A New Era

A NEW ERA CHINAʼS ECONOMY GLOBALIZES

EDITED BY DEXU HE AND CHAOYANG WANG

A New Era

Dexu He  •  Chaoyang Wang Editors

A New Era China’s Economy Globalizes

Editors Dexu He National Academy of Economic Strategy Chinese Academy of Social Sciences Beijing, China

Chaoyang Wang National Academy of Economic Strategy Chinese Academy of Social Sciences Beijing, China

ISBN 978-981-10-8356-3    ISBN 978-981-10-8357-0 (eBook) https://doi.org/10.1007/978-981-10-8357-0 Library of Congress Control Number: 2018943395 Based on a translation from the Chinese language edition: 新时代: 中国经济全球化 Copyright © Social Sciences Academic Press, 2018 All Rights Reserved © The Editor(s) (if applicable) and The Author(s) 2019 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, express 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. Cover illustration: © Yaorusheng / Getty Images Cover Design by Ran Shauli This Palgrave Macmillan 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

As Chinese President Xi Jinping noted in the report he delivered at the 19th National Congress of the Communist Party of China, “With decades of hard work, socialism with Chinese characteristics has crossed the threshold into a new era. This is a new historic juncture in China’s development.” In the new era, the principal contradiction facing Chinese society has evolved. “What we now face is the contradiction between unbalanced and inadequate development and the people’s ever-growing needs for a better life.” The period between now and 2020 will be decisive in finishing the building of a moderately prosperous society in all respects. In this period, “We must take tough steps to forestall and defuse risks, carry out targeted poverty alleviation, and prevent and control pollution.” How should we understand China’s growth model, the characteristics of its economic development, and its status in the global economy in the new era? What does “unbalanced and inadequate development” mean? What are the risks facing the Chinese economy and regulatory means at present and in the near future? All these questions are apparently very important to foreign readers and scholars who are keenly interested in China. The Editorial Department of Finance & Trade Economics has selected ten research papers published in the journal in 2016 which address the questions mentioned above from an academic perspective and in a rigorous, objective, and organized fashion. Finance & Trade Economics is an academic journal on applied economics run by the National Academy of Economic Strategy under the auspices of the Chinese Academy of Social Sciences. With a well-deserved reputation in China’s academic community, v

vi  

Preface

the journal carries articles that address major theories and practical issues concerning China’s economic development, committed to introducing new ideas and views, advancing theoretical research, and informing decision making. The paper “International Mismatch of Economic Cycle, Supply-side Reform and Medium- and High-speed Growth of Chinese Economy” looks at the international mismatch of economic cycle and its deepening after the US financial crisis, examines the background to and external environment in which the Chinese economy entered a new normal, and then discusses the supply-side structural reform which will remain an important focus of China’s economic endeavors in a few years to come. “Roles of Chinese Regions in Global Value Chains and Their Changes” proceeds from the perspective of global value chains, creates a global input-output table that covers subregions, develops the breakdown method for valueadded exports and total export value added on the basis of the table, and offers an empirical analysis of the roles that different regions of China played in global value chains as well as their changes from 1997 to 2007. It concludes that trade surpluses of regions were overestimated in traditional statistics on trade in gross value but Chinese regions have become increasingly involved in global value chains, particularly since China’s accession to the WTO. “Why China’s Gini Coefficient Has been Falling?” tries to explore the reasons why the Gini coefficient has been falling in recent years in China by decomposing the Gini coefficient and the Gini coefficient differences by income source. It concludes that while the decline in the wage inequality contributes the most to the decline in the Gini coefficient the share effect of transfers is the main factor preventing the decline. China’s macro-financial risks and monetary policy for macro-regulation always capture the close attention of researchers in other countries. “China’s Leverage Ratio and Systemic Financial Risk Prevention” is the outcome of a research led by Ma Jiantang, Executive Vice President of the Chinese Academy of Governance. The paper concludes that China’s leverage ratios are not high compared with developed countries but substantial implicit liabilities and fast growth of debt indicate considerable potential risks. The rapidly rising leverage ratio is closely connected with such factors as overreliance on indirect financing, low efficiency in the use of funds, large amounts of funds occupied ineffectively because of overcapacity, and overuse of monetary and credit policies as instruments. “A Study of the Monetary Policy Framework with Mutually Supportive Price- and

 Preface    

vii

Quantity-based Tools”, authored by two researchers from the People’s Bank of China, compares and analyzes the effect of monetary policy tools in six combinations of price- and quantity-based tools by building the Dynamic Stochastic General Equilibrium (DSGE) model encompassing residents, non-financial enterprises, and the government, and concludes that mixed policy tools are better than single policy tools, which may inform future monetary policy operations of China’s central bank. “Does the Upgrading of Industrial Structure Mean Service Sector is More Important?” constructs a DSGE model for the manufacturing and service sectors and analyzes from the perspective of the two-sector integration the economic impact of the transformation of industrial structure on the economy. It concludes that in terms of the potential of economic growth, as the contribution of the service sector to GDP grows, technological innovations in manufacturing will have a reduced impact on GDP growth, while the impact of innovations in the service sector will not change much. The service sector acts as an “economic stabilizer”, but the externality of its technological progress on the overall economic system is weaker than that of the manufacturing sector. Therefore, a balance needs to be achieved in developing the manufacturing and service sectors. Whether land finance is conducive to economic growth and how land finance affects economic growth are important issues that interest researchers very much. “How Does Land Finance Affect Economic Growth in China?” divides the economy into two sectors—traditional Malthus sector and modern Solow sector under the framework of the New-Classical Theory of Economic Growth, and concludes, based on mathematical induction for economic growth model, that land finance contributes to economic growth, which was also proved by the panel data about China. Smog is a direct cause of the general public’s concern about China’s environmental pollution. What factors can influence smog pollution? What effects have the environmental policies produced? How can China strike a balance between environmental protection and economic growth? “Does Urban Sprawl Aggravate Smog Pollution?” calculates the sprawl index of cities in China using ground-level PM2.5 concentration data, global nighttime light data, LandScan population distribution data, and economic statistics, and concludes that urban sprawl increases local PM2.5 concentrations and the size of population has a similar effect. Moreover, the correlation between urban sprawl and smog concentrations weakens as the size of cities increases, and the spatial spread of small cities lead to more serious air pollution. “Regulatory Governance, Public Appeal, and Environmental

viii  

Preface

Pollution” classifies the effects of environmental regulation into main effect, neighborhood effect, direct effect, and indirect effect; examines the modes of environmental governance in different parts of China; and concludes that there are policy interactions between different regions in environmental investment and other aspects that have significant negative impact on environmental governance. “Will Environmental Regulation Help Reduce Pollution and Improve Efficiency?” uses the State Council’s 2003 air quality control program with set deadlines as a natural experiment and adopts the difference-in-differences (DID) method based on trend score match to evaluate how effective China’s environmental regulations are in reducing pollution and improving efficiency and reveal the systematic heterogeneity therein. It concludes that by reducing pollution and emissions, environmental regulation helps grow the economy, and a positive institutional environment amplifies the positive effect of environmental regulation and reduces distortion. Hopefully, these research papers will answer some questions of foreign scholars and arouse their interest in China studies and more readers in foreign countries will follow Finance & Trade Economics and other research papers it carries that cover important topics about China’s taxation, finance, trade, industries, regions, cities, real estate, environment, internet economy, and so on. It is heartening that the journal has been included in EconLit, an economics research database run by the American Economic Association. By then, foreign readers are able to get the basic ideas of every article through a quick search. They may also visit the official website of the journal at http://www.cmjj.org/ and use its English version to check the details of the every article. Beijing, China Beijing, China

Dexu He Chaoyang Wang

The research papers are selected from the Finance & Trade Economics.

Contents

1 International Mismatch of Economic Cycle, Supply-Side Reform and Medium- and High-­Speed Growth of Chinese Economy   1 Hongju Wang and Chuan Wang 2 Roles of Chinese Regions in Global Value Chains and Their Changes  31 Hongfu Ni and Jiechang Xia 3 Why China’s Gini Coefficient Has Been Falling?  59 Tianyu Yang and Zhinan Cao 4 China’s Leverage Ratio and Systemic Financial Risk Prevention  83 Jiantang Ma, Xiaojun Dong, Hongxiu Shi, Jie Xu, and Xiaofang Ma 5 A Study of the Monetary Policy Framework with Mutually Supportive Price- and Quantity-Based Tools 113 Xiandong Yan and Yantao Zhang

xi

xii  

Contents

6 Does the Upgrading of Industrial Structure Mean Service Sector Is More Important? 137 Shenning Qu and Tie Lyu 7 How Does Land Finance Affect Economic Growth in China? 155 Shumin Yue and Yi Lu 8 Does Urban Sprawl Aggravate Smog Pollution? 175 Meng Qin, Xiuyan Liu, and Yiting Tong 9 Regulatory Governance, Public Appeal, and Environmental Pollution 203 Chao Han, Weiguang Zhang, and Shuang Shan 10 Will Environmental Regulation Help Reduce Pollution and Improve Efficiency? 237 Yu Qi, Hongyou Lu, and Ningchuan Zhang Index 269

List of Figures

Fig. 1.1

Global economic growth and contribution made by China and the US. Source: WIND 5 Fig. 1.2 Inflation rate in US, Japan, Europe. Source: WIND 7 Fig. 1.3 USD index and RMB exchange rate 11 Fig. 1.4 ETF and VIX fluctuation in emerging markets. Source: WIND 11 Fig. 1.5 Year-on-year growth of auto sales and retail sales of consumer goods13 Fig. 1.6 Fixed asset investment and real estate investment 14 Fig. 1.7 Proportion of China’s import and export in world trade (January 2007–August 2015). Source: IMF 15 Fig. 1.8 CPI and PPI (year-on-year) 16 Fig. 1.9 Contribution of accumulative final consumption and capital formation to GDP growth 18 Fig. 1.10 Funds outstanding for foreign exchange in China decreased. Source: WIND 21 Fig. 2.1 Domestic and international value chains for value-added exports and outflows in 2007 by Chinese regions 50 Fig. 3.1 Gini coefficient in China from 2003 to 2015. Source: National Bureau of Statistics 60 Fig. 4.1 Growth Rates of Local Government Debt and the Economy. Notes: (1) The local government debt in the figure only includes the debts that local governments have the obligation to repay. Contingent liabilities and debts the government has rescue responsibility for are complicated, so it is not desirable to do simple addition. There will be a detailed analysis in later sections. (2) The growth rate of local government debt for 2002 is the average annual growth rate in the period 1998–2002, xiii

xiv  

List of Figures

the figure for 2007 is the average annual growth rate in the period 2002–2007 and that for 2013 is the average annual growth rate in the period 2010–2012. The data for 2012 only cover the debts of 36 local governments audited by the National Audit Office of China. Sources: (1) GDP growth rates: National Bureau of Statistics of China, China Statistical Yearbook 2014. (2) Local debt growth rates: Audit Results of National Audit Office (No.35 of 2011, No.24 of 2013, and No.32 of 2013) 87 Fig. 4.2 Funds turnover of industrial enterprises above a designated size, 2003–2013. Note: Asset turnover ratio = sales revenue/ total assets. Source: Database of the National Bureau of Statistics94 Fig. 4.3 Working capital turnover ratio of industrial enterprises above a designated size, 2003–2013. Source: Database of the National Bureau of Statistics 94 Fig. 4.4 Receivables turnover ratio of industrial enterprises above a designated size, 2003–2013. Note: Receivables turnover ratio = revenue from main business/average balance of accounts receivable. The dotted line in the figure indicates the growth trend. Source: Database of the National Bureau of Statistics95 Fig. 4.5 Selling expenses ratio of industrial enterprises above a designated size, 2003–2013. Note: Selling expenses ratio = (selling expenses + financial expenses)/sales. Source: Database of the National Bureau of Statistics 96 Fig. 5.1 Impact of the six combinations of policy tools on GDP 129 Fig. 5.2 Impact of the six combinations on inflation 129 Fig. 5.3 Impact of the six combinations on employment 130 Fig. 6.1 Pulse reaction process of GDP under 1% manufacturing and service technology shocks. Note: The left figure shows the technology shocks based on the annual average industrial structure in 1990–2001 period (the service sector accounted for 0.36 of GDP), the right figure illustrates the technology shocks based on average annual industrial structure in the 2002–2013 period (the service sector accounted for 0.43 of GDP)152 Fig. 7.1 Local Governments’ General Public Budget Revenue, Land Transfer Revenue and Local Governments’ General Public Budget Expenditure. Source: China Statistical Yearbook and China Yearbook of Land and Resources157 Fig. 7.2 The Changes of Labor between the Malthus Sector and the Solow Sector. Note: Data of labor in the Malthus sector is

  List of Figures    

Fig. 9.1

Fig. 9.2

Fig. 10.1 Fig. 10.2 Fig. 10.3 Fig. 10.4

Fig. 10.5

xv

obtained from the number of employees in the primary industry, and data of labor in the Solow sector from the secondary and tertiary industries. Source: China Statistical Yearbook166 National Investment in Environmental Pollution Control. Source: Data was collated based on China Statistical Yearbook on Environment. However, the Yearbook does not include nitrogen oxide emissions from 2001 to 2005 205 Emissions of Major Exhaust Pollutants in the Country. Source: Data was collated based on China Statistical Yearbook on Environment. However, the Yearbook does not include nitrogen oxide emissions from 2001 to 2005 205 Air quality in Chinese cities. Source: Calculated based on data provided in China Environment Yearbook240 Distribution of city air quality. Source: Calculated based on annual average air quality data provided in China Environment Yearbook249 Air quality of cities up to standard and cities below standard, 1996–2012. Source: Calculated based on annual average air quality data provided in China Environment Yearbook250 Influence coefficient of the program on air quality, 2004–2012. Note: The Ministry of Environmental Protection and China Environment Yearbook stopped announcement of API after 2008 257 Change of the coefficients showing the impact of environmental regulation on tfp, technological progress and economic growth, 2004–2010 263

List of Tables

Table 1.1 Forecast of economic growth rate worldwide 4 Table 1.2 Forecast of global inflation 6 Table 1.3 Forecast of main macroeconomic indicators in China for 2015Q4 and the whole year 24 Table 2.1 Value-added international trade and traditional international trade by Chinese regions in 2007 (%, million USD) 45 Table 2.2 China’s foreign trade in 2007 by region (%) 47 Table 2.3 Value-added inflows and outflows by region in 2007 (million USD, %) 49 Table 2.4 WWZ-based gross exports and outflows by Chinese regions 51 Table 2.5 Breakdown of Chinese regions according to VS sources 53 Table 3.1 National Gini coefficient and shares of income sources 70 Table 3.2 Impact of wage on Gini coefficient difference 72 Table 3.3 Impact of operational income on Gini coefficient difference 73 Table 3.4 Impact of transfers on Gini coefficient difference 74 Table 3.5 Impact of property income on Gini coefficient difference 75 Table 4.1 Leverage ratios in China by sector (%) 85 Table 4.2 China’s total leverage ratio in comparison with that of other countries, 2014 (%) 90 Table 4.3 Leverage ratios in China, 2014 (%) 91 Table 4.4 Financing structure of businesses 92 Table 4.5 Benchmark interest rates in major countries, 2014 93 Table 4.6 Credit granting in China, 2008–2014 97 Table 5.1 Calibration of basic parameters 126 Table 5.2 Prior distribution and posterior distribution of modified models127 Table 5.3 GDP variance decomposition in different models % 131 xvii

xviii  

List of Tables

Table 5.4 Table 5.5

Variance decomposition for inflation in different models (%) 131 Variance decomposition for employment in different models (%) 131 Table 6.1 Fluctuation rate difference of GDP in different steady states 148 Table 7.1 Descriptive statistics of data used in the model of per capita GDP in 30 Chinese provinces 168 Table 7.2 Estimated coefficients of regression models of economic growth in China 170 Table 8.1 Statistical description of variables 189 Table 8.2 Cities with highest sprawl and most concentrated cities 190 Table 8.3 Distribution range of urban sprawl in China in the last decade191 Table 8.4 Estimated results of the panel data model 192 Table 8.5 Robustness tests by using other urban sprawl indicators 195 Table 8.6 Robustness test based on sample of non-heating cities in South China 197 Table 9.1 Definition of related variables 215 Table 9.2 Statistical description of key variables 217 Table 9.3 Spatial Durbin fixed-effect model for environmental regulatory decisions 223 Table 9.4 Test for the effects of the implementation of environmental regulation225 Table 9.5 Robustness test 231 Table 10.1 Regression results for the influence of environmental regulation on air quality 255 Table 10.2 Regression results for the general economic performance of environmental regulation 258 Table 10.3 Regression results for specific indicators 259 Table 10.4 Regression results for specific indicators 261 Table 10.5 Institutional heterogeneity of the influence of environmental regulation on environment quality 262 Table 10.6 Institutional heterogeneity of the influence of environmental regulation on productivity 264

