapid economic growth on the following basis: preferential treatment of state-owned enterprises (SOEs), local government intervention in business management, the promotion of innovation backed by historical institutions, and local government’s competition to attract companies through the auction of the land usage right. We have conducted the following empirical analysis in this book. First, we analyze whether a clear difference in the labor distribution rate exists between SOEs and non-SOEs, using data from industrial enterprises. Our analysis shows that wage share throughout the analysis period was much higher at SOEs than at non-SOEs. Second, we focus on Shenzhen as an innovation hub and examined sustainable innovation and its institutional context in China. And we found how firms avoid the prisoner’s dilemma in the context of the manufacturing industry in China, which faces uncertainty from the constant entry of micro-companies or traders. Third, we empirically clarify the questions by matching the databases of industrial enterprises and information on elected representatives of the Local People’s Congress of Yangzhou City, Jiangsu Province. Our analysis shows that the acquisition of political capital as a representative of the Municipal People’s Congress certainly improved the performance of local enterprises in terms of scale, but not in terms of efficiency. Finally, we focus on the competition by local governments to attract companies by auctioning land usage rights. The existing studies on the competition among local governments tended to focus on the competition of growth rate through the performance evaluation system or the tax competition for attracting companies. v

vi

Preface

This book was published as part of the Kobe University Social Science Research Series by Springer Nature. And Kobe University Center for Social Systems Innovation (KUSSI) fully supported the publication of this book. The description of the empirical research in Chap. 5 is based on a paper coauthored with Dr. Daisuke Fujii of Osaka University of Economics, who kindly agreed to allow us to use part of the co-authored paper as the content of this book. We would also like to thank Takahiro Sato of the Research Institute for Economics and Business Administration, Kobe University, for his helpful comments on the revision of each chapter. We would like to express our deepest gratitude to all those who contributed to the publication of this book. Kobe, Japan October 2021

Kai Kajitani

Contents

1 Introduction: The Future of Chinese Economy Under the U.S.– China Economic Conflict . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 The U.S.–China Economic Conflict . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Chinese Economy and "Another Capitalist Economy" . . . . . . . . . . . . 1.3 Structure of This Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Ownership Structures and Disparities in the Labor Share of Chinese Industrial Enterprises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 The Guojin-Mintui Phenomenon and the Decline of the Wage Share . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Previous Research on Disparities in Wages and Wage Share at SOEs and Non-SOEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Analysis of Wage Share Differentials Using Propensity Scores . . . . . 2.4 Analysis Methodology, Data, and Analysis Results . . . . . . . . . . . . . . . 2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Evaluating the Sustainability of Innovation in China from the Perspective of Institutional Economics . . . . . . . . . . . . . . . . . . . . 3.1 Is China’s Innovation Sustainable? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Shenzhen’s Innovation Ecosystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 The Three Groups of Companies Based on Their Consideration of Intellectual Property Rights . . . . . . . . . . . . . . 3.2.2 Ignoring Intellectual Property Rights and Vertical Disintegration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3 Protection Through Patents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.4 Innovations Through the Open-Source Concept . . . . . . . . . . . . 3.3 The Institutional Background of the Shenzhen Ecosystem . . . . . . . . . 3.3.1 Coexistence of Shanzhai and Innovations . . . . . . . . . . . . . . . . . 3.3.2 Shenzhen’s Ecosystem and Independent Design Houses . . . . 3.3.3 Solving the Prisoner’s Dilemma . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 4 6 8 9 9 14 16 18 23 24 27 27 30 30 31 32 32 33 33 34 35 vii

viii

Contents

3.4 Institutional Background of Innovation in China . . . . . . . . . . . . . . . . . 3.4.1 Brokerage Economy as the Traditional Economic Order of the Market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.2 Brokerage as a Form of an Incomplete Contract . . . . . . . . . . . 3.4.3 Nash Equilibrium and the Prisoner’s Dilemma . . . . . . . . . . . . 3.4.4 Historical Institution of the Brokerage Economy . . . . . . . . . . . 3.5 Contemporary Significance of the Brokerage Economy . . . . . . . . . . . 3.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Political Capital and Enterprise Performance in China: Empirical Analysis Using Data from the People’s Congress of Yangzhou City and Industrial Enterprise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction: Local People’s Congress as “Chinese-Style Democracy” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 The Methodology of Empirical Studies . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Robustness Check . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Land Market Competition Among Local Governments: A Spatial Analysis of Zhejiang Province . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Land and Real Estate System Reform and Price Restraint Policies . . 5.2 Land and Real Estate Markets and Regional Governments’ Fiscal Revenue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Land Market Structure and Rent-Seeking . . . . . . . . . . . . . . . . . . . . . . . 5.4 Empirical Studies on the Fiscal Competition by the Local Governments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5 Data and Empirical Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.2 Empirical Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6 Results of the Empirical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

36 36 37 38 41 42 44 45

47 47 52 57 61 65 65 67 67 72 76 80 82 82 84 85 88 89

About the Authors

Kai Kajitani is a professor at the Graduate School of Economics at Kobe University. He received his Ph.D. in economics from Kobe University in 2001 and specializes in modern Chinese economics. His research interests also include development economics and comparative economics. He was a visiting scholar at the Center for Chinese Studies, University of California, Berkeley, from September 2005 to June 2006. He is the author of Lectures on Chinese Economy, Chuou Koron Shinsha, 2018 (in Japanese); Chinese-Style Capitalism Evolving Beyond a Dual Trap: Empirical Studies on Ambiguous Institutions (ed. with Hiroyuki Kato), Minerva Shobo, 2016 (in Japanese); The Fiscal and Financial System in Modern China: The Economics of Globalization and Central–Local Relations, Nagoya Daigaku Shuppankai, 2011 (in Japanese); and other works. Tomoki Kamo is a professor at the Faculty of Policy Management, Keio University, where he received his Ph.D. in media and governance in 2004. His research interests include Chinese politics and comparative politics. He was appointed as a consul in the consulate-general of Japan in Hong Kong from October 2016 to October 2018. He was a visiting scholar at the Graduate Institute of Political Science, National Taiwan Normal University (Taiwan) from September 2010 to October 2010 and the Center of Chinese Studies, University of California, Berkeley, from March 2011 to March 2012, as well as a visiting associate professor at the College of International Affairs, National Chengchi University (Taiwan) from February 2013 to September 2013. He is the author of Political Institutions in Contemporary Chinese Politics: The Politics of Temporality and the Rule of the Chinese Communist Party (ed. with Jaehwan Lim), Tokyo: Keio University Press, 2018, (in Japanese); The Sources of China’s Foreign Policy, Tokyo: Keio University Press, 2016, (in Japanese); The Rise of China as a Major Power, Tokyo: Ichigeisya Press, 2016, (in Japanese); Contemporary Chinese Politics and People’s Congresses: Reforms of People’s Congresses and Changes in the “Guiding-Guided” Relationship, Tokyo: Keio University Press, 2006 (in Japanese); and other works.

ix

Chapter 1

Introduction: The Future of Chinese Economy Under the U.S.–China Economic Conflict

1.1 The U.S.–China Economic Conflict Undoubtedly, the most crucial factor affecting the future of the Chinese economy is the U.S.–China economic conflict, which began in March 2018. As a result, there is pessimism that the rivalry between the U.S. and China could become increasingly severe and prolong beyond being solely a trade war. This pessimism is, for example, typically embodied in the speech delivered by U.S. Vice President Mike Pence at the Hudson Institute, a conservative think tank, on October 4, 2018. Pence’s speech was received with shock not only because it comprehensively included the Trump administration’s tough stance on China regarding political, military, and human rights issues, in addition to trade issues, but also because his argument was completely caught up in the idea that “China is a state of a completely different nature.” Since the launch of China’s reforms and more open policies, previous U.S. administrations have promoted a policy of engagement based on the notion that Chinese economic development would help foster a middle class, which would lead to the establishment of the rule of law and democratization. However, since the second half of the Democratic Obama administration, pessimism about China’s efforts to change its political and economic systems has begun to grow. Xi Jinping announced the abolishment of presidential term limits at the National People’s Congress in March 2018, allowing his administration to last as long as he so desires. This event appears to have played a critical role in the shift in the U.S. policy toward China from engagement to deterrence. Undoubtedly, the core of this shift is the argument that “China is a state of a completely different nature.” Since then, the economic friction between the U.S. and China has dramatically become apparent. On March 22, U.S. President Donald Trump declared that the U.S. government would impose a 25% tariff on Chinese imports worth up to 60 billion USD under Article 301 of the U.S. Trade Act. This article allows the U.S. to impose unilateral sanctions on countries engaging in unfair trade practices, claiming that U.S. companies suffered damages from intellectual property infringements by © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 K. Kajitani and T. Kamo, Political Economy of Reform in China, Kobe University Social Science Research Series, https://doi.org/10.1007/978-981-19-0202-4_1

1

2

1 Introduction: The Future of Chinese Economy …

the Chinese government and Chinese companies. Prior to that, the U.S. government restricted imports of steel and aluminum from China under Article 232 of the Trade Expansion Act. In response, the Chinese government immediately announced that they were considering countermeasures. As a result, concerns over an all-out trade war spread immediately. On January 15, 2020, U.S President Donald Trump and Chinese Vice Premier Liu signed a "Phase 1 trade deal." However, the conflict between the two countries was caused not only by their trade imbalance but also by the significant progress of innovation in China as represented by the number of international patent applications almost reaching the level of the U.S. Currently, the U.S. and China struggle over dominance in advanced technologies, such as Information and Communications Technology and Artificial Intelligence (the so-called “high-tech cold war”), making the problem even more difficult to solve. Behind the strong stance of the U.S. government is its wariness toward "Made in China 2025," which is an Internet and high-tech industrial policy announced by the Chinese government in 2015. When imposing sanction tariffs on Chinese products, the Trump administration accused the Chinese government of intellectual property infringement by using the following four methods: (1) forcing U.S. companies to transfer technologies under foreign investment restrictions; (2) discriminating against U.S. companies in technology transfer agreements; (3) Chinese companies acquiring advanced U.S. technology companies; and (4) the People’s Liberation Army and other Chinese entities conducting cyberattacks on U.S. companies. The U.S. government seems desperate to stall the development of China’s rising information technology industry, even if U.S. companies’ profits are temporarily depressed, to prevent Chinese companies from creating next-generation platforms. This high-tech cold war had a serious impact on the global supply chain. Taking Huawei as an example, Huawei disclosed a list of 92 key suppliers worldwide at the end of 2018, and the company purchases parts worth around $67 billion (approximately 7 trillion yuan) a year from around the world. The largest supply source is the U.S., and the procurement amount of Huawei amounts to $10 billion a year. A large portion of the U.S.–China trade in the electronics industry is accounted for by the “U.S.–U.S. trade,” where U.S. companies in China assemble core parts, such as semiconductors, and integrate circuits that are imported from the U.S. Then, they export the finished products back to the U.S. Inevitably, prolonged friction will cause a significant negative impact on industries in the U.S. and China. Prolonged trade friction will also have a considerable impact on the Japanese economy and companies. The reason is that, after China became a member of the World Trade Organization (WTO) in 2001, the pattern of trade and the division of labor in the East Asian regions have been firmly established. That is, China imports intermediate goods from Japan, Korea, and ASEAN countries and then exports finished goods to the U.S. and the European Union (EU). Comparing the structure of the world trade in 1999–2016, before and after China entered the WTO (Figs. 1.1 and 1.2), as exports of finished goods from China to the U.S. and the EU increased, the exports of finished goods from Japan to Europe and the U.S. decreased significantly, whereas exports of intermediate goods from Japan to China increased.

1.1 The U.S.–China Economic Conflict

3

Finished goods Intermediate goods

508.3 Japan

Korea

944.4

153.9 113.8

377.4

250.7 392.1

China

72.5

EU

923.6 110.7

The U.S.

ASEAN 456.9

335.5

Fig. 1.1 World trade flow centering on Asia (1999). Source White paper on International economy and trade 2011, RIETI-TID2016 (http://www.rietitid.com/)

Finished goods Intermediate goods

416.8 Japan

Korea 1195.9

733.7

898.6

597.2

China

709.0 3061.2

2568.8 1267.9

1162.7

EU

The U.S.

ASEAN 841.4

857.3

Fig. 1.2 World trade flow centering on Asia (2016). Source White paper on International economy and trade 2011, RIETI-TID2016 (http://www.rietitid.com/)

In other words, a structure has been established, where the higher the presence of Chinese products in Western markets is, the more profitable the parts industry in Japan and Korea becomes. Despite concerns about the prolonged trade war, the U.S. seems to be unwavering in its hardline stance. Satoru Mori, who is familiar with the U.S. diplomacy, said that there is a basic recognition that efforts to protect the U.S. manufacturing industry and advanced technologies should be regarded as important national security issues, and a phenomenon can be observed in which traditional economic nationalism is combined with a new techno-nationalism affecting the development of defense policy (Mori 2019). Despite many questions on the economic rationality of the U.S.–China trade

4

1 Introduction: The Future of Chinese Economy …

battle, finding clues to the solution will not be easy so long as a political background involving security issues exists.

1.2 Chinese Economy and "Another Capitalist Economy" As we saw in the previous section, the U.S. policy toward China seems to have undergone a shift from engagement to deterrence; this is casting a shadow on the economic relationship between the two countries and the future of the Chinese economy. However, these two stances have something in common in terms of a certain point of view: a dichotomous recognition that the current Western political and economic system is the sole, universal system, and other systems are recognized to exist only when they appear to be on a path of convergence with Western systems, but that systems of a different nature must not exist. In this context, Japanese scholars in China raised the question of whether the Chinese model will converge with the Western "universal model" in a gentler way to avoid that dichotomous view. Mio Kishimoto, specializing in Chinese history during the Ming and Qing dynasties, stated that the postwar paradigm of Chinese studies is roughly classified into the "typology" and the "stages of growth theory." The former focused on the differences between the Chinese society and Western and Japanese societies based on prewar critical thinking, whereas the latter positioned each period of Chinese history within the development steps modeled on the West under the influence of Marxist historical studies (Kishimoto 2013). The “typology” is not always consistent with the American argument that "China is a state of a completely different nature." That is, some typology-based arguments about China focus on, and positively evaluate, the significance of the originality of the Chinese economy. This notion cannot be grasped by the unilineal development-step theory while incorporating that the economy develops with “freedom like scattered sands” originating from the lack of system infrastructure and risks and instability. The late Hiroyuki Kato, who actively studied the Chinese economy, attributed the source of the dynamism of Chinese capitalism, which is full of uncertainty, to the ambiguity of a system that supports a market economy. In his posthumous work, “An Introduction to Chinese Economics,” he paid attention to the ethical discipline of Bao (包), which was set forth by Sukekata Kashiwa, a famous sociologist during and after World War II, as the original system supporting the Chinese economy (Kato 2016). Bao is a general Chinese term for a subcontract, which can even include "subcontracting the fulfillment of a certain service completely to a third person." In China, since the reforms and more open policies, the Bao production subcontracting system between farmers and local governments, or the financial subcontracting system between superior and subordinate governments, made a significant contribution to the advancement of the Chinese economy in later years. Kato positively evaluated ambiguous systems, represented by Bao, for their flexibility and unintended strength against the risks caused by changes in the global economy. Kato’s argument that emphasizes the originality of the modern Chinese

1.2 Chinese Economy and "Another Capitalist Economy"

5

economic system was a developed version of the traditional type theory-based Chinese argument that had taken over Chinese studies in Japan since before World War II. He adopted the approach of interpreting an orientation toward type theory, that is, the irregular phenomena observed in African and Chinese economies. Such phenomena are regarded not as “transitional phenomena in the underdeveloped stage of the market” but as something that constitutes “another capitalist economy” that greatly differs from mainstream capitalist economies in terms of types. The argument against the view that the capitalist economy will converge to the universal Western model is not limited to Chinese researchers. In Anthropology of Hand-to-Mouth Life, Ogawa Sayaka focuses on people who work in the urban informal sector,living a “hand-to-mouth life,” as the third type of people living in modern society, who are neither the “rational homoeconomics,”nor the “people left behind by modernization, bound by tradition.” Moreover, Ogawa Sayaka described "another capitalist economy" created by those people (refer to the chart below). Mainstream capitalist economy versus another capitalist economy Mainstream capitalist economy

Another capitalist economy

People playing a central role Rational homo-economics

People who lead a hand-to-mouth life

The uncertainty of the market

Small

Large

The quality of competition

Orderly competition

Rushing economy

The purpose of behavior

Maximizing benefits in the long term

Securing sufficient benefits to live

The tool of cooperation

Long-term trust

Short-term deal by intermediary

The characteristics of companies

Rational production system with an internal system of division of labor

Retaining an individual operation system

Note Drawn up by the author according to Ogawa (2016)

The books by Ogawa, whose research field is in Africa, and Kato, whose research field is in China, have several resonances beyond the differences in the geographical area and field. The reason is that their books take the approach of interpreting an orientation toward type theory. That is, the irregular phenomena observed in the African and Chinese economies are not transitional in the under-developed stage of the market but are something that constitutes “another capitalist economy” that greatly differs from the mainstream capitalist economy in terms of types. We should not forget the fact that this “another capitalist economy” is highly compatible with authoritarian political systems. One reason for this is that, in short, economic activities based on short-term and unstable business relationships prevent the formation of intermediate groups independent of political powers, such as labor unions and industrial groups. The absence of such groups creates a kind of freedom,

6

1 Introduction: The Future of Chinese Economy …

including the vigorous entry of small businesses into the market economy, which in turn weakens the autonomy and ability to resist political intervention in the market. Another reason for the high compatibility with authoritarian political systems is China’s traditional business customs. Such custom resolves the prisoner’s dilemmalike mutual distrust inherent in short-term deals through mediations performed by influential intermediaries and not by regulations and a court of law. In traditional Chinese society, influential intermediaries with great reputations of reliability and their own information networks have always maintained close relationships with official powers and supported the economic order. This system allows official powers to efficiently manage and control small businesses through mediation by influential intermediaries. The system has also been taken over for intermediary systems through the Internet in modern China by companies such as Alibaba or Tencent. From another perspective, the method of avoiding deal-related problems by using intermediaries without having to depend on legal systems has been developed at a high level. This notion shows that public confidence in legal systems in society remains low and that the rule of law is hard to establish. On the whole, highly harmonious coexistence between highly dispersive and free private economic activities is the greatest feature of the current Chinese political and economic system and will remain firmly unshakable. The activities are those created by outmaneuvering the rules established by the official powers and authoritarian political systems out of the reach of the rule of law. From another perspective, this harmonious coexistence is what fuels the harsh view of the U.S. on China’s development model. Chinese studies in Japan have accumulated a traditional approach of looking at China as "another type" based on neither the “step theory” nor the view of “China as a state of a completely different nature,” which assumes that the Western model is the one and only goal. As descendants of such tradition, we also try to adopt an approach to the research of contemporary China, from an empirical perspective. That is, we neither align with the current authoritarian Chinese administration nor follow the dichotomous American view of China.

1.3 Structure of This Book This book contains four research papers that attempted to clarify the issues behind China’s rapid economic growth using empirical methods of political economy. The papers focused on the political-economic context of China’s rapid economic growth in terms of the following: preferential treatment of state-owned enterprises (SOEs), local government intervention in business management, the promotion of innovation backed by historical institutions, and local government’s competition to attract companies through the auction of the land usage right. The specific contents of each chapter are as follows. In Chap. 2, we tried to analyze whether a clear difference in the labor distribution rate exists between SOEs and non-SOEs, using data of industrial enterprises.

1.3 Structure of This Book

7

Attributing many characteristics of the Chinese economy to its system of state capitalism has become popular. One contributing factor is that aggressive investment behavior on the part of the government and certain large enterprises has served as an engine for high growth. In that process, large SOEs appeared in some industries, drawing criticisms such as “guojin-mintui (国進民退, state advances, private retreats).” However, the only discussions to date regarding the extent to which the expression guojin-mintui is actually applied to the Chinese economy as it currently exists have been based on vague grounds. Our article tries to analyze whether a clear difference in the labor distribution rate exists between SOEs and non-SOEs, using data of industrial enterprises. Our analysis shows that wage share throughout the analysis period was much higher at SOEs than non-SOEs, even after controlling for various variables. In addition, in Chap. 3, we focus on Shenzhen as an innovation hub and examine sustainable innovation and its institutional context in China. First, it discusses the protection of intellectual property rights (IPR) in Shenzhen’s innovation ecosystem through the classification of companies into three groups—pre-modern, modern, and post-modern—based on their considerations of IPR. Second, it studies the institutional background of the Shenzhen ecosystem, wherein advanced innovation coexisted with counterfeit products that overlooked IPR. Third, the study provides references from the Chinese economy’s history to elucidate the ways in which firms operating in Shenzhen’s manufacturing industry—which is subject to uncertainty resulting from the constant influx of micro-companies and traders—avoid the prisoner’s dilemma. Finally, this study examines the changes in technology and firm dynamics that spurred innovations in the Chinese economy—a space wherein innovations were initially considered unlikely owing to the predominance of short-term transactions. In Chap. 4, we try to empirically clarify the aforementioned questions by matching the databases of industrial enterprises and information on elected representatives of the Local People’s Congress of Yangzhou City, Jiangsu Province. Our analysis shows that the acquisition of political capital through election as a representative of the Municipal People’s Congress certainly improved the performance of some local enterprises in terms of scale by bringing political or economic benefit to these enterprises. With this background, the expansion of their political "credit" allowed them to receive more loans and make more investments. From another aspect, our study also suggests that the acquisition of political capital by local enterprises did not improve the efficiency of their performance. And in Chap. 5, we focus on the local government’s competition to attract companies through the auction of the land usage right. In the existing researches on the competition between local governments tended to focus on the competition of growth rate through the performance evaluation system, or the tax competition for the attracting companies. In China, however, where the system of property tax is still underdeveloped and the local government cannot independently set the tax rate. Therefore, in this study, we have focused on some kind of dumping activity of local governments in the auction of land use rights, to attract industrial companies, and empirically test this hypothesis by the spatial lag model.