CHAPTER 1

International Mismatch of Economic Cycle, Supply-Side Reform and Medium- and High-­ Speed Growth of Chinese Economy Hongju Wang and Chuan Wang

Abstract  The Chinese economy entered the “new normal” in 2015 and China successfully fended off the financial risks through the flexible use of monetary and financial policies. More effective fiscal policies were adopted, the economic growth slowed down steadily and Chinese economy realized soft landing. In the meantime, commodity price in production sector was low, so deflation in the consumption sector must be avoided, but the economic structure was continuously improved, and innovation and entre-

Fund project: 2nd-batch emergency project of the National Natural Science Foundation in 2015: “Study of International Price Trend of Bulk Commodities and Influence of Incoming Deflation” (approval number: 71541014); “Study of the Mechanism of How Debt Disposal Cycle Affects Deflation Expectation” (approval number: 71541015); project under the innovation program of CASS: “Study of the Innovation and Improvement of Macro-regulatory Approaches and Risk Prevention and Control” (project number: 2016CJY008). CLC number: F124.1; Document code: A; Serial number: 1002-8102(2016)02-0000-15 H. Wang (*) • C. Wang National Academy of Economic Strategy, CASS, Beijing, China © The Author(s) 2019 D. He, C. Wang (eds.), A New Era, https://doi.org/10.1007/978-981-10-8357-0_1

1

2  

H. WANG AND C. WANG

preneurship created new driving forces. In 2016, the global economy is likely to continue to recover slowly, but the mismatch of economic cycle will aggravate the risk of economic decline. In the first year of the 13th five-year period, the Chinese economy still faces downward pressure. We need to step up the supply-side reform and work hard on both the supply end and demand end in order to meet the goal of stabilizing the economy while making progress and keep the economic growth rate in the reasonable medium- and high-speed range. Keywords International mismatch of economic cycle • Supply-side reform • Medium- and high-speed growth In 2015, the international mismatch of economic cycle continued, as reflected in the different economic growth rate and goods price tendency in different countries and consequently the enhanced conflicts in monetary policies. The global differentiation in economic tendency and monetary policy made the international financial market more volatile. Emerging market countries faced the serious challenge of capital outflow and exchange rate depreciation, while China should be vigilant against the potential pressure arising from the international mismatch of economic cycle (Wang 2014). The Chinese economy entered the “new normal” in 2015 and China successfully fended off the financial risks through the flexible use of monetary and financial policies. More effective fiscal policies were adopted, the economic growth slowed down steadily, commodity price in production sectors was low, the economic structure was continuously improved and innovation and entrepreneurship created new driving forces. The year 2016 is the first year of the 13th five-year period, and understanding, adapting to and leading the new normal is the paramount task in China’s economic development at present and for some time to come. Generally speaking, China still faces downward pressure on economic growth, but it is likely to bottom out with the emergence of new growth drivers. Against the background of the international mismatch of economic cycle, China will step up the supply-side reform in 2016. In face of the pressure of capital outflow and RMB depreciation, we will need supportive macro policies on the demand side, prevent the supply-side reform measures such as decapacity and de-stocking from dragging down short-term economic growth, prevent economic and financial risks and maintain steady economic operation in the reasonable medium- and high-­speed range.

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

3

1   Deepening of International Mismatch of Economic Cycle1 1.1  Global Economic Growth Slowed Down, Economic Cycle Mismatch Worsened 1. Global economic growth slowed down; Brazil and Russia were in economic decline. According to the forecast released by IMF in October 2015, the global economic growth rate would reach 3.1% in 2015, 0.2% less than the forecast in July. Among that, developed countries would reach the economic growth rate of 2.0%, 0.1% less than the forecast in July; emerging market countries would reach 4%, 0.2% less than the July forecast. The global economic growth rate was expected to reach 3.6% in 2016, 0.2% less than the July forecast, 2.2% for developed countries and 4.5% for emerging market countries. As far as developed countries were concerned, IMF expected the US to reach the economic growth of 2.6% in 2015, 0.1 percentage point less than the July forecast. In comparison, the economic growth in EURO zone and Japan remained sluggish. IMF maintained the 1.5% forecast for the EURO zone and reduced the forecast for Japan’s economic growth by 2 percentage points to –0.6%. As to emerging market countries, IMF expected India to surpass China and be the fastest-growing economy and forecast its economic growth to remain at the 7.3% and 7.5% growth rate in 2015–2016. It maintained the 6.8% forecast for China’s economic growth. On the other hand, IMF reduced the economic growth forecast for other emerging market countries such as Russia and Brazil notably. It expected Russia’s and Brazil’s economy to slow down by 3.8% and 3%, respectively, in 2015, 0.4% and 1.5% down from the July forecast. Its forecast for their economic growth in 2016 was pessimistic too, which was −0.6% and −1%, respectively (Table 1.1). 2. Of the three major economies, the American economy increased, European economy was recovering and Japan was in decline. Of the three major economies, the US maintained steady economic growth with an annualized GDP growth rate of 2.1% in Q3, to which personal consumption contributed 2.05%, government consumption and investment ­contributed 0.29% and the contribution by net export of goods and services and domestic private investment was negative. The steady economic growth boosted employment, which then increased income and eventu1

 Data all come from Wind unless otherwise noted.

4  

H. WANG AND C. WANG

Table 1.1  Forecast of economic growth rate worldwide

World Developed countries US EURO zone Japan Britain Emerging market countries China India Russia Brazil South Africa

2013

2014

2015

2016

3.3 1.4 2.2 −0.4 1.6 1.7 5.0 7.7 6.9 1.3 2.7 2.2

3.4 1.8 2.4 0.8 0.1 2.9 4.6 7.4 7.3 0.6 0.1 1.5

3.1 (−0.2) 2.0 (−0.1) 2.6 (0.1) 1.5 (0) 0.6 (−0.2) 2.5 (0.1) 4.0 (−0.2) 6.8 (0) 7.3 (−0.2) −3.8 (−0.4) −3.0 (−1.5) 1.4 (−0.6)

3.6 (−0.2) 2.2 (−0.2) 2.8 (−0.2) 1.6 (−0.1) 1.0 (−0.2) 2.2 (0.0) 4.5 (−0.2) 6.3 (0) 7.5 (0) −0.6 (−0.6) −1.0 (−1.7) 1.3 (−0.8)

Source: IMF (2015a) Note: The number in the parentheses is the revision of forecast

ally boosted the consumption and economy, forming a benign cycle. Although the expansion of the manufacture industry slowed down a little, the real estate market was still in good shape and consumers were positive about economic prospects. It is expected that in 2015Q4, the recovery of the real estate market will help drive the economy, and the steady improvement in employment will help increase people’s income and consequently boost domestic demand. Moreover, the holiday season represented by Thanksgiving and Christmas will invigorate the consumption even though the bullish USD will still exert limited adverse effects on the economy. The EURO zone registered a year-on-year GDP growth of 1.6% and quarter-on-quarter GDP growth of 0.3% in Q3, basically on a par with the Q2 growth. In terms of specific countries, France had a year-on-year GDP growth of 1.2% and quarter-on-quarter GDP growth of 0.3% in Q3, better than expected; and that in Germany was 1.7% and 0.3%, respectively, dragged by its weak export and investment, indicating a narrowed advantage over other countries in the EURO zone. It is expected that the EURO zone will continue the economic recovery in Q4, manufacture expansion will speed up and the relaxed monetary policy adopted by the European central bank and the weak Euro will further bolster the economy despite the downward pressure of inflation. The actual initial GDP in Japan was decreased by 0.2% quarter-on-­ quarter and 0.8% year-on-year in Q3, with the second quarterly negative

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

5

growth in a row. In Q4, Japan’s sluggish economy does not improve, corporate investment is reduced considerably, inventory decreases quarter-­ on-­quarter and enterprises are reluctant in production and investment. While the employment market is steadily reaching the state of full employment, Japan will suffer from weak domestic demand and export as well as low inflation. Under the impact of a range of factors, such as weak demand, exchange rate fluctuation and the adjustment of global industrial chains and economic structure, global trade had the biggest slump in 2015 since the financial crisis in 2008. In the first half of the year, the total trade in goods of G7 and BRICS countries had a year-on-year slump of 10.9%. 3. China made prominent contribution to global economic growth. Since China is the second largest economy in the world, its economic growth is of great significance for the world as the global economy is in the stage of profound adjustment. Figure 1.1 shows that China’s contribution rate to global economic growth has been increasing steadily since 2000 and approached 30% in 2014, and the combined contribution rate by China and the US was higher than 50% in the past three years.

5.6

%

% 60

4.9

50

4.2

40 30

3.5

20

2.8

10

2.1

0

1.4

-10

0.7

-20

0.0 1990

-30 1992

1994

1996

1998

2000

2002

GDP: actual YOY growth (global) World Bank: contribution rate to world economic growth (exchange rate method): U.S. (right axis)

2004

2006

2008

2010

2012

2014

World Bank: contribution rate to world economic growth (exchange rate method): China (right axis)

Fig. 1.1  Global economic growth and contribution made by China and the US. Source: WIND

6  

H. WANG AND C. WANG

1.2  Differentiation of Commodity Price Worldwide The continuous drop of international commodity price since 2015 has brought downward pressure on the PPI in raw material importers, which was transferred to the consumer price index (CPI) and led to the low price of consumer goods. According to IMF’s forecast, the international crude oil price in 2015 dropped by 46.4%, and the price of other commodities than crude oil dropped by 16.9%. As a result, developed countries were under obvious deflation pressure and their CPI growth is expected to be 0.3% in 2015, while that in emerging market countries will be 5.6%. It is also expected that the price of crude oil and other commodities will continue to fall in 2016 but at a lower rate compared with 2015. In 2016, CPI growth in developed countries will improve a little and emerging market countries will have a higher CPI than them (Table 1.2). The US had a low inflation in October. Its post-adjustment CPI in October increased by 0.2% quarter-on-quarter and year-on-year each, and key CPI excluding food and energy increased 0.2% quarter-on-quarter and 1.9% year-on-year, reaching the 2% target set by the Federal Reserve. PPI reduction in October narrowed to 0.3% quarter-on-quarter and remained 4.1% year-on-year. Generally speaking, the current inflation level remains low in the US, but the prospects are good in the long term. The University of Michigan released the latest one-year and five-year inflation forecast in November, which was 2.7% and 2.6%, respectively. Inflation remained low in the EURO zone and Japan. The falling oil price offset the price hike in other goods and services. Data show that CPI in the EURO zone increased by 0.1% year-on-year in October, and core CPI excluding energy and food price increased by 1.1% year-on-year. The year-on-year PPI reduction in September increased from 2.6% to 3.1%, the largest reduction since January 2015, and the EURO zone faced serious Table 1.2  Forecast of global inflation 2015 Crude oil price Bulk commodity (exc. crude oil) CPI (developed countries) CPI (emerging market economies)

−46.4% (−7.6%) −16.9% (−1.3%) 0.3% (0.3%) 5.6% (0.1%)

2016 −2.4% (−11.5%) −5.1% (−3.4%) 1.2% (0.0) 5.1% (0.3%)

Source: IMF (2015a) Note: The number in the parentheses means the revision range of the forecast made in July 2015

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

3.5

%

%

7 3.5

3.0

3.0

2.5

2.5

2.0

2.0

1.5

1.5

1.0

1.0

0.5

0.5

0.0

0.0

13-12

14-02 14-04 14-06

14-08 14-10 14-12 15-02 15-04 15-06

Euro zone: core HICP (core CPI): year-on-year

U.S.: core CPI: year-on-year

15-08 15-10

Japan: core CPI: year-on-year

Fig. 1.2  Inflation rate in US, Japan, Europe. Source: WIND

deflation risks. Japan’s CPI in October increased by 0.3% year-on-year, and the core CPI excluding food and energy increased by 0.7% year-on-­ year, still on a low level (Fig. 1.2). 1.3  Federal Reserve’s Interest Rate Increase Triggered Differentiation of Global Monetary Policy 1. Federal Reserve’s interest rate increase and future tendency of American monetary policy. The Federal Reserve announced on December 17, 2015 to increase the federal funds rate by 25 base points to 0.25–0.50, which was voted in favor by all the ten FOMC members. This meant the Federal Reserve ended the seven-year-long zero interest rate policy and the first interest raise was finally a reality after a half-year delay. To achieve this goal, the Federal Reserve raised the IOER rate to 0.5% and the ONRRP by 20 base points to 0.25% and canceled the daily upper limit of USD300 billion. It also increased the discount rate by 25 base points to 1%, effective on December 17, 2015. However, the FED maintained the policy of investing its institutional bonds and MBS principal back in institutional MBS and extending the duration of its government bonds through tender off. After FED announced the resolution, JP Morgan, Wells Fargo, Bank of America, Citibank and US Bancorp all announced to raise the “most favorable loan interest rate” by 25 base points.

8  

H. WANG AND C. WANG

The FED decided to raise the interest rate at such a point mainly based on the judgment that the economic growth was stabilizing and inflation and unemployment indicators bettered. Regarding the inflation trend, although FED’s goal of 2% inflation was not met, and the core CPI excluding energy and food was still on the decrease, core PCE, the indicator that truly affected FED’s policy decision, obviously climbed up after 2015. PCE means personal consumption expenditure deflator. Compared with CPI that is based on a basket of fixed goods, PCE can reflect the price of substitute goods bought by the consumers due to price change and is therefore considered better able to gauge America’s inflation more comprehensively and stably. It caught the attention of the FED in 2002 and later became a factor that influenced its decision. In terms of the overall operation of American economy, its GDP was above 2% for six quarters in a row and approached 3% several times, indicating a sound and robust economic status. As to unemployment rate, its recovery was quite slow after the economic shock, and the US government aimed to lower the unemployment rate to 6%, which was actually reached in October 2014. The employment market improved further in 2015 and the unemployment rate was falling steadily. As to the future, FED updated its forecast of economic data and interest rate and the mean number showed that it expected the interest rate to reach 1.375% at the end of 2016, which means the FED will raise the interest rate three or four times in 2016, obviously faster than before. 2. The global monetary policy was divided after FED’s interest rate hike, but it did not change the relaxed monetary policy in Europe and Japan. In the EURO zone, the European central bank further cut the negative deposit interest rate on December 3 and extended the government bond purchase program to March 2017. Considering the low international price of crude oil, deflation pressure and shaky economic recovery, the European central bank is likely to maintain the relaxed monetary policy in 2016. In general, the relaxed environment in Europe will remain unchanged and the QE policy will extend to September 2016 at least. Similar to the situation in the EURO zone, the Japanese central bank did not make new moves by November 2015 after it announced to increase bond purchase at the end of October last year and continued to increase basic currency at the yearly rate of JPY8 billion. Given Japan’s bleak economic prospects in 2016, its central bank will probably intensify the relaxed monetary policy.