8

1 Introduction: The Future of Chinese Economy …

References Kato H (2016) An introduction to Chinese economics: how are “Ambiguous Institutions” Working? The University of Nagoya Press (Japanese) Kishimoto M (2013) The traditional chinese economy as an ideological type: history and economics in Chinese studies, in Research meeting on economic history, available at http://www.cirje.e.utokyo.ac.jp/research/workshops/history/history_paper2013/history1021.pdf. (Japanese) Mori S (2019) US-China relationship over trade and technology (Part 2), in Website of Sasagawa Pease Foundation, available at https://www.spf.org/jpus-j/investigation/spf-america-monitor-doc ument-detail_15.html, (Japanese) Ogawa S (2016) Anthropology of hand-to-mouth life, Kobunsha Sinsho (Japanese)

Chapter 2

Ownership Structures and Disparities in the Labor Share of Chinese Industrial Enterprises

2.1 The Guojin-Mintui Phenomenon and the Decline of the Wage Share It has recently become popular to attribute many characteristics of the Chinese economy to its system of “state capitalism,” which is distinct from that of Western society (Bremmer 2010). One contributing factor, especially as China joins the World Trade Organization and moves in earnest to integrate itself into global markets, is the fact that aggressive investment behavior on the part of the government and certain large enterprises has served as an engine for high growth. In that process, huge government-owned enterprises have appeared in some industries, drawing criticism that “as the state advances, the private sector retreats,” or the Chines expression “guojin-mintui” (国进民退). In addition, the large-scale stimulus measures implemented in response to the 2008 financial crash increased the degree of governmental intervention in the markets and gave credence to the theory that the Chinese economy’s “state capitalism” made China’s economy more acutely dependent on investment. The expression guojin-mintui first found widespread usage in the 2000s, coinciding with the beginning of the Hu Jintao administration. This was due to concerns that the reforms to institutional ownership pushed by the earlier Jiang Zemin administration centered on regional medium-sized state-owned enterprises (SOEs) had stalled and that some large SOEs in which physical and human capital had become concentrated might be crowding out private sector enterprises.1 There have already been many debates inside and outside of China regarding how accurately the expression guojin-mintui represents the reality of the Chinese 1

The term “state-owned enterprise” as used in this chapter refers to ownership systems that include “state-owned enterprises,” “fully-owned state enterprises,” and so on—distinct from share-based enterprises in which the state or a state-owned holding company holds a controlling share. For details, see the definition of “state-owned enterprises” in Sect. 4.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 K. Kajitani and T. Kamo, Political Economy of Reform in China, Kobe University Social Science Research Series, https://doi.org/10.1007/978-981-19-0202-4_2

9

10

2 Ownership Structures and Disparities in the Labor …

economy. There are three main arguments. The first is the insistence that guojinmintui does not exist because the share of SOEs in the total number of enterprises or their share of output value in many industries has been shrinking. Figure 2.1 shows how various statistics demonstrate the share of SOEs in a number of industrial enterprises, gross industrial output value, total asset, tax on principal business, and total profits. It is true that the ratio of SOEs in the total number of industrial enterprises or their value of production has been declining since 1998. So, based on these figures, there is a certain degree of persuasiveness in the view that the term guojin-mintui does not correspond to the actual situation. However, total assets or tax on principal business revenues have remained high, and total profits have started to rise again since 2017 after a period of decline. Looking at these indicators, we cannot simply assume that SOEs are exiting the market. The second argument is that strengthening or maintaining SOEs’ dominance in “controlled highlands” industries (i.e., strategic industries) suggests that guojinmintui had a specific basis (Nakaya 2013). Nakaya considered several industries as “controlled highlands” in the Chinese economy in which SOEs are supposed to occupy a dominant position and examines the capital structure of representative companies in each industry to determine the extent to which the phenomenon of SOEs occupying a dominant position in these sectors is universally observed. As a result, He found that (1) state-controlled industries (i.e., industries such as power generation and transmission, oil and petrochemicals, telecommunications, coal, and aviation), in which it is clearly stated that SOEs should “dominate” the entire industry, have % 90 80 70 60 50 40 30 20 10 1998 99 2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19

0

Number of Enterprises Total Asset Total Profits

Gross Industrial Output Value Tax on Principal Business

Fig. 2.1 The share of state-owned enterprises (SOEs) in the output of industrial enterprises, 1998– 2019. Data China Statistical Yearbook, every year

2.1 The Guojin-Mintui Phenomenon and the Decline of the Wage Share

11

been basically “controlled” by SOEs. (2) On the other hand, in the “major companycontrolled industries” in which the goal is to “control” the major companies, most industries are controlling the “major companies” at a level of 70–90%, indicating that control by publicly owned companies is basically being maintained. The third argument is that certain SOEs enjoy oligopolistic, monopolistic, or favorable funding benefits and, as a result, expand the disparity over non-SOEs in wages and/or compensation. This point is closely related to the claim that as investment-led economic growth continued, the efficiency of the allocation of the means of production was compromised. In other words, despite the intrinsically low productivity of SOEs, they tend to have much a higher labor to capital ratio than privately owned enterprises, giving rise to resource allocation inefficiencies. The research of Brandt and Zhu (2010) is a leading example of a similar perspective. They indicate that (1) the total factor productivity of non-SOEs is consistently higher than that of SOEs; (2) the wages at SOEs are, nevertheless, consistently much higher than those at non-SOEs; and (3) although wage disparities shrunk in the last half of the 1990s, since then they have been rising again. Brandt and Zhu contend that distortions in both the labor markets and the capital markets contribute to such wage disparities. In other words, despite the fact that return on investment is higher in the non-state-owned sector, investment consistently flows toward the state-owned sector, and it is this that has given rise to the gap in wages between the state-owned and non-state-owned sectors. Discussions of guojin-mintui in recent years and calls for reforms in SOEs have not necessarily delineated these points clearly and thus remain vague. Given this situation, this chapter once again reviews the guojin-mintui phenomenon, paying special attention to the change in the wage share in mining and industrial enterprises in the 2000s and the disparities resulting from their ownership structure, focusing on the third point listed above. One fact must be made clear: although China’s economy has exhibited a situation in which its growth depends on fixed-asset investment, the wage share viewed at the macro level exhibits a declining trend. Figure 2.2 illustrates the macro wage share in China over time. Data about the gross domestic product (GDP) distribution during the Mao Zedong era cannot be directly derived, so we used the estimates from Marukawa (2013), who used the ratio of private consumption to GDP plus 3.5 as a substitute estimate for the wage share. Starting in 2004, China’s macroeconomic statistics switched from the previous System of National Accounts (SNA) system to the System of National Accounts, 1993 (93SNA) system. While under the old SNA, the combined income of self-employed persons (i.e., worker compensation plus self-employment profit) were included in “worker compensation,” under 93SNA labor compensation, selfemployment profit was no longer included in “employee compensation.” All else being equal, this results in an underestimation of the wage share (Xu 2003; Bai and Qian 2010). Therefore, we use the following methodology to estimate self-employment labor compensation and exclude it from the statistics for “labor compensation” in the

12

2 Ownership Structures and Disparities in the Labor … 0.7 0.65 0.6 0.55 0.5 0.45

Labor Share A Labor Share B

0.4

Labor Share C

0.35

Labor Share D 2014

2016

2012

2010

2008

2006

2004

2002

1998

2000

1996

1992

1994

1990

1986

1988

1980∼85

1970∼75

1952∼57

1962∼65

0.3

Fig. 2.2 Trends in the labor distribution ratio, 1952–2016. Note 1: Labor Share A is based on the estimate by Marukawa (2013). Note 2: Labor Share B is based on the old SNA. Note 3: Labor Share C is based on the 93SNA and includes labor compensation for the self-employed. Note 4: Labor Share C is based on the 93SNA and excludes labor compensation for the self-employed. Source Department of National Accounts, National Bureau of Statistics of China (2003, 2007), Marukawa (2013), National Bureau of Statistics of China (https://data.stats.gov.cn/)

estimated for wage share in Fig. 2.2. First, since data series are available for both the SNA93 operating surpluses and mixed income under SNA93 from 1993 to 2003 and the operating surpluses under the old SNA (excluding mixed income), we used these two variables to estimate the value for mixed income for each year.2 From Fig. 2.2, we can observe that China’s macro wage share experienced a relatively large increase starting in 2009, but in the 1990s and thereafter, it demonstrated a declining trend. This decline in the wage share is closely related to lower household consumption demand and can be considered the flip side of the investment-led economic growth needed to fill that gap.3 Following the 2000s, as a number of large-scale national development projects were implemented—notably, the Great Western Development—capital flowed into large SOEs in certain strategic sectors, greatly boosting their labor and capital ratios 2

Specifically, using the old SNA series, we obtained the operating surplus ratio, excluding mixed income to GDP, and then multiplied that by the GDP value under the SNA93 to obtain the operating surplus value. We then calculated mixed income by subtracting the operating surplus value from the sum of the operating surplus and mixed income and finally estimated the value of managers’ working compensation as 67% of mixed income. 3 Lu and Guo (2012) estimated the capital and wage share taking into account the redistribution effects of capital and taxes on workers and social insurance from 1978 through 2008. According to their analysis, the wage share after redistribution rose during the first half of the 1980s; thereafter, it began declining. Capital allocation ratios moved in exactly the opposite direction.

2.1 The Guojin-Mintui Phenomenon and the Decline of the Wage Share

13

compared to non-SOEs (Brandt and Zhu 2010). As an increase in the concentration of capital brings about an increase in labor productivity, average wages in the stateowned sector rose far more than those in the non-state sector. Figure 2.3 shows the progress in the wage levels over time for each type of company, considering SOEs as a benchmark. While salary levels at foreign-capital enterprises are consistently the highest, the gap with SOEs shrunk sharply at the turn of the century. This is because wage levels at SOEs enjoying monopolistic benefits in sectors such as finance, power, petrochemicals, and telecommunications rose rapidly to approach the levels at foreign enterprises. In addition, wage levels at stock companies are consistently higher those at SOEs, but the former includes companies in which the government controls a dominant share (so-called “governmentcontrolled enterprises”), and thus they are effectively state-owned. Average wage levels at private companies, whose numbers are growing year by year, are 50–60% of those at SOEs. As previously mentioned, following the 2008 financial crash, the wage levels for unskilled workers rose; as a result, the national Gini coefficient published by the government showed a declining trend after its peak in 2008. However, aggressive investment by regional governments led to an excessive accumulation of capital that was, if anything, more severe than in the first stage, giving rise to new factors for increasing disparities, such as asset disparities arising from the rise in real estate prices and the increase in “gray income” generated using political position (i.e., opaque 2.00

Collective Owned

1.50

Limited Liability Joint Stock

1.00

ForeignCapital

0.50

Private

0.00 1995

98

01

04

07

10

13

16

19

Fig. 2.3 Wage levels over time for enterprises in each ownership system, 1995–2019 Note This graph shows the average level of wage overtime for each form of company, considering SOEs as a benchmark. Source China Statistical Yearbook, 2014

14

2 Ownership Structures and Disparities in the Labor …

income positioned between the legal and illegal4 ). If this type of phenomenon is taken as the strengthening of SOEs’ special privileges, then it is certainly possible to say that there is sufficient evidence to describe the changes in circumstances over the last 10 years or so “as the state advances, the private sector retreats” (guojin-mintui). Having said that, one cannot immediately derive from this the conclusion that workers at SOEs enjoy more special privileges than those at non-SOEs and that those special privileges expanded through the 2000s. It goes without saying that Fig. 2.2 is nothing more than the average of wages aggregated by ownership type, with absolutely no controls for worker or company attributes that affect wage levels such as educational background, length of employment, and industry or company size. To view disparities by ownership type in the strictest sense, one must control for variables that impact wage levels using a methodology such as the one described in the following section.

2.2 Previous Research on Disparities in Wages and Wage Share at SOEs and Non-SOEs Based on an investigation of household finances of urban residents using data survey data from the China Household Income Allocation Study (CHIPS) team in 1988, 1995, 2002, and 2007, Xia et al. showed how wage disparity between urban residents had grown (Xia et al. 2013). Their results indicate that a large proportion of wage disparity is attributable to the company ownership system. Using Blinder–Oaxaca decomposition, they measured the proportion accounted for by “discrimination” (i.e., the portion of the wage differential that is not rationally explicable). Blinder–Oaxaca decomposition refers to decomposing (1) wage differentials by worker categories, such as gender and ethnicity, into two parts: (2) the part that can be explained by individual worker attributes such as age, educational background, and number of years on the job, and (3) the part that cannot be explained by individual attributes (Oaxaca-Ransom 1994). The decomposition also refers to interpreting (3) wage differentials stemming from the differences within categories. According to the analysis by Xia et al. (2013) concerning the wage differential between SOEs and non-SOEs, the ratio corresponding to “discrimination” that could be explained by individual attributes ranged from 43 to 44% between 1988 and 2002 but reached 80% by 2007. Although this research by Xia et al. is truly outstanding, it has several issues. First, the CHIPS data from 1988, 1995, 2002, and 2007 are at five- or seven-year intervals and are therefore unsuitable for capturing continuous changes in the situation. Second, problems can be pointed out in the Blinder–Oaxaca decomposition methodology that they used to analyze wage differentials. For example, Blinder– Oaxaca decomposition first requires estimating a Mincer earnings function. The Mincer earnings function makes the neoclassical assumption that earnings are equal 4

For details on “gray income,” see the series of papers by Wang Xiaolu (2013).

2.2 Previous Research on Disparities in Wages …

15

to marginal productivity levels and are determined solely by individual attributes. It follows that when there are distortions in the labor market or large differences in productivity between enterprises or industries in which workers find employment, errors arise in the estimation of wages, preventing the accurate decomposition of differentials. In other words, it is possible that much of the wage differential shown by Xia et al. stemming from differences in ownership (state-owned vs. non-stateowned), arises from differences in the attributes of enterprises such as industry and capital intensiveness. In fact, Xia et al. noted that the expansion of wage differentials in the 2000s is likely to stem from the decrease in the number of SOEs during that period and their concentration in specific industries.5 To verify the extent to which wage differentials are based on the type of ownership upon adjusting for influencing factors such as industry or company size, the most promising approach is to perform an analysis focused on the differences in the wage share between companies using corporate microdata. A number of studies have inspected the effect of differences in corporate ownership systems on the wage share using microdata. Fleisher et al. (2011) used microdata for 425 industrial enterprises in Beijing, Shanghai, Guangzhou, Tianjin, and Chengdu from 1998 through 2000 to estimate the gap between marginal labor productivity and the wage level. Their results show that wages paid were at a level far below marginal productivity for both skilled and unskilled workers and demonstrate that the gap with marginal productivity was the smallest for SOEs and the largest for foreign-owned enterprises. Bai and Qian (2010), taking note of changes in the distributional aspects of GDP from the 1980s to the present, also analyzed the determining factors for wage share using panel data for industrial enterprises of a certain size from 1998 to 2005. Their results indicate that the share of state-owned capital was positively correlated with the wage share at a significant level and that the decline in the share of the entire sample accounted for by SOEs was an important factor in the decrease in the wage share during the period in question.6 All of these empirical studies suggest that the wage share is higher at SOEs than at non-SOEs; yet, they fail to fully address doubts such as the extent to which the disparity is purely due to differences in ownership structure or time series changes. Furthermore, when a particular function is assumed for purposes of hypothesis testing, it is difficult to deny that errors in identifying the quantitative model may 5

Yue et al. (2010) used the 2005 population census data and a Blinder–Oaxaca decomposition to analyze income disparity between workers in a monopolistic industry (i.e., oil, natural gas, tobacco, power, railways, airlines, finance, etc.) and a competitive industry. Their results show that workers in monopolistic industries, on average, earned RMB 450 more per month than workers in competitive industries and that 50–60% of this disparity could not be explained by educational background or age. 6 Bai and Qian (2010) similarly cited another factor to explain the decrease in the wage share during the period—the effect of corrections to national economic statistics when they were switched from the old SNA to 93SNA in 2004, followed by changes in sectoral weights in the form of a decline in the agricultural sector and growth in the industrial sector. For more on this subject, refer to Qian (2011) as well.

16

2 Ownership Structures and Disparities in the Labor …

give rise to large variations in the estimation results. In the next section, to these problems, we introduce a semi-parametric analysis methodology using propensity scores.

2.3 Analysis of Wage Share Differentials Using Propensity Scores Analysis using propensity scores has recently come into wide use in areas such as policy analysis and disparity decomposition. The advantages of this approach over traditional multivariate regression analysis are discussed below. This methodology enables us to measure the degree to which differences in ownership structure favor SOEs, excluding the effects of industry and company size. In the following, we concisely explain the methodologies of disparity analysis and causal analysis using propensity scores. To date, propensity score-based analysis has been mostly used to analyze estimates of causal policy implementation effects. This method attempts to measure policy effectiveness based on the magnitude of the result variables (y1, y0) observed in two groups: a treated group (z = 1) to which the specific policy was applied and an untreated group (z = 0) to which it was not. In this case, it is impossible to observe a policy’s direct effect from a simple comparison of observables y1 and y0 unless the policy is assigned completely randomly to the subjects in the two groups. In other words, to elucidate the causal effects of the policy, a counterfactual assumption-based comparison is necessary between the outcome variable for the treated group (y1 |z = 1) and the latent outcome variable for the case in which the policy was not applied to the treated group (y1 |z = 0). However, what is observable when the policy is actually applied is the treated group’s outcome variable (y1 | z = 1) and the untreated group’s outcome variable (y0 | z = 0). Considering the existence of other variables (covariates) influencing the policy assignment and outcome variables, this untreated group outcome variable (y1 | z = 0) will, in general, not be equivalent to the outcome variable (y1 | z = 0) for when the policy was not applied to the treated group. However, if the condition holds that the allocation depends solely on the covariates and not on the outcome variables, comparing the outcome variable for the treated group (y1 | z = 1) with the outcome variable for the untreated group (y0 | z = 0) enables us to obtain the magnitude of the causal effect of the policy allocation. With this approach, to obtain the magnitude of the causal effect based on policy allocation, we attach the condition of an appropriate balancing score to the covariate value and can then merely obtain the condition in which the simultaneous distribution of latent outcome variable y1 (when assigning to the treated group) and latent outcome variable y0 (when assigning to the untreated group) are independent of the assignment variable z. The propensity score is normally used for the balancing score. This is generally estimated as the predicted probability of z = 1 being assigned to each

2.3 Analysis of Wage Share Differentials Using Propensity Scores

17

subject, using a logit or probit model that explains the assignment variable z and the covariate x.7 Much progress has recently been made in the application of causal effect analysis using propensity scores to decompose disparity data. In the analysis of causal policy effects, observable variables common to both groups appeared before other variables and influenced both the explanatory and dependent variables. In contrast, when analyzing disparities between categories based on attributes such as gender, one must control for the effect of intermediate variables that have a strong correlation to the categories and influence the outcome variables. One approach to disparity decomposition using propensity scores is to assign certain weights to propensity score estimates for each variable, thereby ensuring the statistical independence of the attribute categories and other explanatory variables (intermediate variables), and then decompose the part of the disparity explainable by the other explanatory variables and the part not explainable. This is called DFL decomposition methodology (Yamaguchi 2014; Cao 2013). There are also several empirical studies of disparity decomposition analysis based on methodologies using propensity score matching (Ñopo 2008; Frölich 2007).8 The advantages using propensity scores rather than traditional regression analysis for this type of analysis will now be presented. First, regarding the relationship between the dependent and explanatory variables of interest (i.e., presence or absence of policy allocation, differences in attributes such as gender and ownership structure, etc.), the influence of other variables is brought together in the propensity score, thereby allowing them to be more rigorously excluded. Second, there is no need to assume a particular relationship between the covariates (intermediate variables) and dependent variables, which enables a more flexible analysis. Third (and related to point two), this type of analysis is more robust in the face of incorrect models. For example, if a linear relationship between the explanatory and dependent variables cannot be assumed, then the bias increases with multivariate regression analysis. In addition, if there are distortions in the market, then bias arises when estimating a parametric model under neoclassical assumptions of relationships between the variables. Propensity score-based analysis does not require the assumption of specific relationships between covariates (intermediate variables) and dependent variables, so this type of bias should not cause major changes in the estimates. Based on these points, this chapter analyzes the “unexplainable disparity” effect of ownership structure on the wage share based on propensity scores matching and covariate analysis and controlling for the effect of the attributes on the wage share. Our analysis will involve the following steps:

7 The strict definition of propensity score is as follows: Given a covariate value xi for subject i and a value zi for the allocation variable, the probability ei (ei = p (zi = 1|xi)) is assigned to Group 1. Since this probability cannot be directly observed, an estimate based on a quantitative model, such as a logit or probit model, is used. 8 Słoczy´ nski (2015) presented a detailed survey of empirical studies on disparity analysis that used propensity scores in this way.

18

(1)

(2)

(3)

2 Ownership Structures and Disparities in the Labor …

Identify other explanatory variables (intermediate variables) that influence both ownership structure (the allocation variable) and wage share (the outcome variable); Estimate a qualitative choice model (logit model) with state-owned as 1 and non-state-owned as 0 and obtain the propensity score by aggregating the intermediate variable information from the estimates; and Create pairs of SOEs and non-SOEs with nearby propensity scores and solve for the average of the differences between the outcome variables.

However, a problem arises due to asymmetry between the number of SOEs and non-SOEs when the same SOE is matched more than once. Thus, the criterion for pairing enterprises with different ownership structures is arbitrary. In the following analysis, we consider this by adding the propensity score as an explanatory variable and measuring the effect of ownership structure. Our analysis also utilizes covariate analysis.

2.4 Analysis Methodology, Data, and Analysis Results Before we show the propensity score analysis results, we first discuss the data and variables being used. We used the China Industrial Enterprises Database from 2000 through 2007. This dataset contains production and financial data for mining and industrial enterprises of a certain size (sales of 5,000,000 Renminbi (RMB) or greater through 2010 and RMB 20,000,000 or greater from 2011). The dataset is available through 2013; however, data such as wage payments to workers are missing for 2008 and subsequent years. Therefore, we removed from the analysis cases in which the industrial value added or number of employees was zero or negative. We also removed other clearly abnormal values.9 We estimated the wage share as total wages as a percentage of industrial value added. We defined guojin-mintui as enterprises with their corporate registration form indicating “SOE” or “fully-owned state enterprise” and those enterprises with government capital in excess of 51%. For corporate “monopoly power,” we follow the analysis of Bai and Qian (2010) and use the markup ratio (revenue divided by the cost of sales) as a proxy variable. The rationale behind this is that if enterprises have monopolistic power in the marketplace, they will be expected to gain more revenue at lower cost, giving them a higher markup ratio. Table 2.1 shows the number of enterprises included in the dataset, the ratio of SOEs, and the change over time in wage share by ownership structure. We see from the table that the proportion of SOEs decreases over the years. In addition, although wage share also shows a generally declining trend, the decline is greater for SOEs than that for non-SOEs. For the propensity score matching procedure, we follow the description in the previous section. We repeat it here as follows: 9

For example, we excluded from the analysis wage shares with a value of 10 or greater.

2.4 Analysis Methodology, Data, and Analysis Results

19

Table 2.1 Difference in wage share by ownership structure, 2000–2007 2000

2001

2002

2003

2004

2005

2006

2007

Number of enterprises

143,069 157,871 164,571 184,441 208,645 263,366 293,828 329,524

Ratio of SOEs (%)

26.8

22.1

17.9

13.7

9.8

7.46

5.93

3.96

Wage share (All)

0.442

0.421

0.407

0.370

0.456

0.358

0.339

0.325

Wage share (SOEs)

0.686

0.682

0.626

0.577

0.775

0.617

0.579

0.471

Wage share 0.352 (non-SOEs)

0.347

0.359

0.337

0.421

0.337

0.324

0.319

Source China Industrial Enterprises Database, from 2000 through 2007

(1)

(2)

(3)

Identify ownership structure (the allocation variable) and other explanatory variables (intermediate variables) that influence both ownership structure and the wage share (the outcome variable); Estimate a qualitative choice model (logit model) with state-owned as 1 and non-state-owned as 0 and obtain the propensity score by aggregating the intermediate variable information from the estimates; and Create pairs of SOEs and non-SOEs with nearby propensity scores and solve for the average of the differences between the outcome variables.

To estimate the propensity scores, we first performed a logit analysis using an SOE dummy as the dependent variable and the number of years in operation, the number of employees, the capital-output ratio (i.e., total fixed assets divided by added value), monopoly power,10 a province dummy, and a two-digit industry classification dummy as the explanatory variables.11 In Table 2.2, we can see the results of a logit analysis to explain the differences between SOEs. The results of the analysis show the following. SOEs have been in existence longer, have more employees, a higher capital-to-output ratio, and a higher percentage of employees with higher education than non-SOEs. However, there was no clear result on the markup ratio. Next, we performed matching and covariance analysis using the estimated propensity scores. The matching was one-to-one but with the same numbers being repeatedly used due to the fact that SOEs accounted for a small number of the entire sample. To confirm whether the analysis using propensity scores excluded the effects of other variables, we conducted an ordinary least squares (OLS) estimation with wage share as a dependent variable and a variable identical to the propensity score estimate and a state-owned dummy as the explanatory variable. We also estimated the covariance analysis regression with the wage share as the dependent variable 10

In the estimation, we use the markup ratio as a proxy for monopoly power. Of all the explanatory variables used in estimating propensity scores, the markup ratio was not significant in nearly every case.

11

143,069

No. of obs

157,871

y

2002

164,571

y

y

(0.00)

0.000

(0.00)

0.024***

(0.00)

0.000***

(0.00)

0.070***

2003

184,441

y

y

(0.00)

0.004*

(0.00)

0.000***

(0.00)

0.000***

(0.00)

0.074***

Note 1 Figures in parentheses are standard errors Note 2 ***, **, * denote significance at the 1, 5, and 10% levels, respectively

y

Industrial dummy

y

(0.00)

y

0.000

(0.00)

(0.00)

(0.00)

0.001

0.027***

(0.00)

(0.00)

0.019***

0.000***

0.000***

(0.00)

(0.00)

2001

0.066***

2000

0.064***

Province dummy

Higher education

Markup

Capital output

Employee

Year

208,645

y

y

(0.00)

0.000

(0.00)

0.012***

(0.00)

0.000***

(0.00)

0.073***

2004

Table 2.2 The results of the logit model explaining the character of SOEs, 2000–2007 2004’

208,645

y

y

(0.05)

1.385***

(0.00)

0.000

(0.00)

0.012***

(0.00)

0.000***

(0.00)

0.075***

2005

263,366

y

y

(0.01)

0.039**

(0.00)

0.012***

(0.00)

0.000***

(0.00)

0.079***

2006

293,828

y

y

(0.00)

0.014**

(0.00)

0.011***

(0.00)

0.000***

(0.00)

0.081***

2007

329,524

y

y

(0.01)

0.039**

(0.00)

0.006***

(0.00)

0.000***

(0.00)

0.078***

20 2 Ownership Structures and Disparities in the Labor …

2.4 Analysis Methodology, Data, and Analysis Results

21

and a state-owned dummy and propensity score as the explanatory variables. The estimates from these analyses are shown in Table 2.3. All of the estimates shown in Table 2.2 show us that there are disparities in wage share due to ownership structure that cannot be explained by differences in company size, capital yield coefficient, or industry. Furthermore, compared with normal OLS, the analysis results using propensity scores for both matching and covariance yield a lower estimate for the effect of the ownership structure on the disparity not explainable by other factors. This is likely due to the fact that in the former case, we succeed in eliminating the influence of other intermediate variables from the ownership structure’s influence on the wage share—a result that could be expected. Still, the effect on disparity due to ownership structure [(3) and (4)] as a percentage of overall disparity is 35–85% (i.e., quite a high level). We also show the estimates exclusively for 2004 (2004’), adding as an exogenous variable educational background of employees, which is available in that year’s data set; however, this had almost no effect on the results.12 What conclusions can we draw from the results of our analysis? Let us first draw attention to the fact that wage share throughout the analysis period was much higher at SOEs than that at non-SOEs, even after controlling for variables such as industry, region, size, monopolistic power, and capital intensiveness. This indicates that workers at SOEs are clearly being favored, as long as the type of company is the same in terms of the variables above. However, as commonly held, it is best to view this not as something that became more prominent in the 2000s. Rather, there had always been a large disparity between SOEs and non-SOEs as regards wage share, which continued in the 2000s. On the other hand, through the 2000s, the wage share at both SOEs and nonSOEs was on a declining trend. The degree of decline was, if anything, greater for SOEs. In other words, the absolute value of the wage share disparity due to the ownership structure is decreasing, but the proportion of the disparity accounted for by “discrimination” not explainable by other factors is exhibiting a tendency to rise, with the disparity attributable to ownership structure becoming fixed. Reinterpreting the research results of Xia et al. (2013) presented in Sect. 2 based on the facts revealed by this study, the following seems likely. Throughout the 2000s, (1) SOEs became capital-intensive and concentrated in particular sectors. At the same time, (2) labor productivity (added value per worker) at SOEs rose. We believe that this accounts for a decline in the wage share and a rise in wages in the state-owned sector happening simultaneously, as shown by Xia et al. The dual phenomena of (1) and (2) arose from greater capital intensiveness and industry tilt in SOEs, and we believe that this explanation is consistent with the rise in value added at those enterprises. 12

This empirical study is highly preliminary, making it necessary to continuously review and enhance the methodology and variable definitions. For example, here, we presented a comparison between the two categories of SOEs and non-SOEs. We still need also to compare the situation in categories within non-SOEs such as proprietorships and foreign-owned enterprises. Pursuing more robust results in this area is a topic we intend to take up in the future.

0.416

0.685

0.751

(0.01)

0.183**

(0.01)

0.203**

(0.00)

0.204**

0.267

2002

0.696

0.756

(0.00)

0.167**

(0.01)

0.181**

(0.00)

0.188**

0.240

2003

0.843

0.704

(0.01)

0.299**

(0.01)

0.251**

(0.01)

0.291**

0.354

2004

0.760

0.735

(0.01)

0.269**

(0.01)

0.262**

(0.01)

0.300**

0.354

2004’

0.654

0.639

(0.00)

0.183**

(0.01)

0.178**

(0.00)

0.213**

0.280

2005

0.663

0.677

(0.00)

0.169**

(0.01)

0.190**

(0.00)

0.202**

0.254

2006

0.743

0.750

(0.00)

0.112**

(0.01)

0.110**

(0.00)

0.120**

0.151

2007

denotes significance at the 1% level and * denotes significance at the 5% level Note 1 “Difference” stands for the difference in wage share between SOEs and other companies Note 2 The figures in parentheses are t values Note 3 We show estimates for 2004’, adding as an exogenous variable the educational background of employees, which is only available for 2004

**

0.557

(4)/(1)

0.353

(0.00)

(0.01)

0.502

0.139**

0.186**

0.117**

(0.01)

0.179**

(0.01)

(3)/(1)

Covariance Analysis(4)

Matching (3)

0.226**

(0.00)

0.277**

(0.00)

OLS(2)

0.335

0.334

2001

Difference (1)

2000

Table 2.3 The result of empirical studies on wage disparities due to ownership structure, 2000–2007

22 2 Ownership Structures and Disparities in the Labor …

2.5 Conclusion

23

2.5 Conclusion Our empirical analysis demonstrates that the wage share throughout the analysis period was much higher at SOEs than that at non-SOEs, even after controlling for variables such as industry, region, size, monopolistic power, and capital intensiveness. This indicates that the workers at SOEs are clearly being favored, as long as the type of company is the same in terms of the aforementioned variables. The results of our analysis call into question why more favorable treatment for workers at SOEs continues, and it also shows that there are no rational grounds for this phenomenon. Going forward, how will the Chinese economy be influenced if this favorable treatment of SOEs continues in the absence of rational grounds? We can conjecture the following three effects: (1) the stability of labor relations in “controlled highlands” industries, (2) a rise in awareness of subjective inequality, and (3) inefficiencies in allocating the elements of production. Depending on which of these is assigned the greatest weight, the view of the phenomenon of “as the state advances, the private sector retreats” (guojin-mintui) itself could undergo a shift and become more ambiguous. Finally, let us touch on the outlook for the impact of the ownership structurebased disparity in the wage share on the Chinese economy. Due to data limitations, the results of the empirical analysis in this chapter dealt only with the situation up until the 2008 financial crash caused by Lehman shock. As is common knowledge, following the financial crash, the disparities changed significantly. The Gini coefficient published by the government peaked in 2008 and has been declining for the last several years. In addition, as shown in Figs. 2.1 and 2.2, the wage share, which had been declining, began to rise in 2009, and the wage differential between SOEs and enterprises with other ownership structures such as proprietorships has also been shrinking over the last several years. Under the premise that the capital coefficient does not change greatly, the rise in wage share (i.e., the decline in capital allocation ratio) depresses return on capital. This would usually be expected to reduce investment. Nevertheless, starting in 2009, the proportion of GDP accounted for in gross capital formation increased significantly, reaching about 50%. Obviously, what led to this situation is the large-scale economic stimulus following the 2008 financial crash as well as the investment behavior of regional governments using funding platforms energized by that stimulus. Notwithstanding the decline in returns, this continuation of a high level of investment, including private capital, was supported by capital gains. In other words, this phenomenon was inseparable from the emergence of the asset bubble. Due to the expectations of capital gains following the 2008 crash, the Chinese government succeeded in reversing the previous decline in the wage share and in reducing the manifestations of the guojin-mintui phenomenon (i.e., the discrimination in worker treatment among sectors) with the accompanying latent social tensions.

24

2 Ownership Structures and Disparities in the Labor …

However, this approach ran the risk of not working due to the end of the bubble, which meant a decline in expected asset prices.13

References Bai C, Qian Z (2010) The factor income distribution in China: 1978–2007. China Econ Rev 21:650– 670 Brandt L, Zhu X (2010) Accounting for China’s growth, IZA Discussion paper, No. 4764 Bremmer I (2010) The end of the free market: who wins the war between states and corporations? New York, Portfolio Cao C (2013) JGSS statistical analysis seminar 2013: applied models using the trend score weighting method. Res Center Japanese General Soc Surv Collected Papers 14:37–51 (in Japanese) Department of National Accounts, National Bureau of Statistics of China (2003) Data on the gross domestic product of China 1996–2002, Beijing, China Statistics Press (in Chinese) Department of National Accounts, National Bureau of Statistics of China (2007) Data on the Gross Domestic Product of China 1952–2004, Beijing. China Statistics Press (in Chinese) Fleisher B, Hu Y, Li H, Kim S (2011) Economic transition, higher education and worker productivity in China. J Dev Econ 94:86–94 Frölich M (2007) Propensity score matching without conditional independence assumption—with an application to the gender wage gap in the United Kingdom. Econometrics J 10:359–407 Lardy N (2019) The state strikes back: the end of economic reform in China? Peterson Institute for International Economics Lu Y, Guo J (2012) Estimation on the factor income distribution of China. Econ Res J 10:27–41 (in Chinese) Marukawa T (2013) The modern chinese economy. Tokyo, Yuhikaku (in Japanese) Nakaya N (2013) A survey of the control of public owned enterprises in China’s ‘controlled highlands’ sector. Survey and Materials, No. 118, Center for International Economic Policy Studies, Graduate School of Economics, Nagoya University (in Japanese) Ñopo H (2008) Matching as a tool to decompose wage gaps. Rev Econ Stat 90:290–299 Oaxaca R, Ransom M (1994) On discrimination and the decomposition of wage differentials. J Econometrics 61:5–21 Qian Z (2011) The analysis for factor distribution of national income in China, Beijing, Chinese Financial Press (in Chinese) Słoczy´nski T (2015) Average wage gaps and oaxaca–blinder decompositions, IZA Discussion Paper, No. 9036 Wang X (2013) Grey income and distribution of national income: summary of 2013 report [Retrieved 2013–9–26], available at http://wangxiaolu.blog.caixin.com/archives/61797 Xia Q, Song L, Li S, Appleton S (2013) The effects of the state sector on wage inequality in urban China: 1988–2007, IZA Discussion Paper, No. 7142 Xu X (2003) Research on problems of Chinese economic accounting and macro-economic issues. Beijing, China Statistics Press (in Chinese)

13

Lardy (2019) pointed out that under Xi Jinping’s administration, SOEs have become more aggressive in raising capital and investing. The gap in the capital-output ratios between SOEs and private enterprises has become increasingly wide. He calls this situation “The State Strikes Back.”.

References

25

Yamaguchi K (2014) Income disparities between male and female full-time white-collar workers: understanding 80% of the causes and mechanisms behind the gap, RIETI Discussion Paper Series, 14-J-046 (in Japanese) Yue X, Li S, Taili S (2010) Reflections on the problem of high income in the monopolistic industrial sector. Chinese Soc Sci 3:77–93 (in Chinese)

Chapter 3

Evaluating the Sustainability of Innovation in China from the Perspective of Institutional Economics

3.1 Is China’s Innovation Sustainable? In 2014, the Chinese Government stated that the Chinese economy had entered the “new normal,” i.e., a phase defined by stable growth, and announced its intention to continue introducing reforms characterized by a shift from the crude, investment-dependent growth strategy previously opted for and an emphasis on market mechanisms. Thus, the Government emphasized innovation as a strategy to develop a new growth pattern. Moreover, the “mass entrepreneurship and innovation initiative” (大众创业, 万众创新) was proposed by Prime Minister Li Keqiang in 2014. A March 2015 report on government activities at the National People’s Congress describes “Internet Plus,” the Chinese Government’s new initiative, which promotes economic development by integrating mobile Internet, cloud, big data, and Internet of Things (IoT) with modern manufacturing. Moreover, 2015 was also the year in which the State Council released a notice on “Made in China 2025,” a comprehensive industrial policy package designed to facilitate innovations through an integration of IT and manufacturing technologies and, therefore, transform China from a manufacturing giant to the world’s largest industrial powerhouse by 2049, the year that will mark the 100th anniversary of the People’s Republic of China. Consequently, technological innovations, including electrical and digital communications, optical machinery, and computer technology, have become more vigorous in new industries. These technologies have partially surpassed those of the United States and other developed countries. In addition, massive innovation has occurred in terms of the social implementation of these technologies through their application to new business models such as the shared bicycle ride service and other shared-ride services, various e-commerce services, P2P lending, and telemedicine. In addition, the Chinese Government has been emphasizing policies to protect intellectual property rights (IPR). In 2008, China announced the Guidelines for the © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 K. Kajitani and T. Kamo, Political Economy of Reform in China, Kobe University Social Science Research Series, https://doi.org/10.1007/978-981-19-0202-4_3

27

28

3 Evaluating the Sustainability of Innovation in China from the Perspective … 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 US

Japan

China

Germany

South Korea

Fig. 3.1 Number of international patent applications Source Prepared by the authors based on WIPO and PCT Yearly Review

State Intellectual Property Rights Strategy and made public a policy to enhance the protection of IPR. Moreover, Chinese companies’ international patent filings backed by a certain level of technological sophistication have increased rapidly. In 2017, China’s patent volume surpassed that of Japan, eventually proceeding to attain a figure of 58,990 and surpassing the patent value of the United States in 2019 (Fig. 3.1). Moreover, many companies in Shenzhen, including Huawei Technologies (华为科技), ZTE (中兴通讯), and Tencent (腾讯), actively engaged in innovations, filing the highest number of international patent applications. Simultaneously, the United States has become increasingly wary of China’s rapid technological advances in such high-tech fields. For this reason, the recent intensification of the tariff hike battle between the United States and China under the Trump administration, coupled with the imposition of a de facto embargo on Huawei, has increased the uncertainty surrounding China’s current approach to economic growth. Can China continue carrying out such attention-drawing innovations in the future? Generally, mainstream economists have negative views on the Chinese economy owing to its institutional flaws, including its inadequate protection of IPR, the inconsistent rule of law, and government intervention in the economy without accountability, seem to only hinder innovations as engines of sustainable growth. In the book entitled Why Nations Fail, Acemoglu and Robinson (2013) use the term “extractive institutions” to refer to systems that hinder such sustainable economic growth and are beneficial only to the privileged classes. Meanwhile, Acemoglu and Robinson (2013) define “inclusive institutions” as those promoting innovations and economic growth. An inclusive institution can be

3.1 Is China’s Innovation Sustainable?

29

bifurcated as follows: (i) an inclusive political institution represented by a parliamentary democracy and ii) an inclusive economic institution represented by a free and fair market economy. Citing abundant historical examples, these authors emphasize the rule stating that ironically, in inclusive political institutions, autocratic authority figures eventually establish extractive economic institutions to eliminate the momentums of economically successful people whom they perceive as potential threats to the powers currently vested in these authorities. Acemoglu and Robinson portray China’s economic trajectory negatively from the dynasty era to the Mao Zedong era, presenting it as a typical example of a situation wherein totalitarian authority figures hinder economic development and prevent the establishment of an inclusive institution. They also insist that China’s current rapid economic growth is a temporary phenomenon resulting from the partial introduction of an inclusive economic institution within an extractive political regime. Therefore, unless China’s political institution becomes inclusive—as in the case of parliamentary democracies such as Korea and Taiwan—the current elevations in economic growth will eventually subside.1 North, Wallis, and Weingast (2009) state that sustainable economic growth is only possible in open-access states characterized by the existence of rich and active civil societies, establishment of the rule of law and property rights, and expansion of social relations established on the grounds of equality and justice. Their idea of "open-access states" roughly corresponds to the inclusive institution by Acemoglu and Robinson.2 This chapter examines the sustainability of innovation in China and its institutional context by considering Shenzhen’s status as a center of innovation, particularly in the domain of hardware (e.g., the maker movement3 ). 1

Roland (2018) developed a database on the socio-economic conditions of the “state” in different regions and periods across the bygone and modern eras. The database is based on historical and archaeological research and describes the major differences between statist and market systems. Roland individually scored various aspects—including land and slave ownership systems, legal systems (property and contract law), the market economy (commerce and trade), concentrations of power, the role of cities and merchants in the economy, clan relations, and ethnic diversity—on a scale of 1–10. He found that the development levels of domestic and international market activities were significantly correlated with the assessment scores obtained for legal, political, and social institutions. Moreover, Roland (2018) argued that the development of individualistic institutions— including private and public property, contract law, and the decentralization of power—have caused an increase in market activities. These empirically drawn conclusions were almost the same as those obtained by Acemoglu and Robinson (2013) a series of works as well as the findings of studies conducted by others. In addition, through their empirical research, Gorodnichenko and Roland (2012, 2017) suggested that market transactions supported by modern legal systems reflected the influence of western cultural values, whereas non-Western regional trading practices were associated with authoritarian governments and closed and communal cultures. The studies by Roland as well as Acemoglu and Robinson draw dichotomous conclusions. 2 In contrast, Acemoglu and Robinson (2019) recently published book argues that economic growth under a comprehensive system represented by the rule of law is a “narrow corridor”; such growth can be achieved only by balancing the Leviathan and society. Such an argument comes across as a modification of the dichotomous argument presented in their previous book. 3 The concept of “maker movement” was pioneered by Chris Anderson. See Anderson (2012).

30

3 Evaluating the Sustainability of Innovation in China from the Perspective …

The remainder of this chapter is organized as follows. Section 3.2 presents a perspective on the IPR protection measures in Shenzhen’s innovation ecosystem and the classification of companies as pre-modern, modern, and post-modern based on their consideration of IPR in their business innovations. Section 3.3 examines the institutional background of Shenzhen’s ecosystem; wherein advanced innovation coexists with counterfeit products that overlook IPR. Section 3.4 considers a historical perspective of the Chinese economy to examine how firms belonging to Shenzhen’s manufacturing industry—which faces uncertainty from the constant entry of micro-companies or traders—avoid the prisoner’s dilemma. Finally, Sect. 3.5 and conclusion suggest mechanisms through which changes in technology and firm dynamics have led to innovations in the Chinese economy wherein innovation was previously considered unlikely owing to the prevalence of short-term transactions.

3.2 Shenzhen’s Innovation Ecosystem 3.2.1 The Three Groups of Companies Based on Their Consideration of Intellectual Property Rights The Chinese electronics industry is concentrated in Shenzhen, Guangdong Province, and is promoted by the Chinese Government as a center for startups and innovations initiated by the general public. In the 1980s, the first special economic zone was established in Shenzhen following the introduction of a policy concerning China’s global expansion. The city experienced rapid growth owing to the demand for industrial material from the processing trade in labor-intensive industries. Subsequently, many labor-intensive industries withdrew from Shenzhen due to an increase in wages and changes in preferential policies governing foreign capital. However, specialized markets and multi-tenant buildings entailing booths run by wholesalers and manufacturers and offering electronic parts were rapidly developed in Huaqiangbei (华强 北), which consequently emerged as an electronic hub. Notably, groups of companies with entirely different views on IPR protection coexist in Shenzhen. These groups have been classified as pre-modern, modern, and post-modern in terms of their attitudes toward IPR. First, the pre-modern group overlooks IPR completely. The keyword symbolizing this group is shanzhai (山寨). The second group in this classification is the “modern” group (consisting of companies such as Huawei and ZTE) comprising firms that protect their proprietary technologies through patents or other modern and secure IPR protection strategies. Finally, the post-modern group actively releases unique technological interventions in the public arena to promote innovations. It typically involves several people and does not protect them with patents. The authors believe that one of the features of Shenzhen’s ecosystem (industrial ecosystem) is that these three groups are blended.

3.2 Shenzhen’s Innovation Ecosystem

31

3.2.2 Ignoring Intellectual Property Rights and Vertical Disintegration A pertinent example demonstrating the pre-modern group’s disregard for IPR is shanzhai mobile phones (山寨手机) which, over 10 years ago, dominated the Chinese market as well as other emerging ones such as the African market. The widespread popularity of shanzhai mobile phones over 10 years ago was possible owing to technological advancements. MediaTech Inc., a Taiwanese semiconductor company, developed a general-purpose integrated circuit (IC) chip for mobile phones. This innovation increased the availability of general-purpose electronic chips and boards, which are necessary for the manufacture of mobile phones. This development allowed anyone to assemble a mobile phone, even without advanced skills. In China, the telecommunications and mobile phone industries used to be strictly controlled by the government. However, the development of general-purpose IC chips has enabled microenterprises to manufacture and sell mobile phones without adhering to administrative procedures, including business license requirements and connection inspections. Soon, the availability of shanzhai mobile phones increased rapidly, allowing microenterprises to save on acquiring licenses costs and enabling them to develop new products more efficiently. Most of these new products launched in the market were replicas of conventional models or those offered by famous companies; however, they served distinct functions. Some Japanese scholars observed that the manufacturing technique of shanzhai mobile phone is closely related to the process of “vertical disintegration” (Marukawa, 2013: Chap. 2). In the typical cases, the different manufacturing processes— including planning and design, assembly, circuit design/software development, board manufacturing, and parts procurement—for mobile phones and smartphones are integrated within one company. However, China’s shanzhai mobile phones are developed under an extreme division-of-labor system wherein independent microenterprises are in charge of each process. Therefore, if microenterprises purchase a chip mounter to mount electronic parts on boards, they can manufacture their products using the inexpensive parts available in multi-tenant buildings and still enter the smartphone market. A product manufacture system based on vertical disintegration significantly reduces the restrictions on companies with regard to entering the smartphone market. This relaxation of restrictions stems from the fact that companies no longer have to make large capital investments for small, individual processes. The fierce price competition among microenterprises significantly reduces the aggregate production cost of the intermediate goods sector as well as the prices of the main units. Therefore, even without prior knowledge of electronics, companies can manufacture products using general-purpose goods such as printed circuit boards, electronic parts, and molds readily available in the market. Such identical products may then be named variously and launched in the market. Incidentally, versatile electronic parts and molds are called public boards and public models because they have already been launched and are available for public consumption.

32

3 Evaluating the Sustainability of Innovation in China from the Perspective …

3.2.3 Protection Through Patents Next, Huawei is used as a representative example of the modern group. It is recognized as a company that manufactures and sells SIM-free mobile phones in Japan. However, Huawei is also a business-to-business (B2B) giant dealing into the manufacture of computers and equipment for telecommunications carriers. These products by Huawei account for approximately 60% of its sales. Huawei have the ability to independently provide all equipment and devices required for data centers supporting e-commerce and cloud computing. In the recent years, Huawei has emerged as a global leader in terms of developing next-generation (5G) communications infrastructures. Moreover, U.S. government has been very wary of Huawei as a threat to U.S. technological dispatch because of its high technological capabilities. For this reason, Huawei has been the target of strong sanctions by some western countries. For example, in May 2019, the U.S. Department of Commerce took measures to prohibit the export of U.S. products to Huawei. Huawei has a research and development (R&D) team with more than 80,000 experts and boasts of advanced technical capabilities. Such expertise and proficiency have placed the brand among the top five companies in the world in terms of the number of international patents filed by it in the span of the past ten years. It is clear from the above that Huawei develops its proprietary technologies by adopting a vertical integration management style. This strategy contrasts with the one adopted by shanzhai mobile phones. Moreover, Huawei adopts the authentic IPR protection mechanism4 of protecting its technological innovations with patents.

3.2.4 Innovations Through the Open-Source Concept Comprehending the post-modern group might be more difficult compared to understanding the two groups already mentioned. This section explains the concept using as an example. Seeed’s business Seeed Technology Limited may be best understood by means of considering the term open-source hardware. Open-source is a concept advocated in software development wherein the software 4

Takasu and Takaguchi (2020, pp. 145–147) suggested that even large Chinese companies (e.g., Huawei) that are seemingly positioned in the modern tier do not emphasize copyright protection through patenting when developing low-end products. Rather, they collect publicly available technology and components to develop these products. In other words, they behave similar manner as pre-modern companies in terms of manufacturing low-end products. This phenomenon is related to the fact that innovation by pre-modern companies is underpinned by what has long been observed in Shenzhen’s manufacturing industry as kongkai (公开), a development system wherein electronic boards, molds, etc. are shared effectively among small manufacturers to achieve significant cost reduction in the development of new products (Huang 2017: chap. 4). In other words, the kongkai system of innovation is not the exclusive property of the pre-modern class; however, it is often used by modern and post-modern companies. Such inflexibility is a major characteristic of Shenzhen’s ecosystem.