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

9

Oil exporters and Hong Kong, the latter’s exchange rate being pegged to USD, also increased their interest rate. The Mexican central bank decided to increase interest rate for the first time since 2008 and raised the overnight rate by 25 base points to 3.25%; Hong Kong Monetary Authority announced to follow in America’s steps to raise the interest rate by 0.25% to 0.75%, the first interest hike for nearly seven years, in order to maintain the USD-pegged HK dollar system; and Saudi Arabia, Kuwait, UAE and Bahrain also followed suit after FED because oil, their main income source, was priced with USD, so they needed to maintain their exchange rate. Argentina deregulated its foreign exchange system and gave up the efforts to maintain its exchange rate, which led to the serious currency devaluation. After the FED announced the interest rate hike, Argentina announced to abandon the four-year-long capital regulation and allowed free float of exchange rate, and its central bank would convert the USD3.1  billion RMB reserve assets into USD for better liquidity. The exchange rate of Argentine peso against USD fluctuated drastically, falling by 41% at most before regaining a little. 3. Emerging market economies faced the pressure of capital outflow and exchange rate depreciation. Owing to a range of factors, such as FED’s interest rate hike and consequently the drastic fluctuation in the exchange rate of main currencies, falling oil price, turbulence in local regions, and insufficient effective demand, cross-border flow of international capital intensified in 2015. To be more specific, the continuous economic recovery in the US and FED’s interest hike accelerated the flow of international capital back to the US, which, combined with the slow recovery in other main economies, led to the strengthening of USD index and then the sharp price drop of USD-priced bulk commodities. In comparison, the USD became a safe choice for investors because it was more secure than other international assets. Regarding the cross-border capital flow, the net capital inflow of emerging markets since 2015 was much smaller than the average monthly net inflow in the past five years. In the third quarter, emerging markets experienced net outflow of non-civilian financial assets, and in August alone, the net outflow of stock capital was USD8.7 billion and net inflow of bond capital was reduced to USD4.2 billion. Net o ­ utflow mainly took place in Asia and Europe because the expectation of FED’s interest rate hike, the falling energy price and the pressure of economic slowdown in emerging markets resulted in the reconfiguration of financial assets.

10  

H. WANG AND C. WANG

Because of the FED’s interest rate raise, the flow of international capital from emerging markets to developed economies was more obvious. It is expected that in 2016, American economy and the financial system will get substantial recovery. They may have potential crisis in 2016, thus creating the trajectory from the financial crisis in the US to the EURO crisis and then to the economic crisis in emerging markets. Going forward, the situation of the strengthening of USD and the weakening of currencies in other countries after the FED’s interest increase will continue in the short term, accompanied by the inevitable short-term backflow of international capital to the US, with the emerging markets under particularly heavy pressure of capital outflow. According to IMF’s forecast, emerging market countries will have capital outflow in both securities investment and direct investment in 2016, but of a reduced scale compared with 2015 4. Global financial market remained unstable. Regarding the impact on the fundamentals, the FED’s interest rate increase, first and foremost, continued the sluggish transaction of bulk commodities. Many bulk commodities in the international market were still priced with USD, and its aggressive position was strengthened after the interest hike, which led to the pressure of the rebound of bulk commodity price. In the foreign exchange market, the USD index has been on the rise since the beginning of the year, which was backed by America’s continued economic recovery, the interest rate increase of USD and the weak economic recovery and exchange rate depreciation in emerging market countries (Fig. 1.3). In the third quarter of 2015, the average GDP-weighted value of 15 emerging market currencies against USD value dropped by 9.4%, a reduction rate second only to that during the Asian financial crisis in the 1990s. As FED’s interest hike further boosted the backflow of international capital, currencies in emerging market countries might depreciate even more. At last, in the capital market, stock markets in emerging countries shrank substantially in Q3 due to the sustained outflow of international capital from those countries, and their stock markets were much more volatile in Q3 than before and even approached the level during the market fluctuation in 2011. Since FED’s interest hike will also exert adverse effects on the capital market of developed countries, it is expected that as emerging market countries make in-depth economic adjustment, capital outflow will exacerbate their market volatility (Fig. 1.4).

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

11

Fig. 1.3  USD index and RMB exchange rate % 40

54

36

48

32

42

28

36

24

30

20

24

16

18

12

12 14-01-31 14-03-31 14-05-31 14-07-31 14-09-30 14-11-30 15-01-31 15-03-31 15-05-31 15-07-31

U.S.: ETF volatility index in emerging markets (right axis)

U.S.: Standard & Poor 500 volatility index (VIX)

Fig. 1.4  ETF and VIX fluctuation in emerging markets. Source: WIND

2   Supply-Side Structural Reform Is a Long and Hard Journey The Chinese economy entered the “new normal” in 2015, which featured reduced economic growth, improved economic structure and the shift of growth drivers. Although it experienced the stock market disaster and capital outflow, structural conflict entailed the supply-side reform. As the reform goes deeper, the space for economic development is expected to be

12  

H. WANG AND C. WANG

further broadened in 2016, laying a solid foundation for the economic and social development in the 13th five-year period. 2.1  Slower Demand Increase and Steady Deceleration of Economic Growth The shift from high-speed to medium- and high-speed economic growth is a typical characteristic of the “new normal” of Chinese economy. After the international financial crisis in 2008, the global economy was faced with great pressure, including the slower demand increase, transformation of development model and upgrade of industrial structure. Starting from 2010, China’s economic growth rate has been reduced continuously. The GDP in the first three quarters of 2015 increased by 7.0%, 7.0% and 6.9%, respectively, indicating a seasonally adjusted quarter-on-quarter growth of 1.3%, 1.8% and 1.8%, respectively. Economic growth in Q4 faces downward pressure, but some economic indicators will bottom out. The Q4 GDP is expected to increase by 6.9% year-on-year and the all-year growth is expected to reach 6.9% too, down 0.4 percentage point from a year earlier. Economic growth in 2015 avoided a nosedive and realized steady deceleration. 1. The trend of industry and service industry differed and industrial structure was adjusted. Due to factors such as the overcapacity in manufacturing, stronger efforts to eliminate backward industries, reduced growth of leading industries such as automobile and communication and weak investment demand, industrial production maintained a low-speed growth in the first three quarters of 2015, but the service industry increased at a faster pace. In the first three quarters, secondary industry in China increased by 6% year-on-year, including a 5.9% year-on-year growth for industry, down 2 percentage points from the same period of last year, and an 8.4% growth for tertiary industry, up 0.8 percentage point from the same period of last year. In November, the industrial value added of enterprises above the designated scale increased by 0.58% from October. The year-on-year growth was 6.2%, 0.6 percentage point greater than in the previous month. The manufacture increased by 7.2% in November and displayed signs of bottoming out. The PMI in the manufacture released by the National Bureau of Statistics fell under the 50% line in August and further down to 49.6% in November; non-manufacture business activity index was 53.6% in November, up 0.5 percentage point from the previous month, and non-manufacture sector continued to expand

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

13

steadily, playing a growing part in stabilizing the economic growth. In the first three quarters, the tertiary industry accounted for 51.4% of GDP, 10.9 percentage points more than the proportion taken by the secondary industry. 2. Consumption growth slowed down and picked up again in Q4. From January to November 2015, resident consumption in China increased by 10.6% year-on-year, 1.4 percentage points loler than a year earlier. Year-on-year growth of resident consumption rose to 11.2% in November, up 1.2 percentage points from the low point of 10% in April. In terms of the category of consumption, auto sales rebounded obviously (Fig.  1.5). The auto sales statistics of enterprises above the designated scale increased by 9% year-on-year in November, up 10.3 percentage points from the low point of −1.3% in March. Consumption was relatively stable among the top three demands, and total retail sales of consumer goods throughout the year are expected to increase by 10.6% year-on-year. 3. Growth of fixed asset investment slowed down notably, investment climbed up a little in Q4, but real estate investment was still in negative growth. Investment growth was reduced considerably. The fixed asset 16.0

14.0 12.0

Auto sales (%) Retail sales of consumer goods (%)

10.0 8.0 6.0 4.0 2.0 0.0

-2.0 -4.0

Fig. 1.5  Year-on-year growth of auto sales and retail sales of consumer goods

14  

H. WANG AND C. WANG

investment in January–November 2015 increased by 10.2% year-on-year, on a par with the increase in January–October and 5.6 percentage points lower than the same period of last year. Among that, accumulative real estate investment increased by only 1.3% year-on-year, 10.6 percentage points lower than a year earlier; manufacture investment increased by 8.4% year-on-year, 5.1 percentage points lower than a year earlier; service industry investment increased by 11.0% year-on-year, 6.1 percentage points lower than a year earlier; and infrastructure investment in that period increased by 18.0% year-on-year, 2.8 percentage points lower than a year earlier. Fixed asset investment rebounded slightly in Q4 but real estate investment was still in negative growth. The fixed asset investment growth in November climbed back to 10.8% from 6.8% in September, but real estate investment was in negative growth for four months in a row and the decline worsened to 5.1% in November (Fig. 1.6). Considering that there would not be a significant increase in investment in December, the all-year fixed asset investment in 2015 is expected to increase by 10.3% year-on-year. 4. Import and export growth slowed down, trade surplus increased and share in export market expanded. From January to November 2015, China’s accumulative export and import dropped by 3% and 15.1% year-­ on-­year, respectively, and accumulative trade surplus increased by 61.8% 20 15 10 5 0 -5 -10

Fixed asset investment YOY (%) Real estate investment YOY (%) I

II

III 2014

IV

I

II

III 2015

Fig. 1.6  Fixed asset investment and real estate investment

IV

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

15

year-on-year to reach USD539.1 billion. The main reason for the growing trade surplus was the massive reduction of import rather than the increase of export, indicating an obvious shrinkage of demand. To be more specific, China’s accumulative export to the US and Association of Southeast Asian Nations (ASEAN) in January–November increased by 4.1% and 3% year-on-year, respectively, and that to Japan and EU decreased by 9.6% and 4.6% year-on-year, respectively. The massive export reduction to Japan and EU was partly caused by the shrinking demand in those two regions due to their economic slowdown, and partly by the significant appreciation of RMB against Japanese Yen and EURO. In Q4, RMB is depreciated and the American economy maintains the recovering momentum, so export is expected to improve slightly in December, and the all-year export and import is likely to drop by 1.7% and 15.2% year-on-year, respectively. Regarding the proportion of China’s import and export in world trade, the average proportion of China’s export was 12.9% in 2014 and 14.5% in January–August 2015, indicating an increase rather than decrease in the proportion (Fig. 1.7). The average proportion of China’s import in the same period was 8.5% and 8.7%, respectively, meaning that the reduced growth of Chinese economy did not dampen the global trade demand.

18%

Proportion of China's import in world import (%, left axis)

16%

Proportion of China's export in world export (%, left axis)

14%

China's trade surplus (mil USD, right axis)

12% 10% 8%

6%

100,000

90,000 80,000 70,000

60,000 50,000 40,000

2%

30,000

0%

20,000

2007/1 2007/5 2007/9 2008/1 2008/5 2008/9 2009/1 2009/5 2009/9 2010/1 2010/5 2010/9 2011/1 2011/5 2011/9 2012/1 2012/5 2012/9 2013/1 2013/5 2013/9 2014/1 2014/5 2014/9 2015/1 2015/5

4%

Fig. 1.7  Proportion of China’s import and export in world trade (January 2007–August 2015). Source: IMF

16  

H. WANG AND C. WANG

10.0 CPI YOY PP YOY

8.0 6.0 4.0 %

2.0 0.0 -2.0 -4.0

2015/11

2015/7

2015/3

2014/11

2014/7

2014/3

2013/11

2013/7

2013/3

2012/11

2012/7

2012/3

2011/11

2011/7

2010/11

-8.0

2011/3

-6.0

Fig. 1.8  CPI and PPI (year-on-year)

2.2  Production Sector Cut Output and PPI Was in Negative Growth for a Long Time The CPI in November 2015 increased by 1.5% year-on-year and PPI decreased by 5.9% year-on-year, a difference of 7.4% (Fig. 1.8). There was great downward pressure on commodity price in 2015. Although we cannot say that China is in comprehensive deflation gauged by CPI, CPI is not enough to measure the commodity price changes because in China, consumption takes up only a small proportion in GDP, and investment and production sectors take up a much larger proportion in national economy than in developed countries. In 2014, the proportion of total fixed capital formation and secondary industry (mainly manufacture) in GDP was 46.5% and 42.7%, respectively. The PPI has been in negative growth for more than three years, longer than the period of economic restructuring from June 1997 to December 1999 because of the Asian financial crisis, and it has seriously cramped the investment and production sectors. First, growth of fixed asset investment kept falling. From January to November, the nationwide fixed asset investment increased by 10.2% year-on-year, 5.6 percentage points lower than the same period last year. Second, increase in industrial production and profit slowed down seriously. From January to October, the total profits

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

17

of industrial enterprises above designated scale decreased by 2% year-on-­ year, 0.3 percentage point more than the growth reduction from January to September. Their total profits in October alone decreased by 4.6% year-­ on-­year, 4.5 percentage points more than in September. Third, NPL ratio increased. In the first three quarters of 2015, commercial banks’ NPL ratio was 1.59%, the highest since 2009Q4. Although the NPL ratio released by authorities remained low, it was still 0.43 percentage point higher than in the same period last year due to factors such as more losses of enterprises. Fourth, growth of fiscal income decreased. From January to November, general public budget revenue nationwide increased by 5.7% year-on-year, and government fund budget revenue decreased by 21.4% year-on-year. Fifth, GDP reduction index became negative. The serious contraction in production sector resulted in the general economic cramp. We estimate the GDP reduction index to increase by −0.31%, 0.1%, and −0.51%, respectively, in 2015Q1, Q2 and Q3 year-on-year. The concept and standard for determining deflation aside, we, in view of China’s special economic structure, believe that the contraction of production sector has led to a series of consequences, such as reduced economic growth, falling price of industrial goods, reduced profits of enterprises and reduced growth of fiscal revenue. We should pay close attention to the rising probability of economic risks. 2.3  Structural Conflicts Entailed Supply-Side Reform, Economic Structure Was Continuously Improved and Innovation and Entrepreneurship Created New Drivers Against the general background of the “new normal”, China’s economic growth has bidden farewell to the age of high growth, high investment and high input. The main reasons for reduced economic growth and consistent negative growth of PPI are structural overcapacity, heavy debts of some industries, enterprises and local governments and the ­unsustainability of real estate investment and land finance. Those structural conflicts entailed the supply-side reform. In 2015, the economic structure was continuously improved, industrial structure was upgraded and innovation and entrepreneurship created new economic drivers. Structural adjustment has taken primary effect. 1. Consumption and investment structure are improved. During the economic decline, resident income increased faster than GDP, employment was stable, the income gap between urban and rural residents nar-

18  

H. WANG AND C. WANG

rowed and consumption functioned as a stabilizer of economic growth. In the first three quarters, the accumulative final consumption spending contributed 58.4% to GDP growth, 9.3 percentage points more than a year earlier; and accumulative capital formation contributed 43.4% to GDP growth. Monthly data show that the growth of fixed asset investment dropped much faster than consumption growth (Fig. 1.9). 2. The structure of secondary and tertiary industries was improved. In the first three quarters, value added of tertiary industry accounted for 51.4% of GDP, 2.3 percentage points higher than the same period of last year and 10.8 percentage points higher than the proportion of secondary industry. In the secondary industry, emerging industries developed ­rapidly. In the same period, value added of high-tech industries increased by 10.4% year-on-year, 4.2 percentage points higher than that of industrial enterprises above the designated scale; and investment in high-tech industries increased by 16.1% year-on-year, 5.8 percentage points higher than the growth of general investment. The service industry saw some transformation. Traditional service industries increased relatively slowly, while financial and other service industries grew rapidly. In Q3, the financial industry increased by 16.1% year-on-year, driving GDP to increase by about 1.3 100.000 90.000 80.000