3.2 Shenzhen’s Innovation Ecosystem

33

source code is made publicly available for copying and modification. Moreover, it is a free-of-charge initiative to promote innovations by facilitating the participation of several engineers in the development process. Software products such as the Linux OS and the Firefox web browser were developed under the open-source concept. The approach to hardware development— together with the strategies for the development of 3D printers and the IoT technology—has begun to mimic the approach followed for software development5 ; consequently, the open-source concept, which is used extensively in the software development process, is now being adopted in the hardware field (Anderson 2012). To date, Seeed has been leading the production of open-source hardware in Shenzhen. Seeed’s products, circuit diagrams, computer-aided design data, bundled software code, and other information is freely available in the public arena. Furthermore, the company utilizes the Internet to seek feedback from the general public in terms of its areas of improvement. Another important business area is the manufacturing of electronic components in small lots (dozens to hundreds) in response to customer orders. Seeed undertakes such small-lot manufacturing activities to support manufacturers or individuals aiming to set up hardware manufacturing businesses. As exemplified by Seeed, Shenzhen has attracted companies supporting smallscale prototype manufacturing, venture capitals providing seed funding, and “makerspaces” providing manufacturers and funders with spaces to gather and exchange information. Moreover, these spaces enable makers and funders each find the partner they need and offer their services to each other.

3.3 The Institutional Background of the Shenzhen Ecosystem 3.3.1 Coexistence of Shanzhai and Innovations As discussed in the previous section, one of the many prominent phenomena of Shenzhen is the coexistence of the pre-modern (represented by shanzhai mobiles), modern (represented by Huawei), and post-modern (represented by Seeed) groups of companies. Is the relationship between these three groups contradictory or complementary? Does the excessive level of counterfeit existent in the market hinder innovation? Conventional economics holds that protecting IPR is essential because companies exert significant efforts on R&D to reap the benefits of innovations. If patents do not protect these benefits, no individuals or companies will engage in R&D. Recently, however, there have been arguments against this mainstream view. Raustiala and Springman (2012) present one such representative argument. They argue that the availability and widespread consumption of counterfeit products might promote

5

This movement was pioneered by Anderson (2012) in his book titled Makers.

34

3 Evaluating the Sustainability of Innovation in China from the Perspective …

innovations. In many cases, patents do not protect new ideas. A typical example is that even restaurants that invent recipes are not protected by the law. How does the circulation of counterfeit products drive innovation? In short, the availability of cheap counterfeits could increase demand tremendously. The global boom in hand spinners is one such example. The hand spinners developed in the United States were expensive toys. However, when many counterfeit hand spinners costing about one-fifth of the original price began being manufactured in Shenzhen, the global market witnessed an abundance of hand spinners. Without counterfeits, this boom would not have been as significant. Nevertheless, some economists have their reservations about the relationship between innovations and IPR. For example, Stiglitz and Greenwald (2014, pp. 444– 453) observe that combining three innovation systems—for instance, patent protection, prize system, and government support—is particularly imperative in the innovation process. They also highlight that in the developed western countries, a strong emphasis on IPR protection typically increases social costs of slow diffusion of necessary technologies. Generally, in developed countries, the technology development systems are based on the protection of IPR. In this system, “winners gain everything”, causing a vast difference in the social and private returns earned. In other words, vehement protection of IPR might ultimately hinder innovations as a result of the monopolies on technologies. Such monopolies are typically developed by means of innovations or restrictions on small- and medium-sized enterprises from engaging in R&D owing to the fear of violating the IPRs of other entities.

3.3.2 Shenzhen’s Ecosystem and Independent Design Houses In a world characterized by the widespread availability and consumption of counterfeits, various companies enter the markets and sell their products. Therefore, consumers are likely to purchase products of substandard quality. However, a solution to the significant challenge of finding trustworthy companies to collaborate with for business endeavors has been unearthed by Shenzhen’s ecosystem in the form of independent design houses (IDHs). According to Fujioka (2017: Chap. 3), the owner of Genesis, an OEM electronic manufacturing company in Shenzhen, IDHs originally operated as subcontractors to major IT companies such as Intel and Qualcomm. However, over time, they have begun serving as guiding sources for electronic companies. When a company like Genesis receives a customer order for a tablet with an Intel chip, it first orders a motherboard from a IDH. The IDH then sends a detailed list of the electronic parts that should be used in addition to the design of the motherboard. In other words, the company can launch a new product even without knowing who has manufactured its components. According to Fujioka, because Shenzhen is akin to a jungle comprising multiple players, reaching optimal solutions is a challenge, which necessitates the existence of

3.3 The Institutional Background of the Shenzhen Ecosystem

35

guides in the process. IDHs undertake platform design activities and guide individuals to find suitable, trustworthy companies with which to engage in business. IDHs can help individuals navigate Shenzhen’s ecosystem more efficiently.6

3.3.3 Solving the Prisoner’s Dilemma As mentioned earlier, startups in Shenzhen face several hurdles because it is difficult to determine who may be trusted. The Prisoner’s Dilemma, wherein both of the traders lose if no exchange of information takes place, is prevalent in Shenzhen. In this scenario, what are some effective solutions for the establishment of successful transactions? Bowles (2003, p. 248), an economist known for his research on interdependent relationships between social institutions and economic activities performed by individuals and companies, presents three solutions to this problem—retaliation, reputation, and segmentation. Retaliation, the first solution, is based on a long-term business relationship between economic agents. This solution is governed by the motto “cooperate unless the other party betrays.” In this condition, acts of cooperation by one of the parties facilitate long-term profits for the other party; such positive consequences, in turn, motivate the other entity to stay committed in the association. In contrast, acts of betrayal by one of the entities in the association cause the other entity to incur losses, which fuels betrayal by the affected party. The “reputation” solution emphasizes determining a company’s reputation to judging whether one should strike a deal with the counterparty and engage in conditional cooperation with this entity. In addition, brokerage, as subsequently described, can be understood as a form of risk avoidance achieved by delegating to a third party the task of gathering information about an entity as a means of gauging its reputation. The third solution—segmentation—is to dwell among a small community; such a measure helps genuine entities alleviate the risks involved in frequent trading. In other words, risks are mitigated by means of the exclusion of unfamiliar entities highly likely to betray each other. This approach could be understood as a transaction within a cognate network or community. How can the prisoner’s dilemma be avoided in actual business transactions? Ecommerce platforms such as Alibaba Group provide a rating system that allows parties to rate one another after each transaction. A higher rating implies a more reliable party. Thus, criteria are established to help determine which entities may be trusted for business transactions against the backdrop of a jungle-like environment. However, such a system entails limitations in professional manufacturing fields such as Shenzhen’s electronics industry. In keiretsu (industrial group) trading by Japanese companies, uncertainty is reduced through the establishment of long-term, stable trading associations between 6

A detailed analysis of the role played by IDH in the development of Shenzhen’s manufacturing industry has been provided by Hioki and Ding (2018a, b).

36

3 Evaluating the Sustainability of Innovation in China from the Perspective …

manufacturing companies and suppliers. According to Bowles’ theory, this arrangement could be regarded as a combination of retaliation and segmentation (Bowles 2003). However, this method is not effective in the rapidly changing field of the current electronics industry because other companies might provide higher quality products at lower costs; alternatively, the product lineup might need to be changed based on the circumstances. In contrast, design houses equipped with advanced skills, knowledge, and design capabilities exist in Shenzhen. These IDHs actively mediate between microenterprises in local areas and manufacturing companies unfamiliar with the local circumstances, striving to provide optimal solutions through an ecosystem that facilitates business transactions in a rapidly changing industry. The next section considers China’s historical background to discuss the institutional significance of the Shenzhen ecosystem in the development of China’s economy.

3.4 Institutional Background of Innovation in China 3.4.1 Brokerage Economy as the Traditional Economic Order of the Market The significant weight of intermediaries as third parties in trading has been an institutional feature of the Chinese economy from the dynastic era to the present. This study refers to economies characterized by the existence of such intermediaries as an brokerage economy. Various existing historical studies primarily substantiate the existence of such economies from the bygone eras. For example, Furuta (2013) specifies, “brokerage is very common in the Chinese economy, in all socio-economic relations between two parties, from the general business relationship between goods, to the sale of land and contracts related to socio-economic services, and employment and guarantee of people. So there are etc. I pointed brokerage such as zhongren (中人) baoren (保人) and pubao out.”7 Adachi (2012) emphasizes that China’s traditional market order was characterized as extremely liberal, open, and competitive, with virtually no government restrictions on entry. Moreover, new entrants were not excluded by guilds and industry associations. Because the market’s operation was not supported by formal institutions represented by the rule of law, the execution of commercial transactions had to ultimately rely heavily on ad hoc business relationships. For example, with regard to the purchase of goods from the place of origin or the sale of goods at the place of consumption, merchants involved in wide-ranging business transactions (geshan, = 客商) would always be entrusted with the task of 7

These characteristics of the market order in traditional China have been pointed out by many Japanese researchers who observed the Chinese economy in the early twentieth century.

3.4 Institutional Background of Innovation in China

37

purchasing a product from its place of origin or selling a product at its place of consumption to middlemen (yaren, = 牙人) who were familiar with the local market situation. The yarens obtained rents by arbitrating the price differences that result from information asymmetries between regions (Adachi 2012, pp. 543–54). The business relationship between geshang and yaren is distinctly ad hoc and short-term in nature. In other words, in the geshang–yaren system, the multifaceted mechanism of sanctioning through sharing information among merchants, including the medieval Maghrib merchants subsequently discussed, was not always effective. The cost of distribution of these merchants’ goods was extremely high because of the involvement of a large number of intermediaries. In contrast, the market was constantly in a hypercompetitive situation because the market economy continued to operate and brokerage businesses faced weak entry restrictions. Therefore, the risk of a fixed organization or producers engaging in capital investments was significant. The essence of this “brokerage economy” is well expressed in the word “bao ( 包),” a commonly used term in Chinese society. The word “bao” originally implied concentrating a task in the hands of a third party. The characteristic of the act of “bao” entails risk avoidance by means of shoving a particular task to a third party i.e. a broker; in other words, there is a mechanism of contract that exchanges risk and return. Pre-war and post-war sociologist Sukekata Kashiwa insists that the key to explaining the constitutive principles of traditional Chinese society lies in the concept of “bao” (Kato 2016). Generally, an economy biased to brokerage and mediation has been perceived as a transient phenomenon in the state. In such economies, the market is typically underdeveloped; in other words, the process of compulsory fulfillment of a contract by a public authority is absent. However, an observation of the market economy in modern China indicates that reputation and mediation remain integral criteria for risk management pertaining to incomplete contracts. For example, packaging systems characterized by the hiring of workers by individuals in charge of packing in non-skilled labor employment hires persist at employment locations such as construction sites. Moreover, the problem of deferring the wages of farmers and workers is regarded as a social problem. Additionally, in the corn trade, there still exist producers and jingji ren (経紀人), among others (Zhang 2014). Across the world, there exist few intermediaries who play pivotal roles in trading commoditized (close to a complete contract) goods such as corn. Today’s platform companies such as Alibaba and Tencent may be defined as huge mediators: These platform companies provide the criteria for trader’s reputation as a public good and obtain a commission to reduce the uncertainty inherent in small transactions.

3.4.2 Brokerage as a Form of an Incomplete Contract In emerging countries such as China, various risks impede the smooth operation of the market economy. For example, contracts businesses pertaining to transactions are not

38

3 Evaluating the Sustainability of Innovation in China from the Perspective …

Table 3.1 Nash equilibrium in incomplete contract games Cooperation (C)

Betrayal (D)

Cooperation (C)

(5, 5)

(−4, 6)

Betrayal (D)

(6, − 4)

(−3, − 3)

Notes 1. The rows represent the player’s strategy, and the columns represent the opponent’s strategy. 2. The first letter in parentheses represents the player’s strategy, whereas the second letter in parentheses represents the opponent’s payoff.

implemented adequately among private economic agents. Therefore, the possibility of “betrayal” exists perpetually. In the economics and contract theory system, contracts entailing transactions involving risks, including the possibility of betrayal, are called incomplete contracts. Bowles (2003, p. 252) analyzes the rationality of various institutions with regard to economic activities and the dynamic evolution process undertaken by means of the applied game theory. He described incomplete contract as conditionally updated contracts, such as price determination through negotiations, incomplete contracts characterized by trust and retribution, transactions through a network by a small number. And he also describes complete contract as characterized by onetime contracts, refusal of price negotiations/proposals, and anonymous participants without many connections. In the incomplete contract, as discussed in the previous section, Bowles (2003) cited three types of conditional cooperation models—retaliation, reputation, and segmentation—that prevent betrayal and establish cooperative business relations. Two such cases of conditional cooperation will be discussed, and the conditions under which such processes are carried out are observed. One of the cases of conditional cooperation discussed in this study is retaliation, and the other pertains to the instance in which a third party obtains the counterparty’s reputation and mediates a cooperative transaction.8

3.4.3 Nash Equilibrium and the Prisoner’s Dilemma First, we examine the prisoners’ dilemma in so-called “recurring” games and review the conditions under which the retaliation mechanism works. Table 3.1 depicts the situation of a typical prisoner’s dilemma as the “gain table” for a strategy-type game. Players 1 and 2 select betrayal in only one game, and the 8

Nunn and Trefler (2014) identify sources of comparative advantage for countries and regions in trade as the formation of institutions. In their paper, they suggest that the formation of formal institutions (e.g., legislation) and informal institutions (e.g., reputation) through repeated transactions and human networks lowers transaction costs and provides an institutional comparative advantage. However, they provided no specific consideration of the mechanisms that such networks use to solve the prisoner’s dilemma. Reducing transaction costs through networks can be explained by the effects of intermediation, as discussed in this paper.

3.4 Institutional Background of Innovation in China

39

Table 3.2 Repetitive game for prisoner’s dilemmas (×(1–λ), λ > 1/9) All-C

All-D

Trigger

Tit for tat

All-C

(5, 5)

(−4, 6)

(5, 5)

(5, 5)

All-D

(6, − 4)

(−3, − 3)

(6 − 9λ, − 4 + λ)

(6 − 9λ, − 4 + λ)

Trigger

(5, 5)

(−4 + λ, 6 − 9λ)

(5, 5)

(5, 5)

Tit for tat

(5, 5)

(−4 + λ, 6 − 9λ)

(5, 5)

(5, 5)

Source Okada (2011, p. 233). Note The gains for each player shown in the table are multiplied by (1–λ) for convenience.

situation in which both parties receive gains is represented by (−3, − 3), which is a Nash equilibrium. However, this situation is not Pareto optimal because both parties’ gains can be improved by selecting cooperation for players 1 and 2. As long as the gains realized by cooperation among the players and betrayal are as portrayed in Table 3.1, avoiding the prisoner’s dilemma in only one game is impossible. However, if the transactions between players 1 and 2 are repeated, the cooperative actions of both players are in a Nash equilibrium; therefore, the prisoner’s dilemma can be avoided. In such a game that repeats itself, each player adopts a strategy that prescribes in advance the type of action to be taken based on the history of the opponent and his/her playing style. Under the condition that the transaction is carried out for a considerable period and each player sufficiently predicts the gains that would be obtained through future transactions, the retaliatory mechanism works effectively, and the risks associated with the transaction are avoided.9 When the player does not satisfy such conditions, that is, if the transaction is performed only in the short term and the player also evaluates the future gain to be lower than the current gain, the combination of cooperative actions is taken as a Nash 9

The following four strategies are representative of players (Okada, 2011, p.232). Players’ strategies in a repetitive game: (1) all-C: always cooperate regardless of the result in the past play; (2) all-D: always betray regardless of the result in the past play; (3) Trigger: cooperate first; subsequently, continue cooperating as long as the opponent also cooperates. However, even if the opponent betrays even at least once, he or she continues to betray; and, (4) tit for tat: cooperate-first; subsequently, take the same action as the opponent’s previous action. Let the discounted present value R of the sum of the gains be expressed as the average gain if the player obtains the same gain every period. Then, R is expressed as follows, assuming that the game would be played infinite number of times in the future, where λ is the discount factor for R future gains (0 < λ < 1): R + Rλ + Rλ2 + ... = 1−λ Based on this formula, when the gains are realized by cooperation among the players, betrayals are expressed, as in Table 3.2, when each player adopts strategies (1) to (4). From this equation, if the value of λ is sufficiently large (if λ  1/9 in Table 3.2), that is, at present, the player highly evaluates the gain obtained over the long term (trigger, trigger) or a combination of strategies of (Tit for Tat, Tit for Tat) are also Nash equilibriums. Thus, combination of cooperative actions can be played over a long period.

40

3 Evaluating the Sustainability of Innovation in China from the Perspective …

Table 3.3 Nash equilibrium in the presence of an intermediary

Conditional cooperation (C)

Betrayal (D)

Conditional cooperation (C)

(5 − δ, 5 − δ)

(−3 − δ, − 3)

Betrayal (D)

(−3, − 3 − δ)

(−3, − 3)

Source Prepared by the author. Note If δ < 8, both D-D and C–C are in equilibrium.

equilibrium. Therefore, the mechanism of cooperative action is examined for cases wherein the counterparty’s reputation can be obtained through third-party mediation in a one-time transaction. Players may be bifurcated into two groups—one than betray in any case (Defect D), and another that cooperates in all cases (Conditional Cooperation C) except if the counter partner belongs to the type D classification. Each player can determine the type of party to be traded by obtaining the opponent’s reputation at a cost δ. Table 3.3 shows the combination of each player’s strategy and gain when such a mediation exists. Table 3.3 also shows that if δ is sufficiently small (if it is less than 8), the combination of the two strategies (C, C) and (D, D) is a Nash equilibrium. The dominant equilibrium in that society depends on the value of δ and the distribution of C and D in the initial condition. Figure 3.2 delineates the relationship between α and the gains earned by players from each type (πC and πD). We assume that the value of α when πC and πD

Fig. 3.2 Relationship between intermediary costs and a cooperative equilibrium. Notes 1. When α > α*, the cooperative solution is dominant in society. When α < α*, the treacherous solution is dominant. 2. As the intermediary cost δ decreases, α* decreases, and society as a whole is more likely to move toward a coordinated solution. Source Prepared by the author based on Bowles (2003, p. 263)

3.4 Institutional Background of Innovation in China

41

are equal is α*. Moreover, πC > πD when α > α*, and members are considered to be change their behavioral patterns from D to C as the game progresses. This phenomenon gradually increases the cooperative equilibrium wherein all members of the organization eventually end up qualifying in the C category. In contrast, the converse mechanism is applicable when α < α*., In this case, the betrayal equilibrium wherein all members belong to the D group is realized. Additionally, as the brokerage cost δ decreases, the value of α* as the threshold decreases, and society is more likely to move toward a cooperative equilibrium (Fig. 3.2).10

3.4.4 Historical Institution of the Brokerage Economy As previously described, in the case of mediation (reputation), the third party’s low intermediary cost of mediating information functions as a mechanism to avoid betrayal. This mechanism has been observed to act as an essencial element. Although the actual economy is not a function of only these mechanisms, combining the three mechanisms of retaliation, reputation (and mediation), and decomposition prevents the problem of betrayal under incomplete contracts. Greif (2006), a leading expert on historical institution analysis, states that Maghreb merchants used the game theory to establish contracts with agents and carry out remote area trade in the 12th-century Mediterranean region. When compulsion would be difficult, that is, transactions would be incomplete, cooperative trading between merchants and agents would take place by means of multi-party punitive mechanisms (Greif, 2006: Chap. 9). Maghreb merchants exclude agents having engaged in betrayal behavior even once in the past, by the development of a human network that shared the information on the history of commercial transactions. A punitive mechanism between the parties was established. This system was aimed at providing the agent with an incentive to comply with the contract or relinquish his long-term benefits after a betrayal (although he may reap the benefits of a betrayal at certain rare instances). The greatest contribution by Greif (2006) to research is the finding that cooperation in transactions is realized even if the enforcement of contracts by a public authority is largely ineffective. The punishment mechanism among Maghreb merchants, as described by Greif (2006), may be perceived as a combination of two mechanisms, namely, (i) retaliation (punishment) on the premise of long-term transactions and (ii) the clarification of reputation by means of information sharing. Meanwhile, Genoa merchants did not possess an information network that shared merchant information to clarify their reputations. In other words, when an agent would engage in betrayal, he would not 10

Bowles (2003) used a more general model of the co-evolution of contracts and behavior to show that individuals in societies that have more economic actors engaged in cooperative behavior and low transaction costs tend to prefer incomplete contracts more. Conversely, complete contracts are preferred in societies in which fewer actors are engaged in cooperative behavior and transaction costs are high.

42

3 Evaluating the Sustainability of Innovation in China from the Perspective …

deal with the Genova merchant he had betrayed; rather, he would strike a new deal with another Genoa merchant. Consequently, Genoa merchant had to spend larger amounts to hire agents and undertake a greater risk of betrayal compared to Maghreb merchants. Greif (2006) specified that in the case of this Genova merchant, a formal contract enforcement authority such as a legal and judicial system must be established to promote the trade. Traditional Chinese business customs have evolved through reputation and brokerage. So, compared with the case of Maghreb merchant described by Greif, multilayered development based on shorter-term trading is possible. This feature is thought to be characterized by the fact that the third party’s intermediary over the long term has realized cooperation between traders.

3.5 Contemporary Significance of the Brokerage Economy It is noteworthy that considering the recent progress of the sharing economy, a platform for trading—such as one established by a huge Internet entrepreneur (e.g., Alibaba and Tencent)—will exist. The business model for such a platform is rapidly spread in China. This possibility demonstrates that the brokerage economy that established itself in the modern society as traditional business values molded by support from the latest technological interventions are marking their presence in the contemporary Chinese society. Additionally, as technological standardization advances in the manufacturing industry, particularly in areas characterized by significant technological innovation, modern business partners exploring their avenues beyond long-term business associations with specific companies have begun proposing that contracting is a more efficient—and therefore, preferable—process. This perspective also reflects the contemporary approach that economic development and progress in terms of institutionalization can replace brokerage with credit. Economy of modern China, among the revitalization movements of the brokerage economy, is characterized as the insufficient rule of law and a market characterized by considerably high uncertainty. Alibaba Group and Tencent represent the popularization of the so-called platform economy by acting as reliable information mediators providing suitable platforms, and establishing stable transactions. Credit transactions, including credit settlement, involving private economic entities are extremely underdeveloped in China. Legal systems supporting crosscompany credit—including a disadvantage system that imposes sanctions such as the suspension of bank transactions on defaulting companies—are inadequate in China. Considering the present industrial structure wherein microenterprises engage in successively long-term transactions, the difficulty in connecting a relationship has also been a factor impeding credit transactions. Alibaba’s innovation is a society in which margin trading is underdeveloped, offering its own payment system (support treasure = Alipay) wherein the first

3.5 Contemporary Significance of the Brokerage Economy

43

Fig. 3.3 Transaction brokerage through Alipay. Source Prepared by the author

execution of a transaction would result in cash exchange. This innovation would contribute in overcoming the challenges posed by credit transactions (see Fig. 3.3).11 In other words, Alipay fulfills the role of third-party mediation in transactions, realizes smooth transactions by reducing the psychological burden faced by both the parties in net transactions upon facing credit risks, and rapidly trims small retail transactions on the Internet. In addition, the mutual evaluation system through the Internet provides useful information to help find genuine individuals with whom to engage in trading against the backdrop of a market structure that is typically rushed owing to homogenized companies. This business model is also consistent with traditional Chinese business practices that have emphasized the intermediator’s role. In other words, Alipay acts as a guarantor of a transaction in a village. If everyone trusts a guarantor in a village, villagers shall engage more actively in transactions, and all individuals shall partake in enjoying the merits associated with trading. In other words, the provision of brokerage services by means of the accumulation of a considerable volume of customer information by IT companies such as Alibaba involves the modification of traditional Chinese business practices through modern technological interventions. The spread of Internet transactions through a platform provided by such a huge enterprise aims to provide consumers with quick access to the latest products and entails the possibility of acquiring more markets through word of mouth. This platform plays the role of reducing the costs involved in product development and manufacturing. It facilitates innovation through practices such as helping individuals to efficiently find suppliers of reliable intermediate goods.

11

McDonald (2020) compares the Alipay and WeChat Pay functions in the Chinese society from an anthropological perspective through a survey of peasant workers in Shenzhen. Many workers perceived Alipay as reliable in terms of relatively large-scale monetary exchanges such as online purchases through the intervention of a “third party.” In contrast, the latter is used to facilitate communication between people involved in close social relationships. Such facilitation of communication is evident in the practice of exchanging hongbao (紅包) through WeChat. McDonald (2020) observes that a “summation” exists between the two types of communication.

44

3 Evaluating the Sustainability of Innovation in China from the Perspective …

3.6 Conclusion Given these considerations, is China’s innovation sustainable? We return to the question posed at the beginning of the book. Indeed, China’s traditional society was dominated by short-term transactions among small firms, hindering innovation and creative pursuits. Even today, shortterm transactions persist at almost the same rates among small firms. However, the rise of companies such as Alibaba that provide Internet platform services has significantly altered the technological means adopted for carrying out such short-term transactions. Such changes in the trading environment have increased the sustainability of innovations in Shenzhen and other large cities as conceptually illustrated in Fig. 3.4. As mentioned in the previous section, the expansion of IDHs—which has taken place due to IDHs having served as intermediaries between companies in Shenzhen— was triggered by the exponential growth of shanzhai mobile phones earlier this century. These IDHs developed counterfeit business activities by mediating between amateur mobile operators and miscellaneous manufacturers of microparts. By acting as an intermediary between amateurish mobile phone operators and ragtag microparts companies, which entered the market to win a fortune, the IDHs

Fig. 3.4 Intellectual property rights: The mutual relationship between the three IPR groups Source Prepared by the author

3.6 Conclusion

45

could engage in puckering a viable business. In the environment in which companies can only engage in the act of pirates, they can easily enter the market. This is also an environment where they can save on the fixed costs of product development. In other words, because counterfeit activities were so prevalent, a system was built autonomously to make such activities viable as a business, and this system was a mechanism that supported creative companies. The pre-modern group of companies might appear not to be connected with IPR; however, it provides a kind of background for creating an ecosystem that supports post-modern startups. In contrast, an insurmountable technological barrier exists between the other two tiers and large R&D-oriented companies (e.g., Huawei), which are in the modern tier. Nevertheless, the pre-modern and post-modern tiers are becoming more active through the development of infrastructure, such as networks and e-commerce provided by Huawei. As depicted in Fig. 3.4, the pre-modern, modern, and post-modern groups are considered complementary to each other. In other words, the three groups are, in fact, complementary to one another (Fig. 3.1). Contemporary China is experiencing unprecedented advancements resulting from technological innovations. In some respects, these advancements have even surpassed those witnessed in Japan. In contrast, certain sections of society are almost entirely beyond the reach of modern institutions, including the rule of law and the values that support them. Such lack of uniformity in development has given rise to a kind of dynamism and lopsided social progress. A discussion on contemporary Chinese society and economy would be impossible without consideration of these aspects. In modern China, short-term transactions through the Internet have supported innovations. Meanwhile, the current Chinese political and economic system has facilitated the informality of the private economy. Such informality has been established by overthrowing the rules established by power. Moreover, the current system has proactively utilized the resultant diversity as a useful tool to maintain its regime. As described in this review, active innovations have occurred in a situation wherein some allowance has been made toe protect IPR. This phenomenon might be because a certain complicit relationship exists among the authoritarian government, non-democratic society, and the free and vigorous private economy. When considering innovations in China, these problems might also have to be considered in the future.

References Acemoglu D, James R (2013) Why nations fail: the origins of power, prosperity and poverty. Currency Acemoglu D, James R (2019) The narrow corridor: states, societies, and the fate of liberty. Viking Anderson C (2012) Makers: the new industrial revolution. Crown Business Adachi K (2018) Historical argument about tyranny. Chikuma-Gakugei-Bunko (in Japanese) Bowles S (2003) Microeconomics: behavior, institutions, and evolution. Princeton University Press Fujioka J (2017) Lessons from “Hardware’s Silicon Valley, Shenzhen”: upcoming manufacturing trends and ecosystem. East Press (in Japanese)

46

3 Evaluating the Sustainability of Innovation in China from the Perspective …

Furuta K (2013) Market order and information asymmetry in modern China: Late 19th century— early 20th century, In: Furuta K (eds) Market order in China: from the 17th century to the first half of the 20th century. Keio University Press (in Japanese) Gorodnichenko Y, Roland G (2012) Understanding the individualism–collectivism cleavage and its effects: Lessons from cultural psychology, In: Masahiko A, Kuran T, Roland G (eds) Institutions and Comparative Economic Development, Springer Gorodnichenko Y, Roland G (2017) Culture, institutions and the wealth of nations. Rev Econ Statistics, 99:402–416 Greif A (2006) Institutions and the path to the modern economy: lessons from Medieval Trade. Cambridge University Press Hioki S, Ding K (2018a) The role of a technological platform in facilitating innovation in the global value chain: a case study of China’s mobile phone industry, IDE Discussion Paper, 692 Hioki S, Ding K (2018b) Knowledge and information acquisition of cluster firms through personal networks and value-chain linkages: a case study of China’s mobile phone manufacturing industry, IDE Discussion Paper, 698 Huang AB (2017) The hardware hacker: adventures in making and breaking Hardware. No Starch Press Kato H (2016) The introduction of China economics: how do ambiguous institution work. Nagoya University Press (in Japanese) McDonald T (2020) ‘Social’ money and working-class subjectivities: digital money and migrant labour in Shenzhen, China. China Q 242:397–417 Marukawa T (2013) Chinese dream. Chikuma-Shinsho (in Japanese) Nunn N, Trefler D (2014) Domestic institutions as a source of comparative advantage, In: Gopinath G, Helpman E, Rogoff K (eds) Handbook of international economics, vol. 4. North Holland Okada A (2011) Game theory: Second Edition. Yuhikaku (in Japanese) Raustiala K, Sprigman C (2012) The knockoff economy: how imitation sparks Innovation. Oxford University Press Roland G (2018) Comparative economics in historical perspective: presidential address for the 2018 association for comparative economic studies meetings. Comp Econ Stud 60:475–501 Stiglitz JE, Greenwald BC (2014) Creating a learning society: a new approach to growth, development, and social progress. Kenneth J. Columbia University Press, Arrow Lecture Series Takasu M, Takaguchieds K (2020) Prototype city: Shenzhen and worldwide innovation. KADOKAWA (in Japanese) Zhang X (2014) The development process of China’s corn industry. Keiso Shobo (in Japanese)

Chapter 4

Political Capital and Enterprise Performance in China: Empirical Analysis Using Data from the People’s Congress of Yangzhou City and Industrial Enterprise

4.1 Introduction: Local People’s Congress as “Chinese-Style Democracy” Despite being an authoritarian state, China also has “democratic institutions” aimed at reflecting the “people’s will” in politics at the central and local levels albeit to a limited extent. Elections are regularly held to elect delegates. Recent studies on Chinese politics, such as Kamo (2013), Kamo and Takeuchi (2013), and O’Brien (1994, 2009), have demonstrated how the Chinese Communist Party (CCP) pursues its survival by designing incentive mechanisms to co-opt and manage key players, such as entrepreneurs, local officials, and various groups of citizens. According to O’Brien (1994), the representatives of the People’s Congress play the role of “agent” by communicating the policies of the CCP and local governments to their own electoral districts. O’Brien additionally pointed out that the representatives of the People’s Congress work as “remonstrators.” In other words, delegates of the People’s Congress who are familiar with the situation of the constituency will provide the necessary information for policy planning to the CCP and local governments as well as convey policy inequities and mistakes of the local governments to their constituency. Kamo (2013), using detailed data on the Yangzhou Municipal People’s Congress (MPC) obtained from its website, insisted that delegates of the People’s Congress acted not only as “agents” or “remonstrators” by the CCP and local governments but also as “representatives” of constituencies. According to these previous studies, local people’s congresses’ decisions regarding the allocation of policies and funds have been closely related to the profitinducing behavior of the lower-level region. If this is true, how do the elections of local council members as “political capital” affect its economic performance at the local company level? Many of the previous studies have examined the effects of political factors on firm performance. Li et al. (2008), using a nationwide survey of private firms, found that

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 K. Kajitani and T. Kamo, Political Economy of Reform in China, Kobe University Social Science Research Series, https://doi.org/10.1007/978-981-19-0202-4_4

47

48

4 Political Capital and Enterprise Performance in China …

the Party membership of private entrepreneurs had a positive effect on the performance of their firms, and that Party membership helped private entrepreneurs obtain loans from banks or other state institutions and afforded them more confidence in the legal system. Wu et al. (2012), using data on listed Chinese firms, found that private firms with politically connected managers enjoyed tax benefits and outperformed those without such managers. Cao et al. (2017), using a sample of listed non-stateowned enterprises (SOEs), showed that politically connected chief executive officers (CEOs) had a lower probability of turnover, leading to weaker turnover–performance sensitivity than nonpolitically connected CEOs. Ling et al. (2016), using a sample of 103 listed real estate firms from 1998 to 2012, examined the influence of firms’ political connections on external financing. They demonstrated that political connections were negatively related to return on assets and that firms with stronger connections were financed with more long-term bank loans. Boubakri et al. (2012) examined the cost of equity capital of politically connected firms using a propensity score matching model. They found that politically connected firms enjoyed a lower cost of equity capital than their non-connected peers. Liu et al. (2016) investigated how managerial, political, and professional connections played a role in helping firms obtain trade credits. They documented the fact that non-SOE firms with professionally well-connected managers received more trade credit. Of course, political capital often has a negative effect on the performance of firms as well. Chen et al. (2011) investigated how the rent-seeking incentives of local governments motivated private firms to establish political connections and whether such connections led to more concentrated corporate control structures. He, Wan, and Zhou (2014) analyzed the stock market’s reaction to CEO succession and found that there were cumulative abnormal returns when CEO succession was accompanied by increased political connections. Cheng (2018) tried to estimate the value of the political connections that independent directors of private firms have by examining stock price reactions to the sudden death of retired government officials acting as independent directors of private firms. Most previous studies used the political connections of individuals in a firm as proxy, e.g., the firm manager or CEO who is also a member of the CCP or the National People’s Congress and focused on the effects of these political connections on firm economic performance using the data of listed firms. However, listed firms account for less than 1 percent of all Chinese firms, so we cannot consider this sample size to be representative of the general situation of Chinese firms. Most Chinese firms have a close relationship with the local rather than the central government. It would, therefore, make sense to investigate the effect of localized political connections. Nevertheless, exceptional studies also exist. Cheng et al. (2019a), using the data of China Employer-Employee Survey (CEES), found that innovation subsidies are preferentially allocated to state-owned firms and politically connected firms with National or Local People’s Congress or Chinese People’s Political Consultative Conference (CPPCC) membership, which is a proxy for political favoritism. Cheng et al. (2019b), also using CEES data, established that having political connections

4.1 Introduction: Local People’s Congress as “Chinese-Style …

49

could help firms obtain economic benefits such as tax preferences, government subsidies and that political connections contributed to innovative activities for firms with innovative entrepreneurs. While these studies are excellent, the problem is that the CEES is a three-year study (2013–2015) and therefore, does not capture the effects of changes over a relatively longer time period, such as changes in the political status of managers. In our study, we used data from the MPC of Yangzhou city in Jiangsu Province, focusing on the effect of the connections between the managers of local firms and the local People’s Congress against local firm performance. We also used 10 years of data in our analysis, corresponding to two periods of the MPC to provide an explicit analysis of how the acquisition of new political status of corporate managers affects corporate performance. Firstly, we describe the structure of the Yangzhou MPC. We took Yangzhou City in Jiangsu province as an example of local councils for two reasons. The first is information disclosure: the Yangzhou People’s Congress provides their official website, “Yangzhou People’s Congress” (http://rd.yangzhou.gov.cn/) with published information and materials that are abundant compared with other local councils. The website includes personal data of the People’s Congress’ delegates and materials on the proposals, criticisms, and opinions submitted by delegation members to the People’s Congress. These materials provide important information and clearly show the economic functions of the local council. In addition, Yangzhou City has sufficient representation as a medium-sized city in the coastal area wherein a certain degree of industrialization has occurred. Thus, we considered Yangzhou City suitable for our research. Figure 4.1 shows the structural image of delegates of X Municipal People’s Congress. This figure represents the structure of the People’s Congress of a fictitious city, “X City.” The representative of X City’s People’s Congress will be elected by the People’s Congress of A district, B district, C county, and D county located in the lower administrative class of X City. The elected representatives will act as the representatives of the delegation of each district or county in the X City People’s Congress. Many representatives are senior executives of parties and state agencies or executives of local enterprises. They can be categorized by the delegation to which they belong. The People’s Liberation Army also elects representatives that make up its own delegation. Reports on the activities of government agencies in the X City People’s Congress and deliberations on bills are conducted for each delegation (Kamo 2013). Yangzhou MPC is also composed of delegates of the People’s Congress elected from seven administrative districts under Yangzhou City and elected as representatives of the People’s Liberation Army. The Yangzhou MPC delegation is not elected by residents living in each administrative district but through elections that the representatives of People’s Congresses have established for each of the seven administrative districts under Yangzhou City. Table 4.1 shows the composition of the delegates in the Yangzhou MPC from the fourth to the seventh congress. Table 4.2 shows the number of bills submitted and the recommendations made in each session of the Yangzhou MPC between the fourth and sixth congresses (Kamo

50

4 Political Capital and Enterprise Performance in China …

Fig. 4.1 Structural Image of Delegates of X Municipal People’s Congress. Source Kamo (2013), p. 26

Table 4.1 Composition of the delegates in the Yangzhou MPC 4th period congress

5th period congress

6th period congress

7th period congress

Period

1998–01

2001–06

2007–11

2012–16

Baoying County

64

66

71

74

Gaoyou City

67

66

63

69

Guangling District

78

55

48

54

Hanjiang District

39

41

46

72

Jiangdu District

83

85

85

87

Yizheng City

53

56

54

53

Weiyang District

–

46

47

–

Jiao-Qu

20

–

–

–

PLA

12

13

12

12

Total (CCP)

416(310)

428(327)

426(332)

421(331)

Source Yangzhou People’s Congress official website (http://rd.yangzhou.gov.cn/)

2013). The submitted bills address a wide range of concerns and questions, including cultural and educational issues; environmental protection issues; urban development issues, such as road construction and maintenance, bridge and port construction projects, and requests for establishing economic development zones (jingji kaifa

4.1 Introduction: Local People’s Congress as “Chinese-Style …

51

Table 4.2 Bills and recommendations to the Yangzhou MPC Year

Period

Congress

Bills submitted

Recommendations

2001

Fourth

Fourth

63

224

2002

Fifth

Fourth

51

103

2003

First

Fifth

64

230

2004

Second

Fifth

68

214

2005

Third

Fifth

82

202

2006

Fourth

Fifth

67

222

2007

Fifth

Fifth

62

177

2008

First

Sixth

74

229

2009

Second

Sixth

76

237

2010

Third

Sixth

96

264

2011

Fourth

Sixth

86

253

2012

Fifth

sixth

71

165

Source Yangzhou People’s Congress official website (http://rd.yangzhou.gov.cn/)

qu); and administrative issues, such as taxation, public transportation management, and road management (2013). As Kamo and Takeuchi (2013) have pointed out, most of the bills proposed to the Yangzhou MPC were jointly submitted by delegates from the same district. For example, among the 63 bills proposed in the fourth session of the fourth congress in 2001, 58 bills (92%) were submitted by multiple delegates from a single electoral district. For example, Bill No. 12 in the fourth session of the fourth congress required the municipal government to grant the district government the authority to approve economic projects and confer preferential taxation and other economic measures upon the special economic zone in Hangji Town in Hangjian District (Kamo and Takeuchi 2013). In this case, the delegates who submitted this bill were in a position that would benefit from the economic development of Hangji and Hangjiang. The president and the vice president of the Jiangsu Sanxiao Group, which was the largest and most influential private enterprise in Hangji that would benefit most from the designation of a special economic zone were also included as submitters of the bill. Based on this consideration, we raise the following research question: How does the election of local council members as the “political capital” affect its economic performance at the local company level? We aim to clarify this empirically, using the following steps. First, by matching the data regarding the elected delegates of the Yangzhou MPC with the Chinese industrial enterprises’ database provided by Huamei Information. Next, by empirically investigating whether the manager’s inauguration as a delegate in the People’s Congress changed the performance of the company managed by him or her.

52

4 Political Capital and Enterprise Performance in China …

4.2 The Methodology of Empirical Studies In our research, we used two different kinds of datasets: data on MPC delegates and data on manufacturing enterprises. Regarding the former, the official website of the Yangzhou City’s People’s Congress, which opened in the latter half of the 1990s, contains the personal information of the delegated representatives of the People’s Congress. This website also has published materials on proposals, criticisms, and opinions submitted to the People’s Congress. Therefore, we can access detailed information about the behavior of individual delegates in the Yangzhou MPC, including each delegate’s personal background (name, gender, year of birth, affiliation and employment unit, ethnic identity, education, party affiliation, and other personal information) for each congress. We also have access to the bills proposed by the delegates in each session of each congress. In this research, we use data on the representatives of the Yangzhou MPC in the fifth (2002–2006) and sixth (2007– 2011) sessions, as provided by the official website of Yangzhou People’s Congress (http://rd.yangzhou.gov.cn/). Regarding the latter, we used a database provided by Huamei Information, the most comprehensive database on Chinese mining and manufacturing enterprises. It covers data on production, workers, financial statements, etc., in industrial enterprises with annual sales of 5 million yuan (2011 to 20 million yuan). From this database, we extracted the data on manufacturing enterprises located in Yangzhou City. We then matched the data regarding elected delegates of the Yangzhou MPC with the Chinese industrial enterprises database provided by Huamei Information. Specifically, we checked the names of the managers recorded in the latter database and checked to see if they were listed in the former, and if they were, they were deemed to have been appointed as delegates to the Yangzhou MPC for the relevant year.1 We defined those as the treatment group wherein the enterprises’ managers were elected delegates of People’s Congress in the sixth session and not elected in the fifth. The other enterprises were defined as the control group. Then we determined whether their election as delegates influenced economic performance, such as the profit rates of the enterprises, using difference-in-differences (DID) analysis. Then, we reconsidered the economic function of the local People’s Congress, based on the results obtained from the empirical analysis.2 1

In the Chinese industrial enterprises database, there are many enterprises that changed their names or merged with other enterprises. In such cases, we checked the location of the enterprise and the name of the manager, and counted them as the same enterprise if the continuity of the main business was confirmed. 2 Recent empirical studies that have analyzed policy effects in China using DID analysis include the following: Wen and Zhao (2020) examined the impact of China Manufacturing 2025 (CM2025), an industrial policy aimed at improving the innovation capacity and technological capability of the manufacturing industry, on firms’ R&D investment using DID and Coordinated Exact Matching (CEM). Using a panel data set of 1440 listed companies from 2012 to 2018, they found that R&D investment increased substantially after policy intervention for companies whose business was in the industries targeted by CM2025. Xu et al. (2021) conducted a DID estimation by using

4.2 The Methodology of Empirical Studies

53

Here, we will briefly explain the basic concept of DID analysis. When certain policies are implemented in a specific region, it is not sufficient to compare the expost outcomes among regions for measuring the effect of policy intervention because the improvement in performance may also be seen in areas where the policies were not implemented. Therefore, after policy intervention, we not only compared the outcome differences between the “treatment group” as the object of intervention and the “control group,” which is not subject to intervention but also measure the differences before intervention and tried to obtain an estimate of the pure policy effect by comparing the differences before and after the intervention. DID analysis should satisfy two assumptions: (1) parallel trends and (2) common shocks. The former is the assumption that the treatment group and control group would draw parallel trends if the policy tested had not been introduced. The latter assumes, between the measurements before and after the policy’s introduction, a “separate event” that would affect the outcome had not occurred or, if it had, only works for two groups in the same way. In this study, we also examine whether the data we used meet these conditions. In the next section, we will test three working hypotheses on the relationship between the performance of Chinese local enterprises and their political capital using DID analysis as follows. H1. The performance of enterprises whose managers are elected as the delegates of the MPC will be improved in the terms of volume because profits can be channeled to those enterprises through the introduction of bills and by setting the agenda. H2. Because the effect of profit inducement as described in H1 does not provide incentives for firms to improve their management voluntarily, the performance of firms whose managers are elected as delegates to the MPC will not show any improvement in terms of efficiency. H3. These enterprises can receive preferential treatment from the local branch of the financial institution. This hypothesis seems plausible because the increase in the political capital of enterprises in Chinese local cities will increase its financial credibility and make it easier to obtain loans from the local branch of financial institutions. In Tables 4.3, 4.4, 4.5 and 4.6, we provide the descriptive statistics of manufacturing enterprises subject to our analysis. In Tables 4.3 and 4.4, we provide descriptive statistics of all of the enterprises in our datasets and enterprises that belong to the treatment group. Comparing Tables 4.3 and 4.4, we find that enterprises of the treatment group have more employees, more total assets, higher SOEs ratio, higher labor productivity and profit ratios, and higher paid taxes than the average of all of the samples. These descriptive statistics indicate that the character of the enterprises belonging to the treatment group is quite different from that of the control group. unique provincial-level data set on coal mines from 2004 to 2015 to investigate the effect of an anti-corruption campaign on fatality rates in the coal-mining industry. They found evidence that provinces with stronger exposure to the anti-corruption campaign experienced significantly larger decreases in coal mine death rates.

54

4 Political Capital and Enterprise Performance in China …

Table 4.3 Descriptive statistics of all enterprises, 2002–2011 Variable

Definition

Obs

Employees

Number of Employees at the end of year

23,695

183.62

385.10

1.00

12,840.00

Years

Years that the company has been running

23,695

9.87

9.41

0.00

208.00

State-owned Enterprises

Variable = 23,695 1 if the company is SOE

0.01

0.10

0.00

1.00

Total assets (1000 yuan)

Total of fixed assets, liquid assets, and intangible assets owned by the company

23,695

300.10

739.17

1.51

38,290.88

Total debt (1000 yuan)

Total of long-term debt, current liabilities, accounts payable

23,695

27,936.20

125,384.20

−10,738.77

5,774,512.00

Operation profit (1000 yuan)

Operation 23,695 profits from the main business of the company

4307.79

21,527.35

−97,241.00

1,148,881.00

23,695

93,465.87

352,451.70

18.00

17,000,000.00

Total taxes 23,695 paid by companies for the year

3273.11

13,560.89

−666,195.00

531,075.00

Main business Total main revenue(1000 business yuan) revenue Tax payment (1000 yuan)

Mean

Std. dev

Min

Max

(continued)

4.2 The Methodology of Empirical Studies

55

Table 4.3 (continued) Variable

Definition

Obs

Profit per labor (1000 yuan)

The total added value per number of total employees

23,695

27.56

62.83

−763.28

1341.76

Revenue per labor (1000 yuan)

Main 23,695 business revenue per number of total employees

630.66

880.02

0.40

22,717.95

Profit ratio (%)

Ratio of total Profit against total sales

2.96

11.44

−938.89

40.69

23,695

Mean

Std. dev

Min

Max

Source Chinese industrial enterprises database, provided by Huamei Information

In Tables 4.5 and 4.6, we provide descriptive statistics of the enterprises belonging to the treatment group in the pre-treatment and post-treatment periods. From these descriptive statistics, we easily find that some figures on the economic performance of treated enterprises have explicitly improved after treatment. However, as can be seen in Tables 4.3 and 4.4, the nature of the enterprises in the treatment group has a clear bias compared to the other enterprises in the analysis, so it is not possible to determine whether the improved performance of these enterprises can be explained by the impact of management’s appointment as the representative of the MPC. So, we try to analyze whether election as delegates of MPC influenced economic performance in a more stringent manner by using the following induction model of DID analysis. Per f or manceit = β1 T r eati ∗ Postt + β2 Postt + β3 Enter priseContr olsit + β4 T + μi + ∈it (4.1) In Eq. (4.1), we use asset per capita, debt per capita, operating profit (per capita), the profit-to-revenue ratio, total main business revenue (per capita), and tax payment3 as the dependent variables (Performanceit). As the independent variable, Postt, which is also a dummy variable, takes the value of 1 for the data after 2007, and Treat i × Postt is the cross term of these two variables. EnterpriseControlsit stands for other variables specific to each enterprise, including SOE dummy, two-digit industry classification dummy, years in business, number of employees, and total assets per 3

Total taxes paid by the enterprises for the year include value-added tax and corporate income tax.