Accumulative final consumption Accumulative capital formation

70.000 60.000

50.000 40.000

30.000 20.000

10.000 0.000

Fig. 1.9   Contribution of accumulative final consumption and capital formation to GDP growth

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

19

percentage points year-on-year. Other service industries increased by 9.5% year-on-year, 0.8 percentage point faster than Q2 and driving GDP to increase by 1.8 percentage points year-on-year, more than the contribution by the financial industry. Investment in high-tech service industry increased by 22.6%, 11.4 percentage points higher than the investment in the service industry in general. 3. Resident income increased steadily, annual target of employment was met in advance and income gap between urban and rural residents narrowed further. In the first three quarters, the per capita disposable income of national residents increased by 9.2% year-on-year nominally and 7.7% actually, 2.6 and 0.8 percentage point higher than the nominal and actual GDP growth, respectively. There are 10.66 million new jobs created accumulatively, meeting the annual target in advance, and the surveyed unemployment rate in 31 big cities basically stayed around 5.1%. The per capita disposable income of rural residents increased by 9.5% nominally and 8.2% actually, which was 8.4% and 6.8% for urban residents. 4. Innovation and entrepreneurship created new economic drivers. Production that complied with the general trend of transformation and upgrade increased rapidly. From January to October 2015, high-tech industries increased 4.3 percentage points faster than industries above designated scale; and online retail sales nationwide increased by 33.0% year-­ on-­year, faster than the growth of total retail sales of consumer goods. In October 2015, express delivery business nationwide increased by 50.4% year-on-year; online traffic of cellphone doubled for the 10th consecutive month; output of new energy vehicle increased by 1.9 times year-on-year and that of smart TV increased by 66.5%. Business convenience measures invigorated the market. Since 2015, 12,000 new enterprises are registered every single day on average. In October alone, the value added of industrial enterprises of non-public ownership above designated scale increased by 7.7% year-on-year, 2.1 percentage points higher than enterprises above designated scale in general. From January to October, profits of private enterprises above designated scale increased by 6.2% year-on-year, better than the performance of industrial enterprises in general in that period. 2.4  Successfully Preventing Financial Risks Through Flexible Use of Monetary and Financial Policies Economic growth has slowed down, and hazards of potential financial risks in the stock market, real estate market, bond market and foreign

20  

H. WANG AND C. WANG

exchange market are mounting, but the central government has successfully prevented financial risks by flexibly wielding a range of policies. 1. Stabilizing the stock market. The Chinese stock market fluctuated drastically in 2015 due to a number of factors. Domestic reasons included excessive speculation earlier on, consequences of the deleveraging and bubble-squeezing measures, future reform of the registration system and the increase of stock supply by launching the NEEQ and other measures; external reasons included worries about FED’s interest rate raise, reduced global economic growth, capital outflow and underselling caused by the expectation for RMB depreciation. In the transaction week from August 18 to 26, Shanghai Composite Index slumped by more than 1000 points, but the stock market resumed steady growth after a range of recovering measures were put forth at the end of August. By December 25, 2015, the Shanghai Composite Index climbed up 12.16% from the beginning of the year, higher than the stock market growth in developed countries such as the US, Britain, Germany, France and Japan as well as in Asian countries such as Thailand, Vietnam, Indonesia, the Philippines, Malaysia and Singapore. 2. Stabilizing the housing market. To cope with the pressure of real estate sales, the central government put forth a series of favorable policies in 2015 to boost housing sales, including interest rate cut, reduction of down payment and adjustment of the housing accumulation fund policy, which proved very effective. In the second half of 2015, more cities saw quarter-on-quarter and year-on-year increase in the housing price, and the increase was more substantial. 3. Stabilizing the foreign exchange market. RMB faced major depreciation pressure and there was a mounting pressure of capital outflow in 2015. Capital and financial accounts, the items in international balance of payments that reflect international capital flow, began to reveal signs of capital outflow in 2014. Financial account became negative in 2014Q2 and kept falling, even at a faster pace after 2015 (Fig. 1.10). In 2015Q3, China’s financial account deficit reached the historical new high of USD63.4 billion. Capital outflow directly changed the trend of China’s foreign exchange reserve, which has been shrinking since the second half of 2014 till the present day. In November 2015, China’s foreign exchange reserve fell to USD3438.284 billion from the peak of USD3993.213 billion in June 2014, a contraction of USD554.929 billion within one and a half years. To mitigate the pressure of capital outflow, the exchange rate reform came into force on August 11, whereby the RMB depreciated

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…    100 million USD 800

21

100 million USD 800

600

600

400

400

200

200 0

0 -200

-200

-400

-400

2007-12

2008-12

2009-12

Actual FDI

2010-12

2011-12

Balance of trade in goods

2012-12

2013-12

2014-12

2015-12

Central bank: new funds outstanding for foreign exchange

Fig. 1.10  Funds outstanding for foreign exchange in China decreased. Source: WIND

against the USD in a controllable range. By December 25, 2015, the exchange rate of RMB against USD depreciated by 4.3% from August 10. 4. Financial marketization and internationalized reform were promoted in an orderly manner. On March 26, the central bank issued a notice to simply the issuance and management procedures for credit asset-based securities; on April 1, the State Council approved the Implementing Plan for Deposit Insurance System; on June 2, the central bank issued and officially implemented the Provisional Methods for the Management of Certificate of Deposit; on July 21, ten central departments issued the Guiding Opinions on Promoting the Healthy Development of Internet Finance; on August 11, the market-based determining mechanism of the central parity of RMB exchange rate was launched; on September 15, the central bank reformed the RRR assessment system from the point-in-time methods to averaging method; on September 30, the central bank allowed overseas institutions to engage in the Chinese interbank foreign exchange market; on October 23, it lifted the upper limit on deposit rate and completely deregulated the interest rate; on November 22, it cut the SLF rate in branches and sub-branches, and an interest rate corridor of China’s monetary policies primarily took shape; on November 30, RMB joined the SDR, marking a substantial progress on its internationalization.

22  

H. WANG AND C. WANG

2.5  More Effective Fiscal Policies Were Adopted to Boost Economic Growth Relevant authorities changed the regulatory thinking in 2015, adjusted the fiscal policies, created new methods and approaches, improved the macro-regulatory measures and approaches on the supply side while stabilizing the total demand and earnestly promoted the stable economic growth of better quality and higher efficiency. (1) Give play to the role of government investment in stabilizing growth and adjusting structure. The central finance continued to exert the critical role of investment, sped up general budget expenditure, strongly promoted the PPP model and encouraged private investment to participate in the construction, management and operation of municipal public facilities, toll road, eco-­ environmental projects and so on. Efforts were made to continue to support urban security housing projects and further intensify the renovation of dangerous rural housing. The spending structure of the special fund for water conservancy projects was improved, and industrial upgrade and SME development were bolstered. The central finance insisted on market orientation and government guidance, and fiscal capital was mainly channeled to key areas of manufacture transformation and upgrade, including smart manufacturing and high-end equipment. Steps were also taken to set up the national fund for entrepreneurship and investment in emerging industries, and measures were adopted to innovate in the ways of input, systems and mechanism, and to strengthen water pollution prevention and control. (2) Local bonds were issued steadily to dissolve the local government debt risks. The central finance took measures in 2015 to steadily issue local bonds, appropriately enlarged the scale of local government bond issuance, and continued to issue local government bonds to replace inventory debt. It also issued special bonds to support investment projects in 2015. 2.6  Outlook of China’s Macroeconomic Status in 2016 Based on the analysis above, industrial production and investment recovered slightly in 2015Q4 and consumption growth was basically stable. The macroeconomic trend in 2016 will still be affected by factors such as de-capacity, bubble squeezing and debt clearing, enterprises’ inventory may increase and economic growth still faces downward pressure. We should prevent excessive de-capacity from causing negative price growth of consumer goods. Real estate sales showed signs of recovery, but hous-

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

23

ing de-stocking still has a long way to go, and whether the recovered sales can boost real estate investment is still to be seen. The international mismatch of economic cycle and the turbulent financial market pose a serious challenge to domestic economic and financial stability, making it more difficult to apply the monetary policies, and more fiscal expenditure is needed. Let us assume that the turmoil in the international financial market would not lead to systemic risks in 2016, and the basic forecast is based on the following assumptions: (1) USD fluctuation. The US has seen faster economic recovery than other developed countries, and the fluctuation in USD exchange rate is closely related to the relative speed of its economic recovery against the economic recovery in Europe, Japan and other regions, but an excessively aggressive USD is not good for America’s economic development. Let us assume the USD index sways in the 90–100 range. (2) Interest rate increase in the US. The FED’s interest rate increase in the future may be less frequent than any time in history. Let us assume it will raise the interest rate twice in 2016, 0.25% each time. (3) Oil price remained low. The crude oil price fell back to about USD40/barrel. Let us assume the New  York crude oil price will stay at USD45/barrel in 2016. (4) According to the members announced by the FED and the December bitmap forecast report made by the president of the Federal Reserve Bank, we assume the American economic will grow by 2.4% in 2016; and according to the World Economic Outlook that IMF released in October, we assume the 1.6% and 1.0% economic growth in the EURO zone and Japan, respectively. Assumptions about the policy variable are as follows: (1) One interest rate cut and three RRR cuts. Given the economic growth, commodity price changes, financial market conditions and international economic environment, there would not be much room for interest rate cut in the future, but the RRR may be cut again. Let us assume there will be one interest rate cut and three RRR cuts in 2016. (2) RMB depreciates by about 4% against USD, and China’s foreign exchange reserve is reduced by about USD400 billion. (3) Fiscal deficit increased to about RMB2 trillion and deficit ratio may reach 2.76%, a net increment of about RMB380 billion. Based on the above assumptions and using calculating models, we forecast the economic operation (Table 1.3). Based on the above analysis, we expect the GDP to increase by 6.9% and consumer goods price by 1.5% in 2015Q4, and 6.9% and around 1.4% for the whole year, respectively. We also expect the economic growth to bottom out in 2016, all-year eco-

24  

H. WANG AND C. WANG

Table 1.3  Forecast of main macroeconomic indicators in China for 2015Q4 and the whole year Main economic indicators 1. PPI growth rate (YOY, %) 2. CPI growth rate (YOY, %) 3. Actual GDP growth rate (YOY, %) 4. Nominal growth rate of total retail sales of consumer goods (YOY, %) 5. Nominal growth rate of total fixed asset investment (YOY, %) 6. Nominal growth rate of real estate investment (YOY, %) 7. Nominal growth rate of total export (YOY, %) 8. Nominal growth rate of total import (YOY, %) 9. M2 balance (YOY, %) 10. Credit (YOY, %)

2015Q3

2015Q4 forecast

2015 all2016 allyear forecast year forecast

−5.0 1.4 6.9 10.5

−5.9 1.5 6.9 11.3

−5.2 1.4 6.9 10.6

−1.8 1.3 6.7 10.8

10.3

10.4

10.3

9.8

4.6

−4.0

1.0

1.1

−1.6

−2.5

−1.7

1.6

−15.2

−15.3

−15.2

−0.8

13.1 15.7

13.5 15.1

13.1 15.6

12.8 14.3

nomic growth to be around 6.7%, total retail sales of consumer goods to increase by about 10.8%, and fixed asset investment to increase by about 9.8%. Meanwhile, export will increase by 1.6% or so year-on-year, but import will decrease by −0.8% or so year-on-year. All-year PPI will decrease by about 1.8% year-on-year, a reduced decline compared with 2015, and all-year CPI will increase by about 1.3%, a little lower than in 2015. M2 and credit will increase by about 12.8% and 14.3% year-on-year, respectively. In general, against the background of the reinforced supply-side reform in 2016, we still need supportive macro policies on the supply side and should prevent certain supply-side reform measures from impeding the short-term economic growth and prevent economic and financial risks, so as to ensure the stable economic operation in a reasonable range.

3   Accelerating Supply-Side Reform Medium- and High-Speed Growth

and Maintaining

The Chinese economy entered the new normal in 2015 and the 13th Five-­ year Plan starts with 2016. Therefore, understanding, adapting to and leading the new normal is the general context of China’s economic devel-

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

25

opment at present and for some time to come. International mismatch of economic cycle will continue to exist in 2016 and China’s economic growth will still face downward pressure, but with the deepening of the supply-side reform and the continuous emergence of new growth drivers, the economy is likely to bottom out and maintain medium- and high-­ speed growth. IMF’s report showed that structural reform can considerably improve the TFP and macroeconomic performance (IMF 2015b, c). According to information in World Bank’s database, the 4.9–7.2% range is defined as medium- and high-speed economic growth since 1960. China’s main macroeconomic indicators at the moment are better than those of the economies that have medium- and high-speed growth. China has the conditions and potential to maintain medium- and high-speed economic growth, and it will soon enter the new normal where medium- and high-­ speed growth is matched with medium and high income (Wang 2015a). 3.1  The Five Development Concepts Provide Theoretical Guidance for Economic Transformation and Supply-Side Reform The fifth plenary session of the 18th CPC Central Committee released the 13th Five-year Plan, in which the Five Development Concepts centered on “innovation, harmony, green, openness and sharing” were put forth for the first time, providing the theoretical foundation for economic transformation and supply-side reform, the fundamental goal of which is to improve the TFP. Economic growth is decided by the amount and efficiency of factor input. Having enjoyed satisfactory growth for more than 30 years, China is now faced with a number of challenges, including the reduced increase of international demand, vanishing demographic dividend and mounting resource and environmental pressure. The old path of extensive growth that relied on the increase of capital, labor force and natural resource input does not work anymore. Guided by the Five Development Concepts, the government will launch a series of strong and practical reform plans in areas concerning regional development, industrial development, SOEs, fiscal and tax matters, finance and livelihood, with a view to promoting economic transformation. 1. The medium- and long-term structural factors decide that China has to shift to the path of innovation-driven development in order to maintain medium- and high-speed growth. The first is government innovation. The government improves the efficiency of public resource allocation and its administration through its own reform. The second is innovation by

26  

H. WANG AND C. WANG

enterprises and public institutions. Stimulate their innovative vigor and enhance their innovation capability. Set up a batch of national labs in major innovation areas; foster a group of leading innovative enterprises with international competitiveness at a faster pace, and support the healthy development of high-tech SMEs. Systematically carry out reform experiments on comprehensive innovation in provinces and regions with the necessary conditions, set up a number of platforms for industrial technological innovation and strongly cultivate and develop strategic emerging industries. The third is small and micro innovation. Carry out mass entrepreneurship and innovation and foster supportive platforms for that purpose through Internet-based mass entrepreneurship, crowd-sourcing, mass support and crowd funding. 2. Regional and urban-rural disparity creates immense space for China’s development. The first is coordinated regional development. The Belt and Road Initiative, Beijing-Tianjin-Hebei coordinated development and the Yangtze River Economic Belt—these three strategies constitute a full landscape of coordinated regional development, while the pan-Pearl River Delta also plays an active role. We will also continue to implement the overall strategy for the development of four old industrial bases in the eastern, central, western and northeastern part of China. The second is integrated development of urban and rural areas. Accelerate the people-­ oriented urbanization, implement the reform of household registration system and ensure equal access to public services for all people. The third is the coordinated material and cultural development. Speed up the cultural reform and development, reinforce the socialist cultural development; build a socialist cultural power; intensify the sense of integrity and honesty, nation, the rule of law and social responsibility and advocate the scientific spirit. 3. Adapt to the global climate change and jointly protect the earth. The first is low-carbon development. Raise the proportion of non-fossil energy and foster an energy system that is clean, low-carbon, safe and efficient. The second is intensive and efficient use of resources and strengthening the management of restrictive indicators. The third is conducting management through market approaches. Determine the standards for water, energy and land consumption and carry out transaction of water right and carbon emission right. 4. Secure new dividend of opening up. Build the community of shared destiny, foster an in-depth integrated situation of mutually beneficial cooperation, promote two-way opening up and support coastal areas to

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

27

participate in global economic cooperation and competition across the board. Improve the law-based, internationalized and convenient business environment, better the service and trade facilitation system and comprehensively apply the management system featuring pre-access national treatment and negative list. Accelerate the construction of free trade zones, the Belt and Road construction, and the economic integration of Hong Kong, Macao, Taiwan and the mainland. Actively participate in global economic governance and undertake international responsibilities and obligations and engage in the negotiation on how to deal with global climate change and the 2030 agenda for sustainable development. 5. Sharing in the fruits of development is the precondition for sustainable development. To achieve the goal of doubling the GDP and the per capita income of urban and rural residents in 2020 from the 2010 level, we need to implement various poverty alleviation policies. In the innovation-­driven modern society, all people shall share in the fruits of development and their creativity should be brought into full play to boost sustainable development. 3.2  Maintaining Stability While Making Progress and Ensuring Medium- and High-Speed Economic Growth The central economic work meeting specified the general tone of the economic work in 2016 is “maintaining stability while making improvement”. “Stability” means we should maintain the medium- and high-speed economic growth and keep economic operation in the reasonable range. “Improvement” means we should create new supply and new demand, boost economic transformation and upgrade and elevate industrial development to the medium and high end through supply-side reform. As real estate investment is hard to rebound, domestic demand in 2016 still faces major downward pressure, and the slow recovery of most economies around the world dampens external demand, which, combined with the reduced growth of financial industry due to the large base number in 2015, will lead to substantial downward pressure on the economy in 2016, and it may be more appropriate to lower the economic growth target slightly. In the 13th five-year period, the bottom line of economic growth is 6.5%, and the growth target for 2016 can be set around 6.7%. With the united efforts of all sides, the Chinese economy will maintain mediumand high-speed growth in 2016 and industrial development will move to the medium and high end.