56

4 Political Capital and Enterprise Performance in China …

Table 4.4 Descriptive statistics of enterprises in the treated group, 2002–2011 Variable

Definition

Employees

Number of 225 Employees at the end of year

1054.91

1248.07

16.00

6500.00

Years

Years that the 225 company has been running

15.21

13.78

1.00

55.00

State-owned enterprises

Variable = 1 if the company is SOE

225

0.04

0.21

0.00

1.00

Total assets (1000 yuan)

Total of fixed assets, liquid assets, and intangible assets owned by the company

225

1254.84

5090.60

15.18

38,290.88

Total debt (1000 yuan)

Total of long-term debt, current liabilities, accounts payable

225

272,285.70

615,506.10

318.00

5,774,513.00

Operation profit (1000 yuan)

Operation profits from the main business of the company

225

54,820.07

140,602.80

−29,805.25

1,148,881.00

Main business revenue(1000 yuan)

Total main business revenue

225

603,466.10

976,022.70

6700.00

7,644,613.00

Tax payment (1000 yuan)

Total taxes paid by companies for the year

225

24,274.57

49,016.84

−135,530.00

322,965.00

Profit per labor The total (1000 yuan) added value per number of total employees

225

49.44

106.28

−152.00

703.65

Revenue per labor (1000 yuan)

225

794.77

1108.70

36.35

8244.79

Main business revenue per number of total employees

Obs

Mean

Std. dev

Min

Max

(continued)

4.2 The Methodology of Empirical Studies

57

Table 4.4 (continued) Variable

Definition

Profit ratio (%) Ratio of total Profit against total sales

Obs

Mean

225

Std. dev 5.55

Min

Max −36.30

6.65

28.94

Source Chinese industrial enterprises database provided by Huamei Information

Table 4.5 Descriptive statistics of the treatment group in pre-treatment Variable

Obs

Mean

Std. dev

Min

Max

State-owned enterprises

101

0.05

0.22

0.00

1.00

Years

101

14.60

13.92

1.00

50.00

Employees

101

935.16

1072.21

16.00

4925.00

Total asset (1000 yuan)

101

702.54

3792.07

15.18

38,290.88

Total debt (1000 yuan)

101

174,558.90

331,282.10

318.00

1,935,092.00

Operation profit (1000 yuan)

101

22,432.28

51,979.29

−8604.00

307,051.00

Main business revenue(1000 yuan)

101

349,612.20

531,705.50

6700.00

3,204,087.00

Tax payment (1000 yuan)

101

8925.20

22,782.01

−135,530.00

97,947.00

Profit per labor (1000 yuan)

101

17.93

34.49

−152.00

158.11

Revenue per labor (1000 yuan)

101

475.11

455.15

36.35

2467.65

Source Chinese industrial enterprises database provided by Huamei Information

capita. In the estimate, the time trend variable(T) and the individual effects of the enterprise (µi) are also included.

4.3 Results and Discussion Now we can see the estimated results in Table 4.7 in which we use the panel data from 2002 to 2011. From the results in these tables, only the estimates in which the profit per capita, revenue per capita, and tax payments are dependent variables are significantly positive.4

4

We adopted a fixed-effects model for the estimation; the dummy variable for the treatment group was omitted in the estimation.

58

4 Political Capital and Enterprise Performance in China …

Table 4.6 Descriptive statistics of the treatment group in post-treatment Variable

Obs

State-owned enterprises

124

Mean

Std. dev 0.04

Min 0.20

Max 0.00

1.00

Years

124

15.70

13.70

2.00

55.00

Employees

124

1152.45

1371.39

16.00

6500.00

Total asset (1000 yuan)

124

1704.71

5920.13

26.11

35,834.67

Total debt (1000 yuan)

124

351,885.80

765,818.70

1458.00

5,774,513.00

Operation profit (1000 yuan)

124

81,200.45

179,562.80

−29,805.25

1,148,881.00

Main business revenue(1000 yuan)

124

810,234.10

1,187,009.00

18,840.00

7,644,613.00

Tax payment (1000 yuan)

124

36,776.88

60,021.24

−16,535.00

322,965.00

Profit per labor (1000 yuan)

124

75.11

134.62

−49.10

703.65

Revenue per labor (1000 yuan)

124

1055.14

1384.89

72.48

8244.79

Source Chinese industrial enterprises database provided by Huamei Information

Now we investigate the three working hypotheses, which were presented in the last section: First, the total operating profit and main business revenue of the treatment group’s enterprises evidently increased in the sixth session of the People’s Congress, which was after treatment. These results mean that the performance of enterprises whose managers were elected as the delegates of MPC had improved, and thus, H1 is supported. After the managers were elected as delegates of the MPC, the tax payment of the treatment group’s enterprises also significantly increased. This result is consistent with H1 because firms’ tax payments are closely linked to the size of firms’ sales and profits. Second, as opposed to total profits and sales, the coefficients of DID on profit and revenue per employee, or profit per revenue of the treatment group’s enterprise, were not significant. These results mean that H2 is supported. Third, the total debt of the treatment group’s enterprises significantly increased in the sixth period of the congress. This result means that H3 in which these enterprises could have received preferential treatment from the local branch of financial institutions is supported. Total debt includes long-term and short-term debt, especially the expansion of long-term debt, which will not occur without an increase in financing from financial institutions. This result also is consistent with the above implication that the election of business owners to seats on the council resulted in an increase in scale rather than an

(390.36)

31.900*** (8.95)

655.890 (327.90) −3142.320 (4840.12) 21.116 (26.76) 33.828*** (7.47) 3.446*** (0.97) 673.029*** (77.32)

(1112.28)

−31,060.14

(31,734.54)

−1.145

(96.88)

158.643***

(40.14)

22.898***

(5.42)

1732.595***

(262.40)

y

Fixed effect

23,695

4990

Industrial dummy

Method

Number of obs

Number of groups

Time trend

Total asset /labor

Employees

Years in business

State-owned enterprises

4990

23,695

Fixed effect

y

−141.449

(28,856.63)

(86,681.72)

−971.837

Post

4990

23,695

Fixed effect

y

(730.96)

16,498.87***

(123.60)

338.282***

(71,556.78)

−44,435.920

(2937.32)

1697.720

(133,302.60)

366,639.600***

58,434.52**

194,699.100**

(3) Main business revenue

Treated* Post

(2) Operating profit

Total debt

Dependent valuable

(1)

Table 4.7 Results of DID analysis, 2002–2011 (4)

4990

23,695

Fixed effect

y

(51.02)

377.834***

(0.21)

0.359*

(5.43)

15.295***

(16.43)

−8.373

(3733.13)

−4191.295

(268.32)

1428.14***

(8012.89)

22,554.620***

Tax payment

(5)

4990

23,695

Fixed effect

y

(0.06)

0.434***

(0.00)

0.001**

(0.00)

0.002**

(0.03)

−0.004

(2.015)

−3.033

(0.35)

−0.681*

(1.36)

1.189

Profit/revenue

(6)

4990

23,695

Fixed effect

y

(0.22)

6.154***

(0.01)

0.053***

(’0.00)

0.010*

(0.06)

0.208***

(8.14)

6.416

(1.88)

−5.325***

(21.17)

28.639

Profit/labor

(7)

4990 (continued)

23,695

Fixed effect

2.29

(2.29)

102.982***

(0.17)

0.792***

(0.04)

−0.086**

(0.87)

2.414***

(80.94)

121.655

(22.41)

−73.670***

(207.34)

93.619

Revenue/labor

4.3 Results and Discussion 59

(1)

0.2320

(2) 0.173

(3) 0.014

(4) 0.100

Note 1: Figures in parentheses are standard errors Note 2: *** , ** , * denote significance at the 1 percent, 5 percent, and 10 percent levels, respectively.

R-sq (within)

Table 4.7 (continued) (5) 0.217

(6) 0.335

(7) 0.515

60 4 Political Capital and Enterprise Performance in China …

4.3 Results and Discussion

61

improvement in efficiency of the enterprises. This is because if the increase in political capital led to an increase in debt from financial institutions, then this process seems to be not necessarily efficient. Overall, while the acquisition of political capital by local enterprises has the effect of increasing the “scale” of the enterprise, including the number of employees, our analysis did not confirm the effect of improving productivity and efficiency.

4.4 Robustness Check In the next step, for the test of the parallel trends and common shock assumption of the last section’s analysis, we resorted to an event-study approach. Specifically, we investigated the panel analysis as we did in the previous section, only using the data from 2002 to 2006, with the interaction of each year dummy variable and the treatment dummy variable included in the regression instead of the treatment dummy variable and post-treatment dummy variable. If the assumptions of parallel trends and common shock were true, we should see no significant effects for the interaction terms before the treatment year. We can see the results of the test in Table 4.8. The results in Table 4.8 confirm that the parallel trends and common shock assumption is true for the total debt and profit rate, as none of the interaction terms are dominant. However, for the other variables, no clear results were obtained, and the parallel trends and common shock assumptions could not be confirmed. Therefore, it is undeniable that the enterprises in the treatment group improved their performance, especially their management size, even before they were elected as delegates of the MPC. For this reason, we followed the description in Chap. 5, pp. 238–224 in Angrist and Pischke (2009), following which we added the enterprise-specific time trend trends to the regression equation as an improvement. Specifically, we added the intersection term between the identification numbers assigned to each enterprise and the variable representing time trends to the regression equation.5 The improved estimation equation is shown in the following equation. Per f or mance = β1 T r eati ∗ Postt + β2 Postt + β3 Enter priseContr olsit + β4 I Di ∗ T + μi + ∈it (4.2) In Eq. (4.2), IDi*T means the enterprise-specific time trend coefficient, which is the intersection term between the ID of each enterprise and the time trend, which will capture the linear trends in enterprise-level characteristics that influence the dependent variables. The estimation results of Eq. (4.2) are shown in Table 4.9. The coefficients, signs, and significance of the main variables, including the intersection terms of Treat and Postt, in the estimating equations shown in Table 4.9 are almost the same as those in Table 4.7. In other words, even if we take into account 5

For more information on this method, also see Friedberg (1998).

Industrial dummy

Total asset /labor

Employees

Years in business

State-owned enterprises

Treated* 2006

Treated* 2005

−109.938

−4.300 (13.34) 20.376 (12.62) 1.010* (0.53)

−3.021

(56.210)

124.515***

(44.403)

7.812*

(4.52) y

11.774

(171.78)

(6835.47)

y

(6.34)

−131.608

−5800.155

(49,903.77)

y

(8.63)

1.123

(243.43)

(87,737.58)

−94,435.070

y

(0.38)

0.511

(7.48)

3.745

(2056.33)

−1552.383

(6440.51)

−50,330.800

(169,735.4)

(17,415.08)

7684.002

(3664.33)

6769.390*

(3488.99)

8388.826**

(2037.09)

2582.599

(64,617.62)

408,520.500**

(114,978.5)

279,201.200**

(78,651.31)

125,375.400

(39,311.67)

81,741.770**

(4) Tax payment

30,732.900*

(10,152.69)

(54,806.00)

(3) Main business revenue

90,833.610

19,465.530*

(6268.48)

(38,584.50)

68,169.710

8326.391

(4021.50)

30,380.220

(18,709.78)

Treated* 2004

5752.693

21,465.380

Treated* 2003

(2) Operating profit

Total debt

Dependent valuable

(1)

Table 4.8 Test of parallel trends assumption

y

(0.00)

0.001

(0.00)

0.007

(0.06)

−0.056

(2.81)

−2.444

y

(0.01)

0.014

(0.00)

0.004

(0.03)

0.010

(2.98)

0.703

(8.37)

14.739*

(1.92)

(6.01)

−0.784

9.568

(1.47)

(4.67)

−0.229

0.434

−0.863 (1.54)

(2.49)

(1.25)

2.218

−0.357

(6) Profit/labor

(5) Profit/revenue

(7)

y (continued)

(0.14)

0.280**

(0.09)

−0.244***

(0.89)

−0.447

(37.72)

32.099

(100.53)

177.110*

(64.17)

102.686

(69.18)

−28.974

(30.86)

1.070

Revenue/labor

62 4 Political Capital and Enterprise Performance in China …

0.192

R-sq (within)

0.116

3,015 0.090

3,015

y

(1621.55)

13,832.520***

(3)

0.004

3,015

y

(84.38)

249.503***

(4)

Note 1: Figures in parentheses are standard errors Note 2: *** , ** , * denote significance at the 1 percent, 5 percent, and 10 percent levels, respectively.

3,015

Number of groups

(78.37)

(332.51) y

299.447***

y

(2)

(1)

2225.911***

Industrial dummy

Time trend

Table 4.8 (continued)

0.012

3,015

y

(0.15)

0.134

(5)

0.126

3,015

y

(0.19)

2.641***

(6)

0.268

3,015

y

(3.54)

77.529***

(7)

4.4 Robustness Check 63

44,834.340*** (6965.92)

(28,783.35) 2737.970***

(86,643.00)

4096.85**

(431.70) 341.848*** (124.50)

(29.40) 34.044*** (7.47) 3.837*** (0.999)

(87.41)

159.136***

(40.03)

23.847***

(5.41)

0.233

R-sq (within)

0.174

4990

23,695

fixed effect

y

0.220

4990

23,695

fixed effect

y

y

(10.01)

39.954***

4990 0.015

0.101

23,695

fixed effect

y

y

(0.00)

0.001**

(0.00)

0.002**

(0.03)

0.022

(2.00)

−3.101

(0.26)

0.390

(1.37)

1.520

Profit/revenue

(5)

4990

23,695

fixed effect

y

y

(0.21)

0.586***

(5.45)

15.425***

(17.05)

21.255

(3765.03)

−4237.825

(323.67)

2634.428***

(7978.31)

22,756.050***

Tax payment

(4)

Note 1: Figures in parentheses are standard errors Note 2: *** , ** , * denote significance at the 1 percent, 5 percent, and 10 percent levels, respectively.

23,695

4990

Number of groups

fixed effect

Method

Number of obs

y

y

Enterprise-specific time trend

y

922.821**

72.289**

123.580

(73,137.35)

(4880.76)

(31,943.21)

−46,929.67

(541.01) −3228.430

(1955.62)

(132,390.70)

−31,295.920

Industrial dummy

Total asset per labor

Employees

Years in business

State-owned enterprises

Post

378,473.400***

Main business revenue

58,814.550**

195,770.200**

Operating profit

Total debt

(3)

Treated* post

(2)

Dependent valuable

(1)

Table 4.9 Results of DID analysis (2002–2011), controlling enterprise-specific time trend

0.336

4990

23,695

fixed effect

y

y

(0.01)

0.056***

0.00)

0.011**

(0.10)

0.567***

(7.19)

5.408

(1.36)

9.176***

(20.78)

33.560

Profit/labor

(6)

0.518

4990

23,695

fixed effect

y

y

(0.17)

0.837***

(0.04)

−0.071*

(1.48)

8.301***

(66.52)

104.569

(14.927)

164.196***

(205.59)

177.479

Revenue/labor

(7)

64 4 Political Capital and Enterprise Performance in China …

4.4 Robustness Check

65

the enterprise-specific time trend is, the hypotheses are supported in our analysis, as in the previous section.

4.5 Conclusion In our study, we first matched the data on elected representatives of the Yangzhou People’s Congress with the Chinese industrial enterprises’ database. Then, we empirically investigated whether the managers’ inaugurations as delegates in the People’s Congress changed the performance of the company managed by him or her, using the method of panel DID analysis. In summary, our analysis demonstrates that the acquisition of political capital by being elected as a representative of the MPC certainly improved the performance of some local enterprises in terms of scale by bringing political or economic benefit to these enterprises. It can be pointed out that the expansion of their political “credit” makes it possible for them to receive more loans and make more investments, or that the local government can offer favors that could increase their tax revenue as they expand their scale. On the other hand, our study also suggests that the acquisition of political capital by local enterprises does not improve the efficiency of their performance. We also added a cross term of enterprise-specific coefficients and a time trend to the DID analysis to account for the possibility that each company had a different time trend throughout the analysis period, but the results did not change. In summary, the conclusions of our analysis are consistent with Kamo’s (2013) point that delegates to the People’s Congress act as “representatives” of the interests of the constituency. We believe that this analysis is still quite undeveloped and that there are many things that might be improved. For example, it should be clarified how political capital positively and negatively influences a company’s outcomes by analyzing the activity of the delegates, such as bill submissions or proposals in the MPC. It is also necessary to test the effects of central policy, such as the anti-corruption campaign that began in 2013. Our research of the impact of “political capital” on the performance of local enterprises through the local People’s Congress in China has just begun. Due to data limitations, our analysis was conducted only for the case of Yangzhou City, but we plan to conduct additional analyses for other cities with different conditions and compare the results.

References Angrist JD, Pischke J-S (2009) Mostly harmless econometrics: an empiricist’s companion. Princeton University Press, Princeton

66

4 Political Capital and Enterprise Performance in China …

Boubakri N, Guedhami O, Mishra D, Saffar W (2012) Political connections and the cost of equity capital. J Corp Finan 18(3):541–559 Cao X, Pan X, Qian M, Tian GG (2017) Political capital and CEO entrenchment: evidence from CEO turnover in Chinese Non-SOEs. J Corp Finan 42:1–14 Chen CJP, Li Z, Xu S, Sun Z (2011) Rent-seeking incentives, corporate political connections, and the control structure of private firms: Chinese evidence. J Corp Finan 17(2):229–243 Cheng H, Fan H, Hoshi T, Hu D (2019) Do innovation subsidies make Chinese firms more innovative? Evidence from the China employer employee survey. NBER Working Paper No. w25432. National Bureau of Economic Research, Cambridge, MA Cheng L (2018) Estimating the value of political connections in China: evidence from sudden deaths of politically connected independent directors. J Comp Econ 46(2):495–514 Cheng L, Cheng H, Zhuang Z (2019) Political connections, corporate innovation and entrepreneurship: evidence from the China Employer-Employee Survey (CEES). China Econ Rev 54:286–305 Friedberg L (1998) Did unilateral divorce raise divorce rates? Evidence from panel data. Am Econ Rev 88(3):608–627 He L, Wan H, Zhou X (2014) How are political connections valued in China? Evidence from market reaction to CEO succession. Int Rev Financ Anal 36:141–152 Kamo T, Takeuchi H (2013) Representation and local people’s congresses in China: a case study of the Yangzhou municipal people’s congress. J Chin Polit Sci 18(1):41–60 Kamo T (2013) The political roles of China’s democratic institutions: agents, remonstrators and representatives, and collaboration between democratic institutions. Quarterly J Inst Developing Econ Jpn External Trade Organization 54(4):11–46 Li H, Meng L, Wang Q, Zhou L-A (2008) Political connections, financing and firm performance: evidence from Chinese private firms. J Dev Econ 87(2):283–299 Ling L, Zhou X, Liang Q, Song P, Zeng H (2016) Political connections, overinvestments and firm performance: evidence from Chinese listed real estate firms. Financ Res Lett 18:328–333 Liu Q, Luo J, Tian GG (2016) Managerial professional connections versus political connections: evidence from firms’ access to informal financing resources. J Corp Finan 41:179–200 O’Brien KJ (1994) Agents and remonstrators: role accumulation by Chinese people’s congress deputies. China Q 138:359–380 O’Brien KJ (2009) Local people’s congresses and governing China. China J 61:131–141 Wen H, Zhao Z (2020) How does China’s industrial policy affect firms’ R&D investment? Evidence from ‘China manufacturing 2025. Appl Econ 48:1–14 Wu W, Wu C, Zhou C, Wu J (2012) Political connections, tax benefits and firm performance: evidence from China. J Account Public Policy 31(3):277–300 Xu G, Liu Q, Zhou R, Wang R, Yano G (2018) Anti-corruption, safety compliance and coal mine deaths. J Econ Behav Organ 188:458–488

Chapter 5

Land Market Competition Among Local Governments: A Spatial Analysis of Zhejiang Province

5.1 Land and Real Estate System Reform and Price Restraint Policies Since the 1990s, the Chinese economy has featured excess liquidity due to a continuously low fixed exchange rate against the United States (US) dollar and has frequently experienced sharp increases in asset prices commonly known as “bubbles.” While some traders amassed great wealth through real estate transactions, skyrocketing housing costs in major cities have been causing dissatisfaction among the masses. There has been one protest after another over the government expropriation of land in agricultural villages by farmers whose land has been taken by force with only minimal compensation, and these practices have drawn wide-ranging concerns both inside and outside of China.1 When this issue is covered by the media, the one-sided exploitation of residents and farmers who are forced off their lands by regional governments alongside developers is often portrayed as problematic.2 As frequently noted, the background to this issue includes rent-seeking by regional governments via intervention in land markets, as well as a tug of war over this issue between the central and regional governments. There are two points that require attention here. First, while reference is made to “government intervention in land markets” as a single concept, the situation differs greatly depending on the region, the land use, The empirical analysis in this chapter relies on Kajitani and Fujii (2016). I would like to thank Daisuke Fujii for permission to reproduce our paper. 1

Huang (2006) analyzed systemic changes concerning such land issues. Li and Jia (2006) and Wang (2006) examined the “landless farmers” problem and the living conditions of such farmers after losing their land. 2 Ren (2009) investigated the behavior of real estate companies; Zhu (2007) analyzed the relationship with public finances; and Zhang (2007) examined the environmental problems accompanying agricultural land development. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 K. Kajitani and T. Kamo, Political Economy of Reform in China, Kobe University Social Science Research Series, https://doi.org/10.1007/978-981-19-0202-4_5

67

68

5 Land Market Competition Among Local Governments: A Spatial …

and how the land was expropriated and turned over to the government. Second, such rent-seeking by regional governments and the tug of war between regional and central governments are not by any means new phenomena, but have rather been noted ever since the earliest stages of China’s reform and opening up policy. Consequently, when discussing the recent problems with regional government intervention in land markets, it is necessary to examine the various facts regarding such land markets and then clarify the points upon which they differ from the past problems that have emerged between the central and regional governments as well as their significance when considering China’s future economic development. To understand these issues, this article focuses on the following three points: (1) the relationship between land expropriation and the structural problems facing regional government finances, (2) the real estate price increase mechanism and the effectiveness of government policies to restrain prices, and (3) changes in the patterns of economic development led by regional governments observed vis-à-vis this land expropriation problem. First, we review the establishment of the main laws concerning land and real estate transactions during the reform and opening up period as well as the historical changes in government policies to restrict real estate prices.3 It was not that long ago that China, which firmly maintained a public land ownership system under its planned economy, began to trade land and developed related laws. First, the Land Administration Law of the People’s Republic of China (hereafter, the “Land Administration Law”) was enacted in 1986 to arrange a legal system for land management assuming a public ownership system. In 1987, in Shenzhen City, the rights to use state land in the city were transferred on a paid basis for the first time. The urban development method whereby the regional government sold its land use rights to private developers and used the funds to invest in the infrastructure and urban construction was initiated. It is said to be modeled after the urban development method used in British Hong Kong during the colonial era. Additionally, the legal grounds for such a system were provided by the Constitution and the Land Administration Law, both of which were revised in 1998.4 Under these circumstances, the first real estate development boom (i.e., the new land enclosure movement) was sparked by Deng Xiaoping’s 1992 South China Tour Speech. The urban development method using paid transfers of land use rights spread throughout the country, centered on cities in the coastal regions. The percentage of real estate investment in national fixed capital investment suddenly rose from around 6.1% in 1991 to 9.3% in 1992 and to 15.6% in 1993. However, this real estate boom of the early 1990s abated due to a series of “bubble smashing” and tightening monetary policies enacted by Prime Minister Zhu Rongji and depressed investment resulting from the 1997 Asian financial crisis. 3

Liu and Zhang (2006) and Huang (2006) were used as primary references in writing this section. This is a system for the expropriation of land in cities. There are still strict restrictions in place regarding the conversion of agricultural land that is collectively owned into non-agricultural land. The current method in use is that agricultural land is first nationalized through government expropriation and then transferred to developers.

4

5.1 Land and Real Estate System Reform and Price Restraint Policies

69

Yet, even during that period, systematic preparations for land and real estate transactions steadily advanced. The most important advances included reforms for home ownership and the commercialization of urban housing, which had previously been allocated by units centered on state enterprises. First, the 1994 State Council “Decision on the Deepening of Urban Housing System Reform” and the 1998 “Notification regarding the Further Deepening of Urban Housing System Reform and Acceleration of Housing Construction” made clear the policy of advancing the commercialization of urban housing, abolishing the system of directly providing housing, and promoting the construction of new housing. A series of movements toward the commercialization of housing stimulated demand among urban residents and became a major cause of the second real estate boom starting in 2003. The preparation of systems for the government to expropriate agricultural land and other collectively owned land for smooth development was advanced. First, national land investment companies were established in Shenzhen and Shanghai in 1996, and a unified “land banking system” was established for expropriating land for development and managing the preparation of plots and infrastructure under consignment from regional governments via land banking centers, a system that gradually spread nationwide.5 The new Land Administration Law revised in 1998 (enacted in 1999) strengthened the control over land development by higher government authorities (Gao and Liu 2007). To develop collectively owned land in agricultural villages as plots for construction, it first had to be expropriated by the state and “nationalized,” and the rights to conduct examinations and give authorizations were limited to the State Council and provincial governments. The terms for the rights to use state land were set specifically according to the purpose of land use, with 70-year terms for residential uses, 50-year terms for industrial uses, 50-year terms for educational, science and technology, culture, and hygiene uses, and 40-year terms for commercial and entertainment uses. The standards for giving compensation to farmers when expropriating agricultural land were prescribed at payments of 6–10 times the value of the average harvests over the 3 years prior to the expropriation (Article 47), establishing systematic preparations for the paid transfer of land use rights. The “Circular of the State Council on Strengthening the Asset Management of State-Owned Lands,” which was issued in 2001, called for openness, fairness, and justice in trading land use rights based on this land banking system. The consultation method accounted for the overwhelming share of paid transfers of state land until that time; the prices and process were criticized as lacking transparency. However, the movement toward a bidding system and other market-based transfers of land use rights advanced full scale thereafter.6 5

This “land banking system” has three different types: the Shanghai type, which emphasizes the market mechanism, the Hangzhou type, which is conducted under government initiative, and the Nantong type, in which the market mechanism is tied to land and resource management by the government (Huang 2006, vol. 1: 23). At present, more than 1,000 cities have established land banking centers (Jiang et al. 2007). 6 For example, the revenue from bidding-method land transfers nationwide grew by 40% per year from 35 billion yuan in 2000 to 49.2 billion yuan in 2001. In addition, the percentage of the total

70

5 Land Market Competition Among Local Governments: A Spatial …

Strengthening the management of land development through the land banking system was done fundamentally to prevent the unlawful diversion of the assets of state enterprises that go bankrupt or are subject to restructuring and to address concerns regarding the decrease in the amount of arable land through protection from rampant development (“The Decipherment of Land,” Caijing Magazine, no. 153). Nevertheless, this unification of the land development authority through expropriation and authorization by regional governments also precipitated the following results that were not necessarily intended by the central government. First, because the land sales market from the government to the private sector (the primary land market) was unified through authorization by the central and provincial governments, a supply shortage emerged from this monopolistic system, inviting land price increases. Second, the land paid use system, which had previously been applied primarily to state land in urban areas, began to be applied full scale to agricultural and other collectively owned land.7 Third, while the percentage of land stockpiled in regions that was transferred at market prices increased, because the compensation paid when agricultural and other lands were expropriated was kept low, the income of regional governments—the differential between the compensation paid and the transfer price— increased greatly. Under such conditions, the second real estate development boom advanced full pace starting in 2002. Unlike the first boom, the method whereby regional governments expropriated collectively owned agricultural and other lands, nationalized them, and then transferred them for profit had now become mainstream. For that reason, the presence of “landless farmers”—those who had lost their land without receiving sufficient compensation—began drawing attention as a social problem. Amid such conditions, the Rural Land Contracting Law, which was implemented in March 2003, stipulated that the land “contracting rights” of individual farmers represented a type of usufruct right over the land and opened a path to protecting farmers’ property rights. Specifically, it recognizes the inheritance, sale, and purchase of land use rights based on the farmer’s volition and stipulates that the government or other parties cannot recall the usage rights to agricultural land within the contracting period (i.e., 30 years for arable land, 30–50 years for pasture, and 30–70 years for forests). Then, in October 2007, the Property Law came into force, which was based on socialist public ownership, and confirmed the principle of joint economic development through diverse ownership systems and stipulated the equal protection of ownership rights of not only the state but also of individuals. It is noteworthy that this law confirmed that land use rights and other usufructs are secured through a registration system, and it included provisions regarding what happens after contract periods expire and set forth compensation guidelines for residents from whom the government expropriates land. land area transferred using the bidding and auction method increased from 15% in 2002 to 33% in 2003 (Huang 2006, vol. 1: 30). 7 For example, the land banking center established in one eastern county expropriated agricultural land accounted for 14.4% of the total land stockpiled through 2001 but 88.3% in 2002 (“The Decipherment of Land,” Caijing Magazine, no. 153).

5.1 Land and Real Estate System Reform and Price Restraint Policies

71

At the Third Plenary Session of the 17th Chinese Communist Party (CCP) Central Committee held in November 2008, the CPC adopted the “Decision on Some Important Questions in Promoting the Development of Rural Reform,” which recognized the purchase and sale of agricultural land use rights under certain conditions including free will and compensation, joint ownership of agricultural land, and no change in the land use. In response to this decision, Rural Land Transfer Trading Centers were established in Sichuan, Zhejiang, and other provinces as a systematic framework for the market trading of agricultural land. In addition to the development of industrial zones to attract foreign enterprises, the demand for the development of commercial housing increased during the latter 1990s, so laws were implemented and active real estate financing was arranged to support this. Due to this series of systematic preparations and the excess influx of domestic and foreign capital against the background of insufficient projects for domestic investment, the price of land and real estate consistently rose faster than the price of goods, especially in major cities. In addition to this sharp increase in real estate prices, a variety of social problems emerged such as forced evictions associated with land development, pressure on the lifestyles of city residents from rent increases, and dissatisfaction with real estate developers and regional government officials making excessive profits. The central government has responded by implementing laws and issuing individual notifications to prevent regional governments and developers from expropriating land illegally by violence and to ensure that sufficient compensation is provided to farmers and other original land usufructuaries. On the demand side, when the real estate market showed signs of heating up, the central government adopted restrictive measures centered on the direct regulation of real estate financing by financial organs.8 Harsh tightening measures were implemented against realtors and markets including total volume limits on development items that are not consistent with national industrial policy objectives, especially during the heated real estate investment in and around Shanghai between 2003 and 2004 (see, e.g., “Missions Dilemmatic,” Caijing Magazine, no. 179). For example, in June 2004, the government ordered the cancellation of a project in which the land was obtained by illegal means for development by Jiangsu Tieben Iron and Steel Company Ltd. in Changzhou City, Jiangsu Province, and those involved were punished (Tanaka 2007). The October 2004 “State Council Decision on Deepening Reform and Tightening Land Management” stipulated volume regulations on land for construction, strict management of the conversion of agricultural land, and sufficient compensation for living expenses to farmers on the occasion of land development.9 8

For example, the “Notification regarding Further Strengthening of the Management of Real Estate Financing,” which was issued in July 2003, required project self-financing of at least 30% (35% starting in 2006) when enterprises develop real estate with bank financing. However, the 30% self-financing rule had various loopholes; for example, about 70% of “self-financing” was actually covered by outside capital. See “Secrets of Housing Loan,” Caijing Magazine, no. 87, and other materials. 9 Regarding the latest trends in the securitization of land in agricultural villages, see “The Restoration of Land Ownership,” Caijing Magazine, no. 222.

72

5 Land Market Competition Among Local Governments: A Spatial …

During the real estate price jumps that occurred in 2007 and 2009, the government repeatedly intervened directly in the real estate market by increasing down payments for condominiums purchases and regulating real estate loans. However, while this series of policies did have some temporary restrictive influence, it was not sufficiently effective in preventing sharp price increases in the real estate market. The special relationship between China’s land market and its regional governments may explain why these policies were ineffective, as seen in the following sections.

5.2 Land and Real Estate Markets and Regional Governments’ Fiscal Revenue During the reform and opening up period, regional governments that suffered a chronic shortage of sources for fiscal revenue repeatedly intervened in factor markets and used the resulting “rent” as a non-regular fiscal source. We now discuss in detail how regional governments secured fiscal sources by intervening in the land market specifically. First, regarding the tax revenues related to land in regional government budgets, we will review the items that are subject to taxation, the basis for their inclusion, tax rates, and the percentages retained by regional governments. As shown in Table 5.1, the first distinctive characteristic of the real estate taxation system in China is the existence of different systems for taxing land and buildings. For example, the provincial real estate tax and the city real estate tax on foreign-affiliated enterprises are charged only on buildings and not on land. In contrast, the city land use tax is systematized as a use expense paid to the government by enterprises that are not publicly owned, so this is significant for adjusting differential rents in accordance with the land’s profitability.10 Another institutional characteristic that should be noted is that while the state holds the ownership rights over land in urban areas, the land in agricultural villages is fundamentally collectively owned. To reflect this fact, the tax structure adopts different systems for agricultural villages and cities in taxing both land and real estate, and there are charges, such as the city maintenance construction tax, that are additional taxes charged on the amount of taxes paid, making the system extremely complex. The following problems have been noted with the current land taxation system (Qin and Li 2007). • The tax base is too small, and the properties subject to taxation are not clear. There are many provisions for tax exemption; for example, real estate taxes are 10

Land use expenses were introduced for the government to collect land rents from state enterprises that had been using state land free of charge up until that time. The collection amounts were initially set at symbolic levels sufficient to demonstrate that the land ownership rights belonged to the state (Onodera 1997). The tax rate of the city land use tax was also initially set at a minimal level but was increased to three times its prior level in 2007.

5.2 Land and Real Estate Markets and Regional Governments’ …

73

Table 5.1 Real estate-related tax revenue Taxes on real estate occupiers

Tax on real estate sales and purchase and transfer profits

Tax on real estate rental income

Type of tax

Tax basis

Tax rate

Real estate tax

Real estate cost

1.20%

City real estate tax

Real estate cost

1.20%

Arable land occupancy tax

Arable land area

5–50 yuan/m2

City land use tax

Land area

0.6–30 yuan/m2

Contract tax

Contract amount

3–5%

Stamp tax

Contract amount

0.03–0.1%

City maintenance construction tax

Amount of taxes paid

1–7%

Enterprise income tax

Enterprise income

3–30%

Individual income tax

Personal income

3–45%

Land value-added tax

Value-added

30–60%

Real estate tax

Rental income

12%

City real estate tax

Rental income

18%

Enterprise income tax

Enterprise income

3–30%

Individual income tax

Personal income

3–45%

Source Prepared by the author from “The Start of Property Tax,” Caijing Magazine, no. 99; Liu (2020); and other materials

not charged if the property is not used for business purposes, and public enterprises are not charged land use taxes. • The rates of land value-added taxes and certain other taxes are too high, providing a strong incentive for false reporting and tax evasion. • There are large distortions in the tax rates; for example, the tax burden on land transactions is heavy while the tax burden on land holdings is light. For that reason, there is a lot of land being held for speculative purposes and not being used. • Rents, taxes, and expense burdens are commingled and inconsistent.11 The scale of tax revenues from these land transactions and land holdings, excluding those that contain revenues from other industries such as business and income taxes, has certainly been expanding in recent years, but the ratio of these tax revenues in total regional government fiscal revenues has not significantly changed. 11

To resolve such problems with tax revenues from land transactions and land holdings, the introduction of a more comprehensive real estate tax (property tax) has been examined (“The Start of Property Tax,” Caijing Magazine, no. 99; “Truth of Property Tax,” Caijing Magazine, no. 179). In January 2011, the cities of Shanghai and Chongqing both announced the introduction of a trial real estate tax on housing holdings.

74

5 Land Market Competition Among Local Governments: A Spatial …

Nevertheless, regional governments collect the following various expenses from developers and other parties in addition to the aforementioned tax revenues related to land transactions and holdings (“The Decipherment of Land,” Caijing Magazine, no. 153). 1.

2. 3.

Arable land reclamation expenses, land use rights transfer fees, land use fees for new construction and additions, management expenses, registration expenses, eviction expenses, and other revenues managed by the land department. Land use expenses, land rental expenses, and other revenues managed by the fiscal department of the regional government. Revenues from diverse expenses paid to the agriculture, real estate, water use, transportation, post and telecommunications, culture, air defense, forestry, and other departments.

It is worth noting that the scale of land use rights transfer fees and other expense revenues managed by the Department of Land and Resources and other departments (hereafter, “land use rights transfer revenues”) has been growing year by year, and that these revenues are also gained by regional governments that sell their own use rights on the market using their de facto ownership rights over agricultural land and can be interpreted as a type of rent gained by government intervention in the real estate market.12 This paid land use system was introduced in Shenzhen City at the end of the 1980s. At the national level, its implementation was stipulated by a series of legal preparations including the July 1988 State Council “Notification regarding the Authority to Ratify the Transfer of State Land Use Rights,” the revision of the Constitution that year that recognized the transfer and lease of land use rights, and revisions to the Land Administration Law. This system was advanced full scale centered on cities in the coastal regions with the active solicitation of foreign capital following Deng Xiaoping’s 1992 South China Tour Speech with the purpose of promoting the development and management of land using foreign capital.13 In addition, the “Notification on Some Issues regarding the Development of the Real Estate Industry” stipulated that collectively owned land in agricultural villages and elsewhere could be expropriated by the state, converted to state land, and then transferred (Onodera 1997, p. 31). Such paid transfers of collectively owned land advanced full scale in the latter half of the 1990s under the land banking system previously described. Initially, land use rights transfer revenues were split by the central and regional governments with a 4:6 division. Thereafter, the amount paid to the central government was reduced to 32%, and with the 1992 “Provisional Regulation regarding the 12

The following land use rights transfer revenues can be broken down into three components: land acquisition expenses (including living expense compensation to farmers), land and infrastructure preparation expenses, and sales and purchase revenues (transfer revenues). 13 In fact, the transfer of these “revenues from expenses” can be summarized as follows. First, the government completes the relocation of, and compensation for, the enterprises and residents that have been using the land, prepares the land, and develops the basic infrastructure (i.e., roads, water supply, sewage, telecommunications, gas, heating steam, and grading), and then transfers the land to real estate developers. See Onodera (1997).

5.2 Land and Real Estate Markets and Regional Governments’ …

75

Collection and Management of Revenues from the Paid Transfer of State Land Use Rights,” it became sufficient for the regional government to allocate only 5% of the transfer revenues to the central government. Since the tax sharing system was implemented in 1994, all of the revenues have remained with the regions and have become an important source of funding for regional governments. During the real estate boom that started in 2003, it was not unusual for the transfer revenues received by regional governments in coastal provinces to reach billions of yuan, and much of it was spent on disorderly urban construction. Under such conditions, in 2004, the State Council demanded that 15% of regional government transfer revenues be allocated to agricultural development (“The Redistribution of ‘Land Leasing Fees,’” Caijing Magazine, no. 155). In addition, the “Notification regarding the Standardization of the Management of Revenues and Expenses from the Transfer of State Land Use Rights,” which was promulgated by the Office of the State Council in December 2006, stipulated the range of land use rights transfer revenues and demanded stronger management of collections. Some of these transfer revenues are posted as “fund revenues” in regional budgets and are ratified by the central government.14 The scale of such revenues was 203.751 billion yuan in 2006, accounting for less than one-third of total revenues (807.764 billion yuan), and it varied by region. The remainder is believed to be retained by regional governments as “funds outside of the system” that are not shown in statistics or as “a second budget” or “hidden reserves.15 ” There are also said to be many cases in which regional governments expropriate land illegally during this process, including cases in which regulatory compensation is not paid to residents or residents are evicted by violent means. Figure 5.1 shows that the revenue from the sale of land usage rights dramatically increased starting in 2003, that is, from the time when the second land development boom began to heat up.16 There is a large difference in growth between the eastern and central and western areas, showing that these revenues reflect regional deviations in land development. In general, such land use rights transfer revenues are allocated about 40% to governments above the township and village level and 40% to village committees, 14

“Fund revenues” are a temporary category born amid the tax reforms that occurred since the 1990s in which funds outside of the budget were incorporated into the budget. These are managed by each department including the industry, transportation, commerce, education, culture, and agriculture departments, but their expenditures must be ratified by the central government. Their partial incorporation into general budget revenues is also being considered. See Ma et al. (2003). 15 The decision was made to make all revenues and expenditures from paid land transfers subject to regional fund budget management in January 2007. At the same time, clear stipulations were made limiting the range of use for land revenues to land expropriation, eviction compensation, land development, farmer assistance, and city construction expenses (“Real Property Tax,” Caijing Magazine, no. 179). 16 For example, in 2007, the total amount of transfer revenues reached about 52% of total regional public finance revenues (estimated by the author from China Land and Resources Statistical Yearbook, China Statistical Yearbook, and other materials). The ratio was believed to be even higher when examining only cities in the coastal regions.

76

5 Land Market Competition Among Local Governments: A Spatial …

Fig. 5.1 Total revenue for sale of land use rights in China. Source Almanac of Chinese Land Resource, 2000–2018. Note Amounts show the totals of land use rights transfer fees, land development expenses, land preparation expenses, and other costs paid by land users to the state

with just 20% or less being given to the farmers themselves.17 The changes in transfer income are shown in Fig. 5.1.

5.3 Land Market Structure and Rent-Seeking In economics, when considering real estate and land prices, the land rent is usually first determined by the rental market and, with that rent as a given, the land price is then determined through arbitration with other investment assets on the asset market. However, as shown in the following example, the conditions surrounding China’s real estate market greatly differ from those assumed by standard economic theory. The most distinctive characteristic of China’s real estate market is that land is publicly owned and only land use rights can be traded, while the ownership of housing and other real estate constructed on land by individuals and corporations is permitted. Because of this, China’s real estate market has a complex, multi-level structure with various characteristics. First, there is a market on which regional governments expropriate agricultural land or old city land and transfer the use rights on a paid basis to developers and other businesses. This is the primary real estate market. Next, there is a secondary market on which developers and other parties develop the land that they acquire from regional governments, construct condominiums and other real estate, and sell 17

See Guo (2005). According to this paper, the average compensation received by farmers was equivalent to approximately 3–5 years of their average annual income.

5.3 Land Market Structure and Rent-Seeking

77

land use rights and real estate ownership rights together as a package to individuals and companies. Finally, the market and rental contracts for such real estate bought and sold on the secondary market can be understood as being traded on a tertiary market.18 The arbitration with other assets assumed by standard economic theory is believed to function on the secondary and tertiary markets. However, the total amount of stock of land on the secondary market and its expected returns are believed to be greatly influenced by the supply of land on the primary market. Accordingly, let us now carefully examine the conditions regarding land sales on the primary market. The primary land market is none other than the market on which the government sells the rights to use land owned by the state in cities and agricultural villages to the private sector. These sales of land use rights can be broadly divided into three methods—gratis transfers, consultations, and bidding or auctions. Until 2000, gratis transfers accounted for the sale of more land area than paid transfers; however, since 2001, the latter has greatly exceeded the former. State land transferred gratis is believed to mostly be in demand by the state itself for the provision of public goods such as roads, parks, green belts, and cultural and educational facilities.19 The following considerations focus exclusively on paid transfers. Among paid transfers, the consultation method is the sale of land to certain developers and enterprises at a low price for the construction of factories and public facilities. Because attracting factories to a locality increases the prospects for longterm tax revenues, regional governments compete in establishing economic development zones and lowering sales prices to enterprises, which are reportedly sometimes almost lowered to the level of the land acquisition costs. However, the percentage of consultation transfers in total paid transfers is on a declining trend due to government policies to promote competitive bidding. The bidding system has suddenly increased as the development of agricultural land advanced full scale from the late 1990s through the aforementioned land banking system, and this format now accounts for most commercial and residential land transfers.20 The consultation method accounted for 60%–70% of the total land area transferred through 2006, but only for about one-half of that percentage on a monetary basis. This also demonstrates the fact that land transferred under the consultation method is sold for prices far lower than land sold under the auction and bidding method. As previously explained, that is the result of regional governments competing for lower supply prices on the primary market to attract enterprises, and considering only this point, one might think that the land market is in a desirable condition 18

This refers to Gao (2007) and other materials. Problems have been noted including regional governments using such land that should properly be used for public uses entirely for the construction of luxury office buildings and other structures. (Jiang et al. 2007). 20 However, it is important to note that among competitive sales methods, in recent years the method called Guapai, which has a stronger collusive nature compared with pure competitive bidding, has become the mainstream (accounting for more than 70% of both area and amount in 2008). See Cai et al. (2009) regarding the problems with how sales under the Guapai method easily give rise to collusion between government officials and developers. 19

78

5 Land Market Competition Among Local Governments: A Spatial …

with market competition working. On the other hand, however, this suggests that the market is under monopoly conditions for land acquisition by the government and that sufficient compensation is not being paid to farmers and others as land sales prices are artificially depressed. This phenomenon whereby clearly different transfer formats with large price differentials are used depending on the purpose of land use is an indication of regional governments’ monopoly over the primary market’s land supply. This phenomenon can be understood as a typical price differentiation strategy by a monopoly enterprise. A price differentiation strategy arises when a monopoly enterprise with pricesetting power on the market faces two types of buyers with greatly different price elasticities of demand and sets lower product prices for those buyers with a higher price elasticity. Well-known examples of typical price differentiation strategies include student discounts and charging higher taxi fares late at night. How does this apply to China’s primary land market? For example, when a manufacturing company is seeking land to construct a factory, except for cases in which a special industrial agglomeration is being formed, the company has no need to insist on any specific location and is strongly attracted to locations with the lowest land prices, personnel expenses, and other costs. As illustrated by the succession of manufacturing company production bases that are relocating from industrialized to developing countries, the production locations of manufacturing firms are not necessarily restricted by the geographic locations in which their products are in demand. Therefore, the demand for land to construct factories is believed to have a very high elasticity to land prices. In contrast, because housing and commercial facilities are originally constructed anticipating the demand of local residents (their location is highly restricted by demand), their substitutability with other locations is considered low. For that reason, if a given area has a large population and a certain level of profitability prospects, there is an incentive to obtain land in that area, even if the costs are somewhat high. In other words, compared to land for the construction of factories, the price elasticity of the demand for residential and other land is rather low. So, if the regional government supplies land on a monopoly basis, it can adopt a price differentiation strategy for the demand for the two types of land. That is, the regional government can supply land for the construction of factories at an exceptionally low price (P2) to attract as many factories as possible and secure future tax revenues (the case depicted on the right-hand side of Fig. 5.2), while enjoying a high price (P1) and high monopoly rents for residential land that has a low price elasticity (the case depicted on the left-hand side of Fig. 5.3). While there may be some difference in the extent to which this phenomenon is caused by price elasticity, in both cases, the supply volume is less than that under full competition because it is set at the intersection of the marginal revenue (MR) and marginal cost (MC) curves, so land rents and prices rise. After deducting the compensation paid to residents and other land acquisition costs from the land transfer

5.3 Land Market Structure and Rent-Seeking

79

The market for the Commercial or Residenal use of lands The market for the industrial use of lands

Fig. 5.2 Price differentiation in the land market. Source The author

Fig. 5.3 Trends in the price of land use rights. Data: CEIC Data (http://www.ceicdata.com). Note Each line stands for the average value of land use rights in national major 105 cities

80

5 Land Market Competition Among Local Governments: A Spatial …

revenues (the shaded area of the figure), the regional government’s revenue is equal to a monopoly rent on the land market.21 The fact that this type of price differentiation is being implemented can be confirmed by comparing the difference in the price elasticity of demand for land uses. Figure 5.3 shows the trend of the average unit price of land use rights in 105 major Chinese cities. This shows that for all cities listed, the unit price of residential and commercial land is much higher than that of industrial land. This wide gap between the price of commercial and residential land and industrial land might be caused by a kind of price discrimination by local governments that calculate the difference in demand elasticity for different land uses.