28  

H. WANG AND C. WANG

3.3  Five Major Tasks Point Out the Work Priorities During the Periodic Adjustment The central economic work meeting put forth five major tasks in 2016, namely de-capacity, de-stocking, deleveraging, lowering cost and strengthening weak links. The first four tasks reflect how hard the periodic adjustment is amid economic decline. De-capacity refers to eliminating the surplus production capacity in traditional industries, especially those that feature massive assets and periodic operation; de-stocking refers to the de-­ stocking of real estate industry; and deleveraging means lessening the debts of enterprises and local governments and controlling financial risks. Cost reduction is achieved mainly through tax and fee reduction and energy price adjustment, including lowering the VAT for manufacture, which is of great significance for controlling corporate cost, and lowering the social insurance fee and possibly integrating the “five insurances and one housing fund”, which will increase the actual income of employees. Strengthening weak links mainly means strengthening the weak links in livelihood, products and industries, for which poverty alleviation, product and industrial upgrade and agriculture are emphasized. De-capacity is good for the long-term and healthy economic development, but in the short term, it may continue to largely reduce the investment in industries with overcapacity and dampen enterprises’ financing demand. Accelerated de-capacity may increase unemployment and impede the rise of resident income, subjecting the economy to major deflation risk. It is based on these considerations that the central economic work meeting stressed the importance of “properly handling the relation between maintaining social stability and promoting structural reform”. This means that de-capacity will be conducted in a gradual and orderly manner and that on the supply side will be carried out slowly. 3.4  Reinforced Fiscal Policies Will Hold the Bottom Line To maintain a relatively high-speed economic growth, the government may step up the efforts to stabilize growth, especially “strengthening the proactive fiscal policies”. (1) “Properly increase the necessary fiscal expenditure and government investment”. This means fiscal expenditure will continue to increase at a fast pace but would not be beyond imagination. (2) Implement tax-reducing policies, which is a reform measure on the supply side aimed to lower the enterprises’ cost and improve their operat-

  INTERNATIONAL MISMATCH OF ECONOMIC CYCLE, SUPPLY-SIDE REFORM…   

29

ing efficiency. (3) Raise the fiscal deficit ratio periodically. China set the fiscal deficit at RMB1.62 trillion in 2015, including RMB1.12 trillion central deficit and RMB500 billion local deficit, with the deficit ratio rising to about 2.4%. The fiscal deficit is expected to rise to around RMB2 trillion in 2016, the deficit ratio may reach 2.76% and net deficit increment will be increased by about RMB380 billion. The measures mentioned above aimed to cut tax and increase expenditure are good for boosting economic growth, but generally speaking, fiscal approaches remain the policies that hold the economic bottom line. To lift the economy out of the low ebb, we cannot totally rely on the central finance and leverage but have to complete the deleveraging of overcapacity industries and local governments first. 3.5  Sound and Slightly Relaxed Monetary Policy Guarantees the Stable Operation of Financial Market The deepened international mismatch of economic cycle has put more pressure on China’s capital outflow and RMB depreciation and posed a serious challenge to China’s monetary policy. The FED’s interest rate hike took place in the same period as China’s reduced economic growth, structural adjustment, debt deleveraging and the internationalization of the RMB. To prevent FED’s interest rate hike from triggering systemic risks in emerging market economies (including China), mitigate the vicious impact on Chinese financial market and real economy caused by the turbulence in international financial market and prevent systemic crisis, we need to consider the interactions of factors such as domestic economic growth and asset price, RMB exchange rate and capital flow, seek the new balance point between internal and external economy (Wang 2015b) and guarantee the stable operation of the financial market. (1) Total social demand remains weak and we need to keep up the stimulus on the supply side. (2) In view of the supply-side reform, provide appropriate liquidity for the real economy and lower its financing cost. (3) In light of the structural fiscal policy, create a favorable interest rate environment for various special fiscal bonds, local government debts and the issuance of national bond, and mitigate the pressure of inventory bond at the same time. (4) After the FED raises the interest rate, the RMB will be depreciated mildly to deal with capital outflow. We expect one interest cut and three RRR cuts in 2016, and the RMB exchange rate against USD will depreciate by about 4%. (5) Giving the contracted foreign exchange reserve and funds

30  

H. WANG AND C. WANG

outstanding for foreign exchange, new channels of currency supply will be created. On the whole, the monetary and financial policies, under the framework of macro-prudential management, will have to meet the goal of stabilizing the financial market, boosting economic growth and facilitating structural reform.

References Wang, Hongju. “Prevent the Aggravation of Economic Cycle Mismatch between the New Normal in China and the U.S.” Economic Perspectives 7 (2014). Wang, Hongju. “Range, Conditions and Policy Selection for Medium and High-­ speed Economic Growth in China.” Globalization 6 (2015a). Wang, Hongju. “Long-term Stagnation, Cycle Mismatch and Financial Stability: New Challenge to Monetary Policy.” The Chinese Banker 10 (2015b). IMF, World Economic Outlook—Adjusting to Lower Commodity Prices, October (2015a). IMF, Structural Reforms and Macroeconomic Performance: Initial Considerations for the Fund. November (2015b). IMF, Structural Reforms and Macroeconomic Performance: Country Cases. November (2015c).

CHAPTER 2

Roles of Chinese Regions in Global Value Chains and Their Changes An Analysis Based on a Global Input-Output Model in Which Chinese Regions Are Embedded Hongfu Ni and Jiechang Xia Abstract  Not only has the division of labor system in global value chains developed extensively between countries, but it also has spread into individual countries in the form of domestic value chains. This is particularly significant in the case of China that is vast in territory and rich in natural resources with great regional differences. This chapter creates a global input-­output table that covers subregions, and on this basis, develops the CLC Number: F74; Document Code: A; Article No.: 1002-8102 (2016) 10-0087-15 Funded study projects: “A Study of Paths and Strategies for Broadening the Opening-up of China’s Services Sector” (14ZDA084), The National Social Science Fund of China; “A Study of the Impact of Population Aging on Regional Economy: Based on a OLG-CGE Model” (71401009), National Natural Science Foundation of China; “Transfer of Industries between the Regions of China during Economic Transformation: Observations, Mechanisms and Policies” (71473054), National Natural Science Foundation of China. H. Ni (*) • J. Xia National Academy of Economic Strategy, CASS, Beijing, China © The Author(s) 2019 D. He, C. Wang (eds.), A New Era, https://doi.org/10.1007/978-981-10-8357-0_2

31

32  

H. NI AND J. XIA

­ reakdown method for value-added exports and total export value added, b and offers an empirical analysis of the roles that different regions of China played in global value chains as well as changes in 1997, 2002 and 2007. Research findings show that: (1) Trade in value added varied widely from region to region in China. In traditional statistics on trade in gross value, trade surpluses of regions were overestimated, and there even were phenomena that some regions had a trade surplus in gross-value terms but a trade deficit in value-added terms. (2) The Chinese inland regions achieved the value-added export directly by offering intermediate products to the coastal regions. Traditional statistics on trade in gross value generally overrated the coastal regions and underrated the inland regions in terms of their contributions to foreign trade. (3) Following China’s accession to the WTO, regions all over the country had a fast-increasing degree of vertical specialization, with the coastal regions outperforming the inland ones in this respect. But there were different structural factors responsible for rising vertical specialization. For the inland regions, it was largely because of increased connection with other regions in terms of vertical specialization, and in the case of the coastal regions, it was with other countries in the world. Keywords  Global value chain • Domestic value chain • Value-added export • Global input-output model

1   Introduction Production of goods and services in today’s world is no longer confined to individual countries but take place in a division of labor system in global value chains (GVCs). That is to say, a country (region) imports intermediate products or raw materials, employs its labor, capital and other factors to manufacture products, adds value to them and sells them to other countries (regions) as final products. Because traditional statistics on the total value of trade focus simply on the total value of trade, thus containing repeated calculations, they can neither actually reflect the value added that is hidden in imports nor embody objectively the competitiveness and international position that individual countries (regions) have in the global production system. In recent years, there have emerged many new m ­ ethods for GVCs measurement aimed to understand more accurately the evolution of GVCs and the positions and roles of countries therein. These new

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

33

methods may roughly be divided into two types. One type is measuring methods based on data from business or product surveys, notably those employed in cast studies on internationalized products such as iPod, iPad, iPhone (Linden et  al. 2009; Dedrick et  al. 2010) and Barbie (Tempest 1996). The other type is indirect methods based on input-output models. This type of methods expand country- (region-) specific input-output models into global (or international) models, as in classic studies by Hummels et al. (2001), Johnson and Noguera (2012), Koopman et al. (2014) (aka KWW (2014)) and Wang et al. (2013) (aka WWZ (2013)). But these indirect methods are based either on a model of input and output of a particular country or on a model of global input and output between countries without linking input and output by region within a country to input and output between countries, making it possible to build intermediate input relations of subregions of a country with other countries and then analyze GVCs directly from the perspective of intermediate input relations between those subregions and foreign countries. That is to say, present studies put focus on GVCs between countries but neglect domestic value chains within individual countries (regions). In a big country (e.g. China, the United States or India) with marked differences between regions within it, the division of labor in value chain terms is much more obvious and even occurred much earlier than that in GVCs. With information technology (IT) developing fast and the cost of information transmission and transportation declining considerably, manufacturing tasks and particular activities concerning a product may be carried out simultaneously not only in different countries but in different regions (areas) of individual countries, and as a result the value chains between regions of individual countries and those between countries will be closely linked together to form GVCs.1 Particularly in China, where there are distinct differences between its regions, positions and roles that the regions have in domestic value chains and GVCs will be widely different. Because these regions differ in the level of development and regional advantages, some of them (e.g. coastal regions) contribute the majority of the c­ ountry’s imports and exports, but main raw materials and semi-finished goods used for production may come from inland regions. That is to say, an inland 1  For the convenience of expounding, this chapter refers to the value chains between regions within a country as “domestic value chains” and those between countries as “international value chains”. GVCs in this chapter is relatively broad concept, that is, including “domestic value chains” and “international value chains”.

34  

H. NI AND J. XIA

region may contribute exports indirectly by supplying intermediate products and play an important role in GVCs. Therefore, it is of great theoretical and practical significance to embed input-output tables regarding the regions of a particular country in global input-output tables and measure and analyze the positions and roles that those regions have in domestic value chains and GVCs. This helps (1) identify concrete forms of manifestation of GVCs on international and national levels as well as relations of international value chains to domestic ones; (2) understand economic and technical relations between regions within a country and through what channels they participate in GVCs and then analyze from GVCs’ perspective the positions and roles of those regions and (3) measure GVCs more precisely so as to offer reliable information to policymakers. This chapter tries to embed input-output tables by Chinese regions into global inputoutput tables by countries, expands and applies a global input-output model that covers countries to a global input-output model that covers subregions and finally estimates and analyzes the positions and roles that Chinese regions have in domestic value chains and GVCs as well as dynamic changes thereof. This chapter has made the following contributions. (1) It tried to develop a new method that incorporates input-output tables by region of a country into global input-output tables by countries. Given data availability, employing this new method it produced global input-output tables by Chinese regions for 1997, 2002 and 2007. Meng et al. (2013) developed only a global input-output table in which Chinese regions were embedded for 2007, but with different methods (e.g. a different balance method was used) from this chapter. (2) It analyzed, from the angle of close relations of the domestic value chains to the international value chains, how GVCs create and distribute value worldwide and in Chinese regions, as well as dynamic changes thereof. By contrast, Su (2016) and Meng et  al. (2012) studied the domestic value chains of China with regional input-output tables and Trade in Value-Added (TiVA) methods, but failed to link Chinese regions with other countries; instead they assumed foreign imports and exports in the input-­ output tables regarding Chinese regions, that is, only using a closed economic model in dealing with the input-output tables by Chinese regions. Although Meng et al. (2013) produced the global input-output tables in which Chinese regions were embedded for 2017, one year’s data cannot be used for analyzing changing trend. (3) It expanded the breakdown methods for value-added exports and the WWZ (2013)

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

35

export value added, broke down value-added exports by Chinese regions through domestic and international value chains, analyzed the levels of participation and the roles that Chinese regions had in GVCs through the WWZ (2013) breakdown method concerning the export value added and further broke vertical specialization (VS) by places of origin down into VS from domestic other regions and from other countries. By contrast, Meng et al. (2013) only employed the KWW (2014) breakdown method.

2   Literature Review and Linking Methods 2.1  Literature Review The outsourcing of large amounts of parts and components to particular specialized enterprises led to sizable trade in intermediate goods globally and even within individual countries, with nearly two-thirds of global trade being trade in intermediate goods. This phenomenon is called “vertical specialization”, “production segmentation”, or more fashionably, “global value chains”.2 While there have emerged some theoretical studies on intermediate goods (Findlay 1978; Grossman and Rossi-Hansberg 2008; Costinot et  al. 2013), GVC measurement has increasingly been improved. Existing GVCs measuring methods may roughly be divided into two types, as described below.  easurements Based on Business or Product Surveys M Measurements based on business or product surveys are relatively direct and easy to understand. Xing and Deter (2010) found in their study that from every USD 178.9 iPhone exported, China could obtain only a value of USD 6.5, or approx. 3.6%, while all the remainder of the value was shared by the United States, Germany, Japan, South Korea among other countries. Such a thorough investigation, limited to value chain analysis with respect to a particular product, could hardly reflect the domestic value-added ratio (DVAR) of a country on the whole as well as international relations between industrial sectors, and is unable hence to provide information on value chains within a country (region) and at industrial level, that is, unable to measure the position and role of a country (region) in GVCs. Another typical microscopic method is based on the DVAR with 2  In different studies this is variably known as vertical specialization, production fragmentation, outsourcing, global supply chains and global value chains.

36  

H. NI AND J. XIA

respect to exports in processing trade; it requires the availability of a large amount of survey data on small businesses as well as statistical trade data from customs to calculate the DVAR for experts. This method was used in studies by Upward et  al. (2013), Zhang et  al. (2013), Kee and Tang (2015) and so on. Because the method needs to distinguish between customs data on processing trade and general trade (e.g. in the cases of China and Mexico), it is impossible to measure DVAR at business level for countries where no statistical data on processing trade is available.  easurement Based on Input-Output Models M Indirect measuring methods based on input-output models, which are free of the flaw of microscopic measurements that can hardly factor in the relations between industries, are able to measure to what degrees a country and its sectors of the economy participate in GVCs. Since the turn of the century in particular, as many international organizations and institutions have developed global input-output tables, indirect measuring methods based on input-output models have increasingly been improved. These methods may further be divided into measuring methods based on non-competing input-output tables regarding individual countries (regions) (Hummels et al. 2001; Ping et al. 2006; Lawrence et al. 2007), and measuring methods based on global (international input-output tables). The former type is flawed: First, input-output tables regarding individual countries (regions) are unable to take into consideration international industrial relations, making it impossible to consider spillover effects and feedback mechanisms in relation to other countries (regions). Second, TiVA for countries A and B that is obtained based on their non-­ competing input-output tables ignores mechanisms for indirect TiVA influence from third-party countries. A product exported to the country B from the country A may be directly consumed in the country B, and it is also likely that the product is used as intermediate input to produce another product which the country B exports to a third country, meaning a decrease in value-added exports from the country A to the country B. And if the product is exported from the country B to a third country as intermediate input and the third country exports the final product back to the country B, there would be an increase in value-added exports from the country A to the country B. In order to overcome the aforesaid flaw, measuring methods based on global (international) input-output models have been developed, as in studies by Johnson and Noguera (2012), KWW (2014), WWZ (2013) and so on. This type of methods has the difficulty

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

37

of working out global input-output tables between countries. While KWW (2014), with a breakdown of total export value added, put under one framework the VS measuring methods developed by Hummels et  al. (2001) and Johnson and Noguera (2012), the nine-component formula used by KWW (2014) regarding total exports were restricted to national levels, unable to look further into specific sectors. WWZ (2013), on the other hand, went further to expand total export breakdown formulae to value-added breakdown formulae regarding bilateral exports by sectors (industries), resulting in a 16-component formula. This chapter applies the WWZ (2013) method to a global input-output model in which regions are embedded. The wide application of the above-mentioned breakdown methods in global input-output models regarding value-added exports and export value added is attributed to the compilation and development of global input-output tables. At present, major global input-output databases include EU’s WIOD (World Input-output Database), Purdue University’s GTAP (Purdue University), the Asian International I-O Table by Institute of developing Economies IDE-JETRO, OECD/WTO’s ICIO (Inter-­ Country-­Input-Output Model) and UNCTAD/Eora’s GVC Database. These databases contain no data on domestic value chains for regions within individual countries, though they each have an abundance of data. A global input-output model focuses on GVCs between countries but neglects domestic value chains within individual countries (regions) like China in particular where there are wide regional differences that make the division of work in the value chain system more prominent. To more accurately analyze to what degree and in what way China participates in GVCs, it is necessary to look deep into domestic value chains and the relationship between domestic and international GVCs in the context of GVCs. It is for this purpose that this chapter tried to embed China’s regional input-output tables in global input-output tables, created the global input-output tables in which Chinese regions were embedded for 1997, 2002 and 2007, and focused on the roles of Chinese regions in the global input-output tables and changes thereof. 2.2  How to Link Regional Input-Output Tables with Global Input-Output Tables This section offers a brief introduction to the main principles and steps of how to link inter-regional input-output tables of a country with global

38  

H. NI AND J. XIA

input-output tables.3 Given that the State Information Center of China generated the inter-regional input-output tables (IRIOTs) regarding eight regions within China for 1997, 2002 and 2007, this chapter chose only the 1997, 2002 and 2007 WIOTs from the WIOD (World Input-Output Database)4 for the linking purpose in a way that is largely consistent with Dietzenbacher et al. (2014). As the IRIOTs are concerned only with 17 sectors, this chapter combined the WIOTs and the IRIOTs according to these sectors. Constrained by availability of data, assumptions were made so as to calculate related data. It was assumed, for instance, that the coefficient of intermediate input from abroad for the subregions in the WIOTs-­ IRIOTs was the same as that for China in the WIOTs, and that the coefficient of distribution abroad for the subregions in the WIOTs-IRIOTs was the same as that for China in the WIOTs. Such fixed proportional coefficient assumptions are in practice often used in researches on inter-­regional or non-competing input-output tables; Johnson and Noguera (2012), for instance, employed fixed proportional coefficient when generating global input-output tables with input-output tables concerning countries. Below is a brief introduction to the steps of embedding the IRIOTs in the WTOTs: (1) Merge 35 sectors in the WIOTs into 17 sectors and convert the IRIOTs into USD-denominated input-output tables using annul average exchange rates. (2) Generate WIOT-IRIOT tables and enter directly data that need not be calculated, with final demand broken down into final consumption, capital formation and stock and with an aggregate value-added account created for remuneration for labor, capital and so on; (3) Based on the fixed coefficient assumptions, measure the distribution worldwide of exports from the Chinese regions by using the structured data in WIOTs and the trade data on the Chinese regions. (4) Adjust the ROW account so that such variables in other accounts as total intermediate input, value-added, total investment, are total output remain the same in value as much as possible as those in the WIOTs and IRIOTs. Total consumption expenditure in final consumption is also made consistent with that in the original WIOTs and IRIOTs as much as possible. Adjust to zero any negative value that appears in the ROW account. (5) Adjust 3  Restricted by length requirements, the chapter did not give a detailed description of the tabulation method, which the authors may make available upon request. 4  The WIOD project, which lasted three years (from May 2009 to April 2012) and involved 11 research organizations with 4 million euros in funding, covered 27 EU countries and 13 non-EU countries, involving 35 industrial sectors and 59 product sectors over the span 1995–2011. WIOD data is available at: http://www.wiod.org/database/index.htm.

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

39

stocks to balance the entire WIOT-IRIOT table and adjust the domestic stocks of the corresponding production sectors so that the row sum is equal to the column sum, that is, total input is equal to total output. It should be noted that, when embedding the IRIOTs, the chapter regarded the total amounts in the IRIOTs as control data and made WIOT-IRIOT calculations based on the structurally proportional coefficients in the WIOT tables relating to China and where imbalance items arose; these items were all attributed to other regional accounts in the WIOT tables. There were several reasons for doing so. First, as China in the WIOTs included trade data on Hong Kong and Macao—that is, mainland China, Hong Kong and Macao were in the WIOTs seen as a whole,5 the total amounts in the IRIOTs were regarded as control data, considering discrepancies that occurred when China in the original WIOTs was replaced by its regions. Second, when producing WIOTs, the WIOD project also treated other regions of the world as a remainder item, so it is justified and consistent that this chapter attributed errors to the accounts of other regions in the world when balance adjustment was made. Third, considering that this chapter focuses on the relations between Chinese regions and other major countries of the world, it is not inappropriate to treat those countries as a remainder item. Finally, because the data on imports, exports and output of the Chinese regions contained in the generated global input-output tables in which the Chinese regions were embedded might be slightly different than statistical data of other sources, all data used in this chapter, for the purpose of data comparison and uniformity and unless particularly noted, came from the authors’ global input-­output tables in which the Chinese regions were embedded.

3   Measurement and Structural Breakdown Methods 3.1  TiVA and Structural Breakdown  alue-Added Exports in a Global Input-Output Model in Which V Subregions Are Embedded VS and TiVA measuring methods are largely derived from the Leontief inverse. To make it easy to expound and understand, use a WIOT-IROIT  See Dietzenbacher et al. (2013). In the WIOTs, mainland China, Hong Kong and Macao together were treated as a whole. 5

40  

H. NI AND J. XIA

model including two countries and three subregions for an explanation. According to the TiVA calculation method used by Johnson and Noguera (2012), the following formula is arrived at:

TiVA rC = V r BY C

(2.1)

Where B is the Leontief inverse in the WIOT-IRIOT model; non-zero elements in Vr correspond to the value-added ratio of region r, and other elements are row vectors of 0; YC is the vector of country C’s demand for final products; and TiVArC is the total value-added exports from region r to country C. To obtain total value-added exports of region (country) r, it is necessary to add together value-added exports of different countries (regions), namely, TiVA r = ∑TiVA rC . And when Vr is substituted by Vi r C ≠r

(where non-zero elements are the value-added ratios of sector i in country r, and other elements are row vectors of 0), the result will be the value-­ added exports ( TiVA ri C ) from industrial sector i in country r to country C. 3.2  Breakdown of Value-Added Exports Through Domestic and International Value Chain Channels  reakdown of Value-Added Exports Between Countries (Regions) B In the global input-output model in which subregions are embedded, value-added exports from subregion E to country H may be through multiple ways; for instance, region E exports intermediate products to country G which then exports the products to country H, with value-added exports from subregion E to country H being in fact implemented via relations between countries (international value chains). If region E has no trade ties with other countries and supplies intermediate products to another region (M) within the country, and if region M exports the products to country H, then value-added exports from region E to country H are carried out through domestic inter-regional relations (domestic value chains). The above two paths of value-added export are obviously different. Therefore, by differentiating the channels for value-added exports from regions to other countries, we may have a clear understanding of how Chinese regions are embedded in GVCs. Below is a breakdown of value-­ added exports by region based mainly on block matrices.

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

TiVA EH = v E  B EW B EM B EE = v E B EEY EH +

∑

B EH

B EL  Y WH

Y MH

Y EH

Y HH

Y LH 

T

v E B ErY rH + v E B EHY HH + v E B ELY LH

r∈{W ,M}

41



(2.2)

where BEr stands for a block matrix in inverse matrix B; YrH stands for final demand of country H from region r; T stands for matrix transposition. The first item in formula (2), VAED1 (value-added exports by domestic segment of GVC), refers to value-added exports as a result of direct exports form region E to country H through domestic value chain channels. The second item, VAED2, means value-added exports when region E exports intermediate products to other regions (W, M) where the intermediate products are used to produce final products that are exported to country H as products of final demand, which exports are also value-added exports from region E to country H through domestic value chain channels. Because these two items are indirectly implemented through mainly domestic inter-regional intermediate input relations (BEE, BEr), they belong to domestic value chain channels. The third item, VAEI1 (value-added exports by international segment of GVC), means value-added exports when region E exports intermediate exports to country H where they are used to produce final products for domestic consumption. The four item, VAEI2, means value-added exports when region E exports intermediate products to country L where they are used to produce final products that are exported to country H as products of final demand. VAEI1 and VAEI2 are value-added exports that stem mainly from intermediate input relations between countries (or between regions and countries) and may therefore be ascribed to international value chain channels. It is very natural to generalize the above breakdown method regarding bilateral national (regional) valueadded exports to the total value-added exports and value-­added outflows ① of region r.  he Extraction Method T The application of the TiVA calculation formula in the global input-­output model in which subregions are embedded as well as of the block matrix method may distinguish the value-added exports and value-added outflows of region r implemented via domestic value chains from those of it

42  

H. NI AND J. XIA

via international value chains.6 Such an addition breakdown based on domestic and international value chains may reveal how Chinese subregions participate in GVCs. While some regions may realize value-added exports and added value outflows largely through domestic value chain channels, others may do so through international value chain channels. In the above breakdown, however, if only block matrices of domestic subregions (E, M, W) are involved, this chapter would consider value-added exports and added value outflows to have been realized through domestic value chain channels; wherever a block matrix involves countries (H, L), it would be considered that value-added exports and added value outflows were realized through international value chain channels. On the other hand, Brs and Brr do not completely represent domestic value chains, and block matrices Brs and Brr are arrived at by using the direct consumption coefficients in the global input-output tables in which subregions are embedded. bRS and bRR in fact also include the influence of international input-output relations indirectly. To arrive at value-added exports and added value outflows of pure domestic value chains, this chapter adopted the extraction method for processing purposes, assuming that there exist no intermediate input relations between countries in the global input-­ output tables in which regions are embedded, that is, converting into Ad the direct consumption coefficient matrix A for the global input-output tables in which regions are embedded:  A RR  A = 0  0  d



0 A HH 0

 AWW 0    RR 0  , where, A =  A MW  AEW ALL  

AWM A MM AEM

AWE   A ME  AEE 

Then define Bd = (I − Ad) − 1express it in the form of a block matrix7:

6  Meng et  al. (2013) further developed the breakdown formula used by KWW (2014) regarding subregions and broke down value-added exports and DVA within a country (region) based on domestic value chains and international value chains (relations between countries in global value chains). Considering that from a country’s (region’s) point of view, the TiVA method is the same as the value-added exports calculation method in the KWW (2014) formula, Meng et al. (2013)’s breakdown method based on domestic and international value chains is consistent with this chapter’s breakdown method that starts directly with the TiVA formula. 7  The inversion of a diagonal block matrix is also a diagonal block matrix

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   



 B dRR  Bd =  0  0 

 B dWW 0    0  , where, B dRR =  B dMW  B dEW B dLL  

0 B dHH 0

B dWE   B dME  B dEE 

B dWM B dMM B dEM

43



We then arrive at the value-added exports and outflows of region through pure domestic value chains and international value chains.   TiVA r =  ∑v r B drr Y rs + ∑v r B drsY ss + ∑∑ vr B drsY st  s ≠r s ≠ r t ≠ s ,r  s ≠r    +  ∑v r Brr − B drr Y rs + ∑v r Brs − B drs Y ss + ∑∑ v r Brs − B drs Y st  s ≠r s ≠ r t ≠ s ,r  s ≠r       r rG Gt r rG GG r rG GK  +  ∑ ∑v B Y  +  ∑ v B Y + ∑ ∑ v B Y  G∈{H ,L} K ∈{H ,L}  G∈{H ,L} t ≠ r   G∈{H ,L}  K ≠G     +  ∑ v r Brr − B drr Y rG + v r Brs − B drs Y sK  ∑ ∑ s∈{E ,M ,W }, s ≠ r K ∈{H ,L}  G∈{H ,L}    v r B drsY sK  +  ∑ v r B drr Y rG + ∑ ∑ (2.3) s∈{E ,M ,W }, s ≠ r K ∈{H ,L}  G∈{H ,L} 

(

)

(

(

)

)

(

(

)

)

In formula (2.3), the first square brackets refer to the value-added outflows (VAODP) of region r through pure domestic value chains. The second and third square brackets refer to the value-added outflows (VAOIP) of region r through international value chains. The fourth and fifth square brackets refer to the value-added exports (VAOIP) of region r through international value chains. The six square brackets refer to the value-added exports (VAEDP) of region r through pure domestic value chains. No more explanation is given here as to detailed breakdowns of each part of the formula.8 8  The difference between BEE − BdEE, BEr − BdErand so on may be considered to be international feedback effects arising from domestic and international value chains. This chapter attributes the value-added exports and outflows thus caused to international value chain channels.