5.4 Empirical Studies on the Fiscal Competition by the Local Governments In following section, we mainly focused on the competition between local governments in China to attract industrial companies.22 The earliest theoretical studies such as Oates (1972), or Zodrow and Mieszkowski (1986) and Wilson (1986), analyzed fiscal competition by the local governments, and focused on interaction due to taxbase mobility, which is known as “tax competition” or “yardstick competition.”23 They pointed out that each region finance provision of a public good with a tax on the locally employed capital.24 When capital can moves among regions in response to tax rate differentials, the tax rate for capital in each region tends to be lower than the optimal level. Inspired by pioneering studies, various types of fiscal competition have been analyzed. Hoyt (1993) examined tax rate competition in case of mobile residents 21

As stated above, it has been noted that land sales prices for enterprises advancing into economic development zones and other lands for factory construction have been decreased nearly to the level of the land acquisition costs. Under such conditions, it might be thought that there is no leeway for acquisition of monopoly rents by regional governments which are land suppliers. However, as noted in the previous section, when enterprises locate factories they pay various expenses to different government departments aside from the land sales price, and the total amount of these expenses is of a scale that cannot be overlooked. These various expenses are not systematized like taxes, and so they are considered to be a type of rent generated because the regional government is the monopoly land supplier. 22 See Kajitani and Fujii (2016) for more information on the theoretical model upon which the empirical analysis in this section is based. 23 Tax competition between local governments consists of tax rate reductions implemented to attract a tax base, such as companies and laborers, into a region. In contrast, yardstick competition occurs primarily due to the propagation of information between regions. It is typical that the residents of each region affect their residential tax rate by voting in regional local elections, by comparing the tax rate of their area to that of other areas. 24 Zodrow and Mieszkowski (1986) shows that capital tax rate competition to attract firms among homogenous local governments causes an undersupply of public goods as a result of a Nash equilibrium.

5.4 Empirical Studies on the Fiscal Competition …

81

and showed that competition results in higher levels of government service when the demand for housing is elastic and lower levels when the demand is inelastic. Wildasin (1988) compared tax rate competition and public expenditure competition as a policy variable and demonstrated that the supply of public goods at equilibrium in both cases was at an inefficiently low level. Furthermore, it also showed that the equilibrium in tax competition is greater than in public expenditure competition. On the other hand, Noiset (1995) and Matsumoto (1998) pointed out the possibility of the overprovision of public goods, because although a higher tax rate drives out capital, the tax base may ultimately expand if the increased revenue leads to an increase in public inputs that attract capital by increasing productivity. Therefore, if we consider leased land as a public good that improves the utility of residents, it may be underprovided, while if we consider land as a public input that improves the productivity of firms, it may be overprovided. Brueckner (2003) overviewed previous studies on such competition between local governments. According to him, the theoretical model for such studies can be classified into two types of basic models, the “spillover model” and the “resource flow model.” In the spillover model, local governments are assumed to directly affect the behavior of residents in other regions. Yardstick competition is a kind of spillover model, because the spillover of information relating to policy decisions in one region directly impacts the voting behavior of residents of other regions. On the other hand, in the resource flow model, policy decisions by local governments are assumed to indirectly affect residents in other regions through the interregional movement of policy elements (i.e., capital or labor). Tax competition in which the tax base moves toward the area with the lower tax rate can be considered an example of the resource flow model.25 From an empirical viewpoint, many existing studies have adapted a spatial lag framework in spatial econometrics to consider the strength of interaction based on distance or contiguity between governments. One of the earliest studies using this method is Case et al. (1993), which estimated the interaction function of public expenditure at the state level in the US. Inspired by their study, many empirical studies have examined expenditure interactions among local governments (Revelli 2003; Solé-Ollé 2006). While Case et al. (1993) examined expenditure competition, studies focusing on tax competition to attract firms can be found. For example, Brueckner and Saavedra (2001) estimated the response function of property tax rates among local governments in the Boston metropolitan area, and its results indicate the presence of strategic interaction. Buetter (2001) also analyzed tax rate competition using data on the local 25

Such recent studies on competition between local governments have also incorporated the results of spatial economics. Ottavianoa and Yperseleb (2005) theoretically investigated tax competition using the concept of the so-called “integration rent,” which is based on spatial economics. According to them, when there is tax competition to attract companies, the externality of integration provides an integration rent to companies located in regions in which enterprise integration already exists. Therefore, the regions with integration have the ability to attract companies, compared to areas in which integration does not exist. This means that the former region might impose a higher asset tax rate than the latter.

82

5 Land Market Competition Among Local Governments: A Spatial …

business tax rate in Germany, and concluded that governments are indeed in tax rate competition. Chen and Hong (2009) also employed the spatial econometric approach to estimate tax competition among regions in Taiwan. With respect to fiscal competition in China, existing empirical studies have mainly focused on expenditure competition. This is because, although Chinese local governments have not been able to independently choose tax rates in the Tax Sharing System (Fenshuizhi) since 1994, they have greater authority in public expenditures. For example, Guo and Jia (2009) examined the strategic interaction of public expenditures at the provincial level during the 1986–2006 period and demonstrated the existence of expenditure competition. Zhang and Chen (2007) also examined expenditure competition among provinces and came to the same conclusion. On the other hand, studies that have examined tax rate or revenue competition are rare. As far as we know, Yao and Zhang (2008) is the only example of a statistical analysis of revenue competition by local governments. Their study took a spatial statistics approach and calculated Anselin’s local spatial autocorrelation index, the so-called LISA (Anselin, 1995). Fujii (2016), with the reference to Case et al. (1993), pointed out that there is an expansion of government spending competition between local governments. Fujii also noted that there is a mechanism for evaluating local bureaucracy in the background of such competition. In short, to evaluate and promote the Chinese bureaucracy system, the regional growth rate is a very important criterion. However, these studies did not focus on tax competition among local governments in China because in China, the property tax system is under construction and local governments cannot set taxes independently. This does not mean, however, that such competition between governments to attract companies does not exist in China. We could not find any studies that have empirically analyzed land lease competition by local governments, to the best of our knowledge. In the following section of this paper, we will therefore consider the possibility of competition among local governments to attract companies through subsidies for investment and land purchases.

5.5 Data and Empirical Method 5.5.1 Data In this section, we postulate the hypothesis that local governments compete with each other to attract firms by dumping land use rights for industrial purposes. This dumping by local governments can be considered as a subsidy for firms. To verify this hypothesis, we use county-level socioeconomic data from Zhejiang Province, including land auction prices for industrial purposes in 2010. We choose Zhejiang Province as the sample for this research for the following three reasons. First, Zhejiang Province is a highly developed region and its scale of public finance is quite large. Therefore, local governments can afford to join intergovernmental fiscal competition. Second, trading land use rights has been done

5.5 Data and Empirical Method

83

since the early stages of the opening up process, and most farmland has already been liquidated. Third, there are many distinctive industrial agglomerations, especially in the coastal area in the province. Our dataset consists of the average land use price for an industrial purpose, population, population density, gross regional product (GRP), public revenue, labor population ratio, and the existence of national industrial parks in 89 counties in Zhejiang Province. In this dataset, the average land use price for an industrial purpose is calculated from the publication of land use rights auctions by the Zhejiang provincial department of land resources, and the average price of each auction in a county for 2010. The GRP and public revenue are per capita data. The labor population ratio is the share of the population aged between 15 and 59. The existence of national industrial parks is treated as a dummy variable. These data are obtained from Zhejiang Statistical Yearbook, Zhejiang Yearbook, China 2010 County Population Census Data with GIS Maps, and the website by the Zhejiang Provincial Department of Land Resources (http://www.zjdlr.gov.cn). The descriptive statistics of our dataset are shown in Table 5.2. In addition to socioeconomic data, we use spatial information to consider the geographical relationships of county-level governments. This spatial information is quoted from “China Geo Explorer,” which is provided by the China Data Center, University of Michigan (http://chinageoexplorer.org/cge/). We combine this spatial information of population census data and socioeconomic data in 2010 and compose a so-called GIS dataset. Using our GIS dataset, we draw a map of the spatial distribution of the average price of land used for industrial purposes in Fig. 5.4. This map shows that the land use price is quite higher in coastal areas including the central business districts (CDBs) of big cities such as Wenzhou, Hangzhou, Ninbo, and Xiaoxin. On the other hand, the price is lower in inland areas such as Lishui and Quzhou. Table 5.2 Descriptive statistics of valuables used for empirical analysis Variable (unit, year)

Obs Mean

The average price for the industrial 89 purpose (Yuan/km2 , 2010)

Stand. dev

353.0

207.5

Min

Max

1239.2 102.1

GRP per capita (Yuan, 2010)

89

45,883.9

22,857.4

16,655.7 152,968.4

Public revenue per capita (10,000 Yuan, 2010)

89

219,979.3

209,412.0

0.0 981,659.0

Labor population ratio (%, 2010)

89

71.4

5.1

Population (persons, 2010)

89

516,074.7

285,392.0

Density of population (persons/km2 , 2010)

89

1271.4

2754.5

National Industrial Park (unit, 2010)

89

0.157

0.364

60.0 84.4 88,362.0 1,233,348.0 78.5 18,308.7 0.0 1

84

5 Land Market Competition Among Local Governments: A Spatial …

Fig. 5.4 Spatial distribution of the average price of the land use rights for industrial purposes, by county. Source The author

It seems that spatial autocorrelation of the price of land use rights exists between counties. To examine spatial autocorrelation, we calculate the Moran’s I index. If this index has a positive value, the spatial distribution has a positive spatial correlation (Moran 1948). The Moran’s I index score of the price is 0.608 and is significant at the 1% level. Therefore, the price can be considered to have a positive spatial autocorrelation.

5.5.2 Empirical Method To examine the dumping interaction by local governments in our hypothesis, we follow the models of Case et al. (1993) and Brueckner (2003) and adopt the spatial econometrics method. They applied the spatial lag model (SAR) with a spatially lagged dependent variable on the right-hand side of the estimation equation to examine the interaction by governments. In other words, the model supposes that the price is influenced simultaneously by both the regional characteristics and the prices in surrounding regions. Whereas Case et al. used taxation and Brueckner used public expenditures as their dependent variables, we use the price of land use rights as the dependent variable to examine our hypothesis. Furthermore, we also consider the possibility of interaction effects among the error terms. The model that considers

5.5 Data and Empirical Method

85

the interaction error terms is called the spatial error model (SEM). To consider both endogenous interaction effects and interaction among the error terms, we combine the SAR and SEM models. This combined model is called the SAC model (Elhorst 2014). Thus, our estimation model in matrix notation is expressed in Eqs. (1a) and (1b) as follows: Y = α I + δW Y + Xβ + u

(1a)

u = λW u + ε

(1b)

In this model, α, β, δ, and λ are unknown parameters to estimate. If δ = 0 in Eq. (1a), this SAC model is equivalent to the SEM model, and if λ = 0 in Eq. (1b), this model is equivalent to the SAR model. In addition, if both δ = 0 and λ = 0, then this model is equivalent to the ordinary least squares (OLS) model. For a robustness check, we also estimated the SAR, SEM, and OLS models. Y denotes the average price of land use rights in each county in Zhejiang Province. X denotes the matrix of explanatory variables, and we use population, population density, GRP, public revenue, labor population ratio, and a dummy variable for national industrial parks. The dependent and explanatory variables (except for the national park dummy and the labor population ratio) are in logarithmic form. W denotes a weight matrix that considers the geographical location of county-level governments. We apply two types of weight matrices here. One is the k-nearest method in which the element of matrix wi j is a binary value. If j county is one of the k-nearest counties to i county, wi j takes the value of one and zero otherwise. We set the value of k to 4. Another is the inverse distance method in which the distance between i county and j county is di j , and the element of the matrix becomes 1/di j . These spatial weight matrices are normalized in each row. In this estimation, the result in which we are most interested is the sign of parameter δ and its significance. If δ has a significantly positive value, this means that the price of land use rights in influenced by the price in nearby countries, and it implies that bargaining competition exists among the local governments. If δ is significantly negative, this means that if a government lowers the land price, then the surrounding governments raise their price. If the value δ is not significant, it means that there is no interaction between the counties. In the next section, we show the results of the estimation.

5.6 Results of the Empirical Analysis Table 5.3a shows the estimation results of the SAC, SAR, SEM, and OLS models in the upper part and the spatial dependence tests in the lower part using a four-nearest

86

5 Land Market Competition Among Local Governments: A Spatial …

Table 5.3 Estimation result (a) K = 4 OLS

SAR

SEM

SAC

−0.047

(0.975)

−0.258

(0.846)

0.753

(0.639)

−0.205

(0.816)

GRP per capita −0.185

(0.193)

−0.207

(0.085)

−0.052

(0.699)

−0.246

(0.004)

Public revenue per capita

0.524

(0.063)

0.452

(0.104)

0.647

(0.023)

0.306

(0.093)

Labor population ratio

0.054

(0.000)

0.031

(0.002)

0.027

(0.001)

0.025

(0.014)

Population

0.238

(0.014)

0.177

(0.003)

0.214

(0.001)

0.121

(0.007)

Density of population

0.106

(0.016)

0.067

(0.141)

0.086

(0.069)

0.032

(0.253)

National industrial park

−0.244

(0.024)

−0.160

(0.110)

−0.054

(0.607)

−0.146

(0.060)

0.539

(0.000)

0.842

(0.000)

−0.816

(0.000)

Constant

W*land price W*u

0.610

(0.000)

Obs

89

89

89

89

R2

0.528

0.551

0.5271

0.6060

Degrees of freedom

82

81

82

81

AIC

86.384

64.111

64.223

Test for spatial dependence

DF

Value

LM (lag)

1

34.451

(0.000)

Robust LM (lag)

1

15.750

(0.000)

LM (error)

1

18.978

(0.000)

Robust LM (error)

1

0.277

(0.599)

LM (lag and error)

2

34.728

(0.000)

(b) Inverse distance OLS

SAR

SEM

SAC

−0.047

(0.975)

0.046

(0.973)

0.965

(0.553)

0.021

(0.981)

GRP per capita −0.185

(0.193)

−0.165

(0.184)

−0.036

(0.791)

−0.192

(0.019)

Public revenue per capita

0.524

(0.063)

0.433

(0.130)

0.599

(0.040)

0.323

(0.062)

Labor population ratio

0.054

(0.000)

0.033

(0.001)

0.030

(0.007)

0.030

(0.001)

Constant

(continued)

5.6 Results of the Empirical Analysis

87

Table 5.3 (continued) (b) Inverse distance OLS

SAR

SEM

SAC

Population

0.238

(0.014)

0.137

(0.030)

0.168

(0.015)

0.091

(0.086)

Density. of population

0.106

(0.016)

0.072

(0.129)

0.105

(0.033)

0.040

(0.169)

National industrial park

−0.244

(0.024)

−0.164

(0.114)

−0.081

(0.441)

−0.150

(0.052)

0.476

(0.000)

0.711

(0.000)

0.562

(0.000)

−0.636

(0.000)

W*land price W*u Obs.

89

89

89

89

R2

0.528

0.6060

0.527

0.626

Degree of freedom

82

81

82

81

AIC

86.384

83.971

81.393

Test for spatial dependence

DF

Value

LM (lag)

1

27.753

(0.000)

Robust LM (lag)

1

13.509

(0.000)

LM (error)

1

15.216

(0.000)

Robust LM (error)

1

0.972

(0.324)

LM (lag and error)

2

28.726

(0.000)

Note Probability values are given into parentheses

spatial weight matrix. Table 5.3b shows the estimation results and spatial dependence tests using an inverse distance spatial weight matrix. Anselin and Rey (2014) demonstrated the procedure of model choice using a Lagrange multiplier and robust Lagrange multiplier tests. Following their procedure, the results indicate that the SAC models are appropriate in both matrices. Even so, the estimation results are also stable in the other models, and the coefficients of determination in all of the models are at a sufficient level. Recall that δ is the parameter in which we are most interested. The estimation results of δ are all positive in the SAC models and significant at the 1% level, and in the SAR models as well. The results in the SAC models indicate that a local government lowers (or raises) the price of land use price rights by 0.841 or 0.711 yuan, if nearby governments reduce (or raise) their price by 1 yuan. This response of the local government is consistent with our hypothesis. The estimation results of parameter β are as follows. In the model using the fournearest spatial weight matrix, the GRP per capita is significantly negative at the 1% level, population is significantly positive at the 1% level, and the labor population

88

5 Land Market Competition Among Local Governments: A Spatial …

ratio is significantly positive at the 5% level in the SAC model. The results in the other models are similar. In the model using the inverse distance spatial weight matrix, the GRP per capita is significantly negative at the 5% level and the labor population ratio is significantly positive at the 1% level in SAC model. Based on the results of the SAC and other models, the land price for an industrial purpose is higher in the areas with a high labor population ratio and that are highly populated. This result is quite reasonable. On the other hand, the results in some models show that the land price becomes cheaper in areas with a high per capita GRP.

5.7 Conclusion The results of our empirical analysis show that the prices of land for industrial use are strongly influenced by the land prices in neighboring regions, in addition to the attribute values that determine the land prices in the region. These findings are consistent with the hypothesis that there is competition between local governments to attract companies by auctioning land use rights. This suggests that, in China, where the property tax system is still underdeveloped, discounting land use prices is a powerful tool for local governments to attract industrial companies. However, competition by dumping land use rights prices might disturb the price mechanisms of the real estate market and induce a shortage in land use rights for residential and commercial uses due to steep increases in those prices. It is therefore necessary that the central government begin reforming land market institutions and breaking down the land supply monopoly system controlled by local governments. Our analysis features many inadequacies that should be confronted in future research. First, in this analysis, we do not directly investigate the correlation of subsidy rate t i because it is difficult to obtain data for t i . But, we can actually estimate the value of t i , by taking the following steps. First, we can estimate the land price fundamentals in each region land using a hedonic function. Next, we can obtain the value of the subsidy from the difference between the fundamental and real land price. We would like to directly investigate the correlation of subsidy rate t i, which is estimated as previously described, in our next study. Second, in the area in which land auction data cannot be obtained, we performed data compensation by making a fairly unreasonable assumption. Therefore, we should revise our data mining methods. Third, it was necessary for the analysis to incorporate geographical factors such as highways, rail, and port access, which would influence the formation of land prices. Fourth, the analysis should be extended to a broader region, for example, the Yangtze River Delta area including Shanghai and Jiangsu Province.

References

89

References Anselin L (1995) Local indicators of spatial association-LISA. Geogr Anal 27(2):93–115 Anselin L, Rey SJ (2014) Modern spatial econometrics in practice. Geoda Press LLC Brueckner JK (2003) Strategic interaction among governments: an overview of empirical studies. Int Reg Sci Rev 26(2):175–188 Buettner T (2001) Local business taxation and competition for capital: the choice of the tax rate. Reg Sci Urban Econ 31(2–3):215–245 Cai H, Vernon Henderson J, Zhang Q (2013) China’s land market auctions: evidence of corruption? RAND J Econ 44:488–521 Case AC, Rosen James HS, Hines Jr R (1993) Budget spillovers and Fiscal policy interdependence: evidence from the states. J Public Econ 52:285–307 Chen J, Hong Z (2009) Tax competition and yardstick competition. Jingji Yu Guanli Luncong 5(1):55–82 (in Chinese) Elhorst JP (2014) Spatial econometrics—from cross-sectional data to spatial panels. Springer, New York Fujii D (2016) Fiscal Competition between local governments and its sustainability in China. In: Kato H, Kajitani K (eds) Chinese-style capitalism evolving beyond dual trap: empirical studies on ambiguous institution, Minerva Shobo (in Japanese) Gao S, Liu S (2007) Land for construction of multifamily housing enters the market: difficult conditions of realities and the law. Guanli Shijie 3 (in Chinese) Guo S (2005) The property rights of farmers must be protected. In: Yongsheng T, Zhong ZZ (eds) Guangming Ribao Chubanshe (in Chinese) Guo Q, Jia J (2009) The tactical interaction among local governments, their competition in fiscal expenditure and regional economic growth. Guanli Shijie 187(10):17–27 (in Chinese) Hoyt WH (1993) Tax competition, nash equilibria, and residential mobility. J Urban Econ 34(3):358– 379 Huang X (ed) (2006) Research on China’s land management in the new era, Vols. 1 and 2. Dangdai Zhongguo Chubanshe (in Chinese) Jiang S, Liu S, Li Q (2007) Land system reform and national economic growth. Guanli shijie 9 (in Chinese) Kajitani K, Fujii D (2016) Spatial analysis of competition among local governments and the price of land: the case of Zhejiang Province. J Chinese Econ Bus Stud 14(3):229–242 Li Y, Jia R (2006) Rational reconsiderations of land expropriation and management problems in the urbanization process: analysis of surveys on 1,538 farming households which lost their land in the eastern, central, and western areas of our country. Jingji Xuejia 5 (in Chinese) Liu H, Zhang H (2006) The real estate industry and social economics. Qinghua Daxue Chubanshe (in Chinese) Liu Z (2020) China Taxation 2020. Economic Science Press (in Chinese) Ma H, Li Y, Shi G, Xu H (eds) (2003) Decoupling revenue-expenditure management system. Zhongguo Caizheng Jingji Chubanshe (in Chinese) Matsumoto M (1998) A note on tax competition and public input provision. Reg Sci Urban Econ 28(4):465–473 Ministry of Natural Resources. China land and resources statistical yearbook, Zhongguo Dizhi Chubanshe, every year (in Chinese) National Bureau of Statistics of China. China statistical yearbook. Zhongguo Tongji Chubanshe (in Chinese) NoisetLuc (1995) Pigou, Tiebout, Property taxation, and the under provision of local public goods: comment. J Urban Econ 38:312–316 Oates WE (1972) Fiscal federalism. Harcourt Brace, New York Onodera A (1997) Land system reform and city formation in China. Ajia Keizai 38(56) (in Japanese) Ottavianoa GIP, van Yperseleb T (2005) Market size and tax competition. J Int Econ 67:25–46

90

5 Land Market Competition Among Local Governments: A Spatial …

Ren Z (2009) Dilemma of government intervention in China’s real estate industry. Ajia Kenkyu 55(1) (in Japanese) Revelli F (2003) Reaction or interaction? Spatial process identification in multi-tiered government structures. J Urban Econ 53:29–53 Solé-Ollé A (2006) Expenditure spillovers and fiscal interactions: empirical evidence from local governments in Spain. J Urban Econ 59:32–53 Tanaka O (2007) Verification: contemporary chinese economic policy decision-making— approaching the limit of the reform and opening policy. Nihon Keizai Shimbunsha (in Japanese) Wang G (2006) Survey of farmers who lost their land. Xinhua Chubanshe (in Chinese) Wilson JD (1986) A theory of interregional tax competition. J Urban Econ 19:296–315 Wildasin DE (1988) Nash equilibria in models of fiscal completion. J Public Econ 35(2):229–240 Yao Y, Zhang X (2008) Race to the top and race to the bottom, IFPRI Discussion Paper, 00799 Zhang H, Chen X (2007) Fiscal competition and the structure of local public expenditure in China. Front Econ China 2(2):237–249 Zhang Y (2007) Spoiled land. Chugoku 21, vol 26 (in Japanese) Zhejiang C, Weiyuanhui NB (eds). Finance yearbook of Zhejiang. Zhonghua Shuju (in Chinese) Zhejiang S, Ju T (eds). Zhejiang statistical yearbook. Zhongguo Tongji Chubanshe (in Chinese) Zodrow GR, Mieszkowski P (1986) Pigou, Tiebout, Property taxation, and the underprovision of local public goods. J Urban Econ 19:356–370 Zhu Q (2007) Research on reform of China’s land fiscal system. Lixin Kuaiji Chubanshe (in Chinese)