44  

H. NI AND J. XIA

3.3  Export Value-Added Breakdown: The WWZ Method To analyze the degree to which the Chinese regions participated in GVCs as well as their trade relations with the rest of the world, this chapter used the WWZ (2013) breakdown formula in a global inputoutput model in which regions were embedded, and further divided the WWZ breakdown items according to places of origin and destinations. To more clearly reveal a specific structure as regards the degree to which the regions participated in GVCs, this chapter, based on the WWZ (2013), further divided the subitems in VS9 (FVA) into VS (FVA) from other regions within the country and VS (FVA) from other countries of the world according to corresponding regions and countries.10

4   Empirical Analysis 4.1  TiVA of Chinese Regions11 TiVA of Chinese Regions at a Glance TiVA of Chinese Regions in International Trade Table 2.1 shows TiVA and traditional trade in gross value of Chinese regions in 2007. Take the eastern Chinese coastal region as an example. In 2007, value-added exports of the eastern Chinese coastal region amounted to USD256.604 billion (Column 8 × Column 2), accounting for 35.52% (Column 2) of the region’s GDP; the ratio of value-added exports to traditional exports, VAXR (value-added export ratio) stood at 0.60, indicating that nearly 40% of the value of the region’s exports came from foreign countries and other Chinese regions, which means quite a high degree of its participation in GVCs. The region has value-added trade surplus (or 9  So treated to keep consistent as much as possible with the VS definition of Hummels et al. (2001). WWZ (2013) defines VS as FVA plus repeated calculation items (items 13 and 16) in PDC in connection with foreign countries, that is, the sum of the 11th through 16th item in the 16-item breakdown formula. 10  See WWZ (2013) for details about the 16-item breakdown formula. 11  Trade data used in this chapter was generated through MATLAB programming based on TiVA calculation formulae. Detailed program codes and data are available on request.

15.30 39.88 26.37 59.02 59.89 21.15 11.73 19.25 35.04

22.96

35.52 33.17

21.93 17.73 18.80 25.64

Exports/ GDP, %

18.70 21.90

Value-added exports/GDP, %

18.71 19.22 19.22 18.42

17.87 18.56

16.21

20.19 19.72

Value-added imports/GDP, %

16.32 22.17 17.77 27.82

37.04 42.42

17.99

28.09 46.08

Imports/ GDP, %

3.22 −1.49 −0.42 7.22

17.65 14.61

4.83 −10.43 1.48 7.22

21.98 17.47

8.39

−12.79 −6.20

−1.50 2.18 6.75

Traditional trade surplus/GDP

Value-added trade surplus/ GDP, %

668,095.46 251,081.02 355,186.79 349,8130.21

722,421.15 515,502.89

501,751.26

300,603.18 183,488.46

GDP, million dollars

Note: (1) For China as a whole, the sum of the trade surpluses of the regions should be equal to the sum of their value-added trade surpluses. But this may not be the case between two countries, for there may be discrepancies, or even opposite directions, between bilateral trade surpluses calculated with traditional gross-value trade data and bilateral value-added trade surpluses. (2) Data used in this chapter on GDP, imports and exports are all derived from the inputoutput tables and thus differ from data in the China Statistical Yearbooks. There are also considerable discrepancies between import and export data in customs statistics and that in the input-output tables. Below is an excerpt from the Tabulation Method for Input and Output of China in 2007: In preparing the input-output tables of 2007, considering the differences between processing output calculated based on processing charges and customs import and export statistics based on commodity prices, we made adjustments to import and export data when producing import and export vectors. That is to say, import data was calculated by deducting processing and assembly imports from imports in all types of trade; export data was calculated by deducting processing and assembly exports from exports in all types of trade and adding processing charges. Import and export data for input-output tables was thus linked with gross output estimation to improve the estimating method for the 2007 input/output tables

Northeastern Beijing and Tianjin Northern Coastal Eastern Coastal Southern Coastal Central China Northwestern Southwestern China

Region

Table 2.1  Value-added international trade and traditional international trade by Chinese regions in 2007 (%, million USD)   ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

45

46  

H. NI AND J. XIA

added value net exports = value-added exports − value-added imports) at USD127.492 billion, accounting for 17.65% of its GDP—4.33 percentage points lower than the region’s traditional trade surplus’s share (21.98%) in the GDP. Traditional gross-value trade statistics overestimated the region’s trade surplus. This also demonstrates the processing trade characteristic of the eastern Chinese coastal region and the fact that other Chinese regions engage in indirect foreign trade through the region. Of the eight Chinese regions in 2007, five had a trade surplus in value-added terms, and five had a trade surplus in gross-value terms. Value-added exports and gross-value trade in 1997 and 2002 were similar to those in 2007, but most of the regions tended to have a widening difference between their value-added trade surplus and gross-value trade surplus, and traditional gross-value trade surplus statistics overestimated China’s balance of trade. There were three regions with a VAXR12 greater than 1 in 2007, namely Northwestern China (1.51), Northeastern China (1.22) and Central China (1.04), suggesting that these regions had part of their value-added exports implemented indirectly through domestic other regions. By contrast, the relatively developed Southern Coastal, Eastern Coastal, and Beijing and Tianjin regions each had a lower VAXR, all below 0.6. Exports of these regions contained the value added from domestic other regions and foreign countries; that is to say, when manufacturing export products these regions used input of intermediate products from other regions. From the perspective of TiVA by region (Table 2.2), China value-added exports, imports and value-added trade surplus came mainly from the Eastern and Southern Coastal regions, suggesting that the coastal regions are the main contributors to China’s TiVA surplus and traditional gross-­ value trade surplus. In 2007, for instance, value-added exports of the Eastern Coastal region and the Southern Coastal region accounted for 28.60% and 19.06%, respectively, of China’s gross value-added exports, 12  VAXR may be greater than 1. Circumstances where the value is greater than 1 are consistent with theoretical predictions, which however is a flaw of the indicator. WWZ (2013) noted that VAXR for a particular industry might not certainly be smaller than 1 and that a greater value than 1 for the industry was because the industry’s value added was a result not only of direct exports through the industry but of indirect exports through other industries. This chapter also found on a subregional level that VAXR for a particular region within a country might also be greater than 1, because the region’s value added came from not only direct exports through the region itself but also indirect exports by supplying intermediate products to other regions.

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

47

Table 2.2  China’s foreign trade in 2007 by region (%) Region Northeastern Beijing and Tianjin Northern Coastal Eastern Coastal Southern Coastal Central China Northwestern Southwestern China

Value-­added exports

Exports Value-added Imports Value-added imports trade surplus

Trade surplus

6.27 4.48

3.75 5.97

9.42 5.61

8.68 8.69

−1.78 1.58

−15.23 −4.51

12.84

10.80

12.62

9.27

13.41

16.67

28.60 19.06

34.78 25.19

20.03 14.85

27.49 22.47

50.50 29.83

62.90 35.68

16.34 4.96 7.45 100.00

11.53 2.40 5.58 100.00

19.40 7.49 10.59 100.00

11.20 5.72 6.49 100.00

8.52 −1.48 −0.59 100.00

12.78 −10.38 2.08 100.00

with their TiVA surpluses accounting for 50.50% and 29.83%, respectively, of the gross TiVA surplus of China as a whole. By comparison with traditional gross-value trade, the coastal regions’ TiVA (exports, imports and surplus) contributed less than tallied in traditional gross-value trade terms to China’s exports, imports and surplus. For instance, the Eastern Coastal region’s exports calculated with traditional gross-value trade data accounted for 34.78% of China’s gross exports, 6.18 percentage points higher than the region’s share of the country’s value-added exports, implying an overestimated contribution by the region in traditional gross-­ value trade statistics. Moreover, the differences between the Chinese regions in terms of value-added exports were also lower than those estimated in traditional gross-value trade statistics. For instance, the difference between the maximum and minimum percentages of value-added exports was 24.12 percentage points, while that in traditional gross-value trade was 32.38 percentage points.  alue-Added Inflows and Outflows Between Chinese Regions V According to source/destination countries (regions) of value-added, value-added output from a Chinese region and absorbed by foreign final demand may be called the region’s value-added exports, and such value-­ added output, when finally absorbed by other Chinese regions, may be called a value-added outflow. And similarly, value-added from abroad and absorbed by a Chinese region’s final demand may be called value-added

48  

H. NI AND J. XIA

import, and value-added inflow if such added value comes from other Chinese regions. Table  2.3 shows value-added inflows and outflows between Chinese regions in 2007. The inflow-outflow balance (value-­ added outflows—value-added inflows) of the Northeastern and Northwestern regions was a surplus (greater than 0), but their balance of international trade in value-added (value-added exports—value-added imports) was a deficit (smaller than 0). These two regions had the highest value-added outflows/(value-added outflows plus exports) ratios, at 72.71% and 60.57%, suggesting that they fed more value-added into the final demand of domestic other regions and provided less value-added for other countries. By contrast, the Northern, Eastern and Southern Coastal regions had an inflow-outflow balance that was a deficit, and an international balance of trade in added value that was a surplus, with lower value-­ added outflows/(value-added outflows plus exports) ratios. From the above analysis on the Chinese regions’ value-added exports and outflows, imports and inflows, it may be concluded that China’s inland regions participated in GVCs indirectly by supplying intermediate products to the coastal regions; traditional statistics on trade in gross value generally overrated the coastal regions and underrated the inland regions in terms of their contributions to China’s foreign trade; traditional gross-­ value exports also overestimated export gaps between the regions. 4.2  A Structural Breakdown of Value-Added Exports and Outflows of Chinese Regions To show more clearly the contributions—and the trends in changes thereof—of domestic value chains and international value chains to the Chinese regions’ value-added exports and outflows, this section employs the preceding breakdown formula regarding regional value-added outflows and exports through pure domestic value chains and international value chains to break down China’s regional value-added exports and outflows in 1997, 2002 and 2007.13 Figure 2.1 shows a breakdown of the Chinese regions’ value-added outflows and exports in 2007 through pure domestic value chains and international value chains. The following can be observed from the figure: (1) 13  Restricted by the length of this chapter, there are only listed the breakdown results of value-added exports and outflows of the regions without including results of their valueadded exports to foreign countries. In addition, only data for the year 2007 is included. Data for 1997 and 2002 is available on request.

52.90 27.86 37.12 47.98 72.71 54.13

12,9386

99,081

10,0950

13,5148 118,618 78,803

21,2986 62,134 97,056

10,4745

13,6430

13,2144

45,016 43,416

Value-­ added inflows

63.01 56.28 58.70

52.26

51.38

61.91

42.58 54.54

Value-added inflows/ value-added inflows

36,832 46,175

−4496 3993 33,863 127,493 75,309 21,500 −3736 −1486

41,328 42,181 −2759 −37,349 −3796 −77,838 56,484 −18,253

−56,338 52,748 −19,739

71,514

90,143

31,104

Value-­ added balance

Value-added exports − imports

Value-added outflows − inflow

Note: Regional value-added trade surplus = (Regional value-added outflows − Regional value-added inflows) + (Regional value-added exports − Regional value-added imports) = Regional net outflows of value-added + Regional net exports of value-added

60.57 68.05

86,344 85,598

Northeastern Beijing and Tianjin Northern Coastal Eastern Coastal Southern Coastal Central Northwestern Southwestern

Value-added outflows/ value-added exports

Value-­ added outflows

Region

Table 2.3  Value-added inflows and outflows by region in 2007 (million USD, %)

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

49

50  

H. NI AND J. XIA

Fig. 2.1  Domestic and international value chains for value-added exports and outflows in 2007 by Chinese regions

Regional value-added outflows to domestic other regions were mostly through pure domestic value chains, with a very small part of them through international value chains. For domestic inland regions in particular, their value-added outflows relied more on domestic value chains. Thanks to the vast domestic supply markets and the advantage of cheap prices, final products consumed in the Chinese regions were mostly manufactured within the country, with comparatively small amounts imported from other countries. (2) The Chinese regions all had a big percentage of value-added exports implemented through pure domestic value chains and international value chains, but wide differences existed in this respect between the regions. Domestic inland regions exported value-added to foreign countries mainly through international value chains, while the coastal regions did so mainly through domestic value chains. This is understandable. According to the VAEIP and VAEDP formulae, the coastal regions mainly exported final products for direct consumption abroad, but relied mostly on domestic value chains for manufacturing export products. The inland regions, on the other hand, largely supplied raw materials and semi-finished products to foreign producers, and they relied on indirect exports to produce valueadded through intermediate input relations with foreign countries. 4.3  Breakdown of Regional Gross Outflows and Exports: Based on the WWZ Method Table 2.4 shows breakdown results based on the WWZ method of gross exports and outflows of the Chinese regions in 1997, 2002 and 2007.

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

51

Table 2.4  WWZ-based gross exports and outflows by Chinese regions Year

Region

DVAR

RDVR

FVAR

PDCR

VSR

Out-EXP

2007

Northeastern Beijing and Tianjin Northern Coastal Eastern Coastal Southern Coastal Central China Northwestern Southwestern Average

0.68 0.49 0.61 0.53 0.48 0.64 0.66 0.68 0.59 DVA 0.78 0.54 0.73 0.64 0.53 0.80 0.73 0.84 0.70 DVA 0.76 0.63 0.73 0.59 0.55 0.66 0.72 0.77 0.68

0.01 0.00 0.02 0.01 0.01 0.03 0.01 0.01 0.01 RDV 0.01 0.00 0.02 0.01 0.00 0.02 0.01 0.01 0.01 RDV 0.01 0.00 0.02 0.01 0.01 0.08 0.01 0.02 0.02

0.19 0.36 0.24 0.30 0.36 0.21 0.22 0.22 0.26 FVA 0.15 0.36 0.21 0.27 0.37 0.12 0.19 0.11 0.22 FVA 0.17 0.29 0.15 0.31 0.36 0.04 0.20 0.15 0.21

0.13 0.14 0.14 0.15 0.15 0.13 0.11 0.09 0.13 PDC 0.06 0.09 0.05 0.09 0.10 0.06 0.06 0.05 0.07 PDC 0.06 0.07 0.10 0.08 0.09 0.22 0.07 0.07 0.09

0.31 0.49 0.36 0.45 0.51 0.33 0.33 0.31 0.39 VS 0.21 0.45 0.25 0.35 0.46 0.17 0.25 0.15 0.29 VS 0.22 0.36 0.24 0.38 0.44 0.25 0.26 0.21 0.30

210,333 254,776 403,110 665,611 569,184 443,409 248,777 215,665

2002 Northeastern Beijing and Tianjin Northern Coastal Eastern Coastal Southern Coastal Central China Northwestern Southwestern Average 1997 Northeastern Beijing and Tianjin Northern Coastal Eastern Coastal Southern Coastal Central China Northwestern Southwestern Average

Out-EXP 47,406 65,062 108,080 191,137 204,845 87,164 33,399 43,135 Out-EXP 29,461 29,114 53,818 114,815 100,032 28,213 21,789 21,706

Note: DVAR refers to the ratio of DVA to corresponding regional gross outflows and exports; RDVR, FVAR and PDCR refer to their ratio to regional gross outflows and exports, respectively

(1) Beijing and Tianjin, Eastern Coastal, Southern Coastal and Northern Coastal regions had a lower DVAR value than did the Northeastern, Central, Northwestern and Southwestern regions, which to a certain extent suggests higher degrees of the coastal regions’ participation in GVCs. In 2007, for instance, the Southern Coastal region had the lowest DVAR value, at only 0.48, while the Northeastern region had the highest DVAR value, at 0.68. (2) From 1997 to 2007, all the regions saw a downward trend in DVAR on the whole, indicating their increasing

52  

H. NI AND J. XIA

participation in GVCs. And this downward trend became more remarkable following China’s accession to the WTO in particular. (3) In contrast to DVAR, the regions’ FVAR changed in the opposite direction. Each of the regions saw a gentle upward trend in FVAR, whose simple arithmetical mean rose from 0.21 in 1997 to 0.26 in 2007, up 5 percentage points, and the results only slightly differed from those for China as a whole in WWZ (2013) and KWW (2014). (4) VSR showed a similar trend to FVAR because FVA is the main part of VS. (5) RDV has the smallest percentages, which averaged only 0.01 and almost remained unchanged through the years, suggesting a very small portion of domestic value-added returned to China.14 In short, the Chinese regions were all more and more involved in GVCs, with the coastal regions outperforming the inland ones in this respect. Further division can be made, according to different VS (FVA) sources for the Chinese regions, between VSD (FVAD) for domestic other regions and VSC (FVAC) for other countries of the world. Table  2.5 shows a breakdown of the Chinese regions according to VS sources. The following observations can be made. (1) After China’s accession to the WTO, All the regions saw a surge in VS/(regional gross exports plus outflows), which in 1997–2002, however, changed little and even tended to be in decline. For the Eastern Coastal region, for instance, VS/ (regional gross exports plus outflows) decreased from 0.3833 in 1997 to 0.3529  in 2002 before increasing again to 0.4503  in 2007. (2) The increase in 2002–2007 of VS/(regional gross exports plus outflows) for the Chinese regions could be ascribed to various structural elements. The growing degree of VS in the Northeastern region after China’s accession to the WTO was largely because of its increased ties with domestic other regions, or more accurately, because of domestic value chains. For the Eastern Coastal region’s VS/(regional gross exports plus outflows) rose from 0.2105 in 2002 to 0.3112 in 2007, up 0.1007, the percentage of which ratio derived from domestic other regions increasing from 0.1534 to 0.2450, up 0.8916, with a contribution rate of 90.97%(0.0916/0.1007  ×  100%). In contrast to the Northeastern region, the Eastern Coastal region’s VS ratio increase in 2002–2007 was 14  This part of domestic value added was first exported to foreign countries but was also contained in imports from other countries and finally consumed domestically, known as RDV. RDV, though not part of value added exports of a country, is part of the country’s GDP hidden in exports.

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

Table 2.5  Breakdown of Chinese regions according to VS sources Region

Source

1997

2002

2007

Northeastern

VS/out-exp Domestic other regions (VSD) Domestic first-place region

0.2246 0.1093

0.2105 0.1534

0.3112 0.2450

0.0309 (Northern Coastal) 0.0280 (Eastern Coastal)

0.0500 (Northern Coastal) 0.0465 (Central)

0.1153 0.0227

0.0417 (Northern Coastal) 0.0354 (Beijing and Tianjin) 0.0570 0.0189

0.0167

0.01109

0.0076

0.3833 0.1733

0.3529 0.1558

0.4503 0.1657

0.0473 (Southern Coastal) 0.0460 (Central)

0.0591 (Central)

0.0595 (Central)

0.2099 0.0495

0.0303 (Southern Coastal) 0.1971 0.0599

0.0590 (Southern Coastal) 0.2846 0.0759

0.0283

0.0271

0.0272

0.2104 0.1700

0.1521 0.1183

0.3098 0.2138

0.0338 Southern Coastal 0.0281 Northwestern 0.0338 0.0086

0.0683 Southern Coastal 0.0476 Central 0.0959 0.0077

0.0045

0.0033

Domestic second-­ place region

Eastern Coastal

Other countries (VSC) First-place country (United States) Second-place country (Japan) VS/out-exp Domestic other regions (VSD) Domestic first-place region Domestic second-­ place region

Southwestern

Other countries (VSC) First-place country (United States) Second-place country (Japan) VS/out-exp Domestic other regions (VSD) Domestic first-place region

0.0600 (Southern Coastal) Domestic second-­ 0.0344 place region Eastern Coastal Other countries (VSC) 0.0403 First-place country 0.0108 (United States) Second-place country 0.0081 (Japan)

Note: VS/out-exp refers VS divided by the sum of regional gross exports and outflows

0.0661 0.0169

53

54  

H. NI AND J. XIA

mostly caused by VS ratio increases in other countries, with a 89.78% rate of VS contribution from other countries. The Southwestern region’s VS ratio increase was derived from VS from domestic other regions and other countries, with a 60.61% rate of VS contribution from domestic other regions. (3) The United States and Japan were the main source countries of VS for the Chinese regions. In 2007, for instance, the Eastern Coastal region’s VS share from other countries was 0.2846, of which 0.0759 and 0.0272 came from the United States and Japan, respectively, accounting for 36.23% of combined VS from other countries of the world. (4) Regional VS from domestic other regions was largely a result of relations with neighboring regions. To the Eastern Coastal region, for instance, the biggest VS share from domestic other regions came from the neighboring Central region.

5   Conclusions This chapter expanded the TiVA and WWZ analysis methods by embedding China’s regional input-output tables in global input-output tables and employed the global input-output tables for 1997, 2002 and 2007 in which the Chinese regions were embedded to analyze the positions and roles of the Chinese regions in domestic and international value chains. The following main conclusions were drawn: (1) Trade in value-­added varied widely from region to region in China. In traditional statistics on trade in gross value, trade surpluses of regions were overestimated, and there even were phenomena that some regions had a trade surplus in gross-value terms but a trade deficit in value-added terms. (2) From the perspective of value-added outflows and exports, the Chinese inland regions also show relatively high degrees of participation in GVCs, mainly by supplying intermediate products to domestic other regions. (3) The Chinese regions participated in GVCs through different channels. The coastal regions participated in GVCs mainly by manufacturing through domestic value chains final products to be directly exported abroad, while the inland regions did so mostly by supplying intermediate products abroad through international value chains. This suggests that the Chinese regions were in different stages of development and they participated in GVCs by making use of their own comparative advantages in terms of specialization. (4) WWZ (2013) breakdown results show that the Chinese regions had a growing degree of participation in GVCs,

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

55

especially following China’s accession to the WTO. The coastal regions were more involved in GVCs than the inland regions. (5) After China joined the WTO, the Chinese regions’ VS/(regional gross exports and outflows) surged because of strengthened relations with domestic other regions (in the case of the inland regions) and increased VS from other countries (in the case of coastal regions). Based on the analysis above, the following suggestions may be made: (1) Considering the important roles that domestic value chains have in involving China into GVCs, China should take further steps to promote healthy development of domestic value chains, lower costs of transport and information transmission between regions, minimize local protectionism and push for integrated development of domestic regions. (2) Given the considerable differences that exist between the Chinese regions in both the degree and the way of their participating in GVCs, it is necessary, when making related policies, to take into consideration local conditions so that they can make full use of their own comparative advantages and actively participate in GVCs. The global input-output model used for this study, which was embedded with the Chinese regions, offers an endogenous perspective from which GVCs are examined and which helps better understand the relations of domestic value chains to international ones. It is worth noting that, constrained by data availability, this chapter made some particular assumptions (e.g. about proportional coefficient) in preparing the global input-output tables in which the Chinese regions were embedded, so prudence need be used when data contained herein is interpreted. Given that the assumptions herein were all made in breakdowns on industrial level, there may be errors in estimates on industrial level, but errors in estimates on regional level may have been very few. Therefore, this study put the focus on the involvement of the regions (country) in GVCs as well as their positions therein. Finally, further research is still required to improve the method and data used to produce global input-output tables in which Chinese regions are embedded, for example, exploring the possibility of embedding Chinese provinces in global input-output tables; apply global input-output tables in which Chinese regions are embedded to research on general equilibrium models, embodied energy, environmental pollution and so on and measure regional comparative advantages and opening-­up degrees against value-added exports and imports.

56  

H. NI AND J. XIA

References Ping, Xinqiao et al. China Center for Economic Research. Vertical Specialization in Export Trade of China and China-US Trade. The Journal of World Economy, Issue 5, 2006. Su, Qingyi. A Breakdown of Provincial-level Export Value Added in China and Its Application. Economic Research Journal, Issue 1, 2016. Zhang, Jie, Chen, Zhiyuan and Liu, Yuanchun. Estimation and Change Mechanism of Domestic Value-added in China’s Exports, Economic Research Journal, Issue 10, 2013. Lawrence, J.  L. et  al. Non-competing Input-Occupancy-Output Model and Its Application: China-US Trade Surplus in Perspective. Social Sciences in China, Issue 5, 2007. Costinot, A., Jonathan V., & Wang, S., An Elementary Theory of Global Supply Chains. Review of Economic Studies, Vol. 80, No. 1, 2013, pp. 109—144. Dedrick, J., Kraemer, K.L., & Linden, G., Who Profits from Innovation in Global Value Chains?: A Study of the IPod and Notebook PCs. Industrial and Corporate Change, Vol. 19, No. 1, 2010, pp. 81–116. Dietzenbacher, E., Joaquim M. G., & Denise I., The Role of Brazilian Regions in the Global Value Chain, Working Paper, 2014, Dietzenbacher, E., Robert Stehrer, Marcel Timmer and Gaaitzen De Vries. The Construction of World Input-Output Tables In the WIOD Project. Economics Systems Research, 2013, Vol. 25, No. 1, pp. 71–98. Findlay, R., An ‘Austrian’ Model of International Trade and Interest Rate Equalization. Journal of Political Economy, Vol. 86, No. 6, 1978,pp. 989–1007. Grossman, G.M., & Rossi-Hansberg, E., Trading Tasks: A Simple Theory of Offshoring. The American Economic Review, Vol. 98, No. 5, 2008, pp. 1978–1997. Hummels, D., Ishii, J., & Yi, Kei-Mu., The Nature and Growth of Vertical Specialization in World Trade. Journal of International Economics, Vol. 54, No. 1, 2001, pp. 75–96. Johnson, R.C., & Noguera, G., Accounting for Intermediates: Production Sharing and Trade in Value-added. Journal of International Economics, Vol. 86, No. 2, 2012, pp. 224–236. Kee H.  L., & Tang, H., Domestic Value-added in Exports: Theory and Firm Evidence from China. American Economic Review, Vol. 106, No. 6, 2015, pp. 1402–1436. Koopman R., Wang Z., & Wei, S.J., Tracing Value-added and Double Counting in Gross Exports. American Economic Review, Vol. 104, No. 2, 2014, pp. 459—494. Linden, G., Kraemer, K.L., & Dedrick, J., Who Captures Value in a Global Innovation System? The Case of Apple’s iPod. Communications of the ACM, Vol. 52, No. 3, 2009, pp. 140–144.

  ROLES OF CHINESE REGIONS IN GLOBAL VALUE CHAINS AND THEIR…   

57

Meng, B., Wang Z., & Koopman, R., How are Global Value Chains Fragmented and Extended In China’s Domestic Production Networks? IDE Discussion Paper No. 424, 2013. Meng, B., Yamano N., & Fang, Y., China’s Regional Economics and Value Chains: An Interregional Input-output Analysis, IDE Discussion Paper, No. 359, 2012. Tempest, R., Barbie and the World Economy. Los Angeles Times, September 22, 1996. Upward, R., Wang, Z., & Zheng, J., Weighing China’s Export Basket: The Domestic Content and Technology Intensity of Chinese Exports. Journal of Comparative Economics, Vol. 41, No. 2, 2013, pp. 527–543. Wang Z., Wei, S.J., & Zhu K.F., Quantifying International Production Sharing at the Bilateral and Sector Levels. Working Paper 19677. November, 2013. Xing, Y., & Detert N., How the iPhone Widens the United States Trade Deficit with the People’s Republic of China. ADBI Working Paper, 257, 2010.

CHAPTER 3

Why China’s Gini Coefficient Has Been Falling? A Study Based on Income Sources Tianyu Yang and Zhinan Cao Abstract  By decomposing the Gini coefficient and the Gini coefficient differences by income source, this chapter tries to explore the reasons why the Gini coefficient has been falling in recent years in China. The chapter’s analytical framework consists of two parts. The first is to decompose the Gini coefficient by income source, and the second is to decompose the differences between the Gini coefficient in a given year and the average value during the period observed into three components—the inequality effect, share effect, and correlation effect of each income source. The

This article is one of the milestones of “Understanding, Adapting to and Leading the New Normal of the Economy”, a research sub-project of the 2011 Collaborative Innovation Center for the Development of Socialist Economy with Chinese Characteristics sponsored by the Ministry of Education. CLC No.:F036; Document Code: A; Paper No.: 1002-8102(2016)11-0034-13 T. Yang (*) School of Economics, Renmin University of China, Beijing, China Z. Cao China Securities Depository and Clearing Company Limited, Hong Kong, China © The Author(s) 2019 D. He, C. Wang (eds.), A New Era, https://doi.org/10.1007/978-981-10-8357-0_3

59

60  

T. YANG AND Z. CAO

breakdown of Gini coefficient differences, in particular, facilitates a better estimate of how each income source contributed to the decrease in the Gini coefficient. It is found that while the decline in the wage inequality contributes the most to the decline in the Gini coefficient, the share effect of transfers is the main factor preventing the decline. Therefore, to decrease the Gini coefficient requires efforts to reduce the inequality in wages and transfers and adjust the share of transfers. Keywords  Gini coefficient • Income sources • Wages • Transfers

1   Introduction The Gini coefficient is used as a measure of the fairness of the income distribution among residents and is often used to compare the gap between rich and poor countries and regions. According to the National Bureau of Statistics, the Gini coefficient in China has remained at a high level of 0.47–0.49 since 2003, reaching a peak of 0.491 in 2008. The latest data released in 2015 show that the Gini coefficient has fallen to 0.462, as part of a declining trend after 2008. Figure 3.1 shows how the Gini coefficient changed in China from 2003 to 2015 based on the data released by the National Bureau of Statistics. A downward trend has been exhibited since 2008.

Fig. 3.1  Gini coefficient in China from 2003 to 2015. Source: National Bureau of Statistics

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

61

What caused the fall in the Gini coefficient in China? The chapter tries to answer this question. While existing academic studies on income inequality focus primarily on the reasons for the widening inequality, few of them direct their attention to the actually narrowing gap. This chapter attempts to examine the reasons for the decline in the Gini coefficient from the perspective of income sources. Income consists of wages, operational income, property income, and transfers depending on the source from which it is obtained. This is also known as the functional distribution of income, which deals with the amount of income relative to the factors of production. It is different from the size distribution of income, which uses the Gini coefficient to measure how income is distributed among individuals and households. Logically, the functional distribution of income determines the level of the size distribution, and those groups having a position of advantage in terms of functional distribution are bound to receive a greater proportion of income in size distribution. The analysis of income sources would help us identify what lies at the root of the decline in the Gini coefficient and which income source is more conducive to the decline. We can thus make a better estimate of the contribution of each income source to the decline and, based on their respective role in reducing income inequality, optimize the income structure.

2   Literature Review Most of the existing studies that examine the income gap in China from the perspective of income sources focus on the widening divide among residents. Generally, they fall into the following three categories: The first category uses the microdata of residents to calculate the Gini coefficient and then decompose it by income source. The advantage of doing so is that we can get a more accurate urban-rural mixed result, but the downside is that survey samples are only available for a limited number of years. Yao (1999) calculated the Gini coefficient of Sichuan Province based on a 1% microdata sample and decomposed it by income source, taking into account the complex income structure in China. He found that the rapid growth of non-traditional income constituted the main reason for income inequality. Li et al. (1998) used the 1988 and 1995 resident income survey data from the Chinese Academy of Social Sciences to decompose the Gini coefficient by residents’ income source, geographical area, and educational level. They found a growing disparity in

62  

T. YANG AND Z. CAO

income across the country and at multiple levels. Liu (2013) used the China Health and Nutrition Survey (CHNS) database to examine the relationship between the different income sources and the total income of rural households from 1991 to 2004, and found that the contribution of earned income to the Gini coefficient rose and then fell, while that of capital income increased. Due to the limited availability of survey data, however, these studies had difficulty tracking the long-term changes in the Gini coefficient and in income sources. In order to maintain the continuity of data, the second category of studies used macrodata to calculate and decompose income inequality indicators (including but not limited to the Gini coefficient). Gu and Wang (2008) used the urban-to-rural income ratio to measure the disparities between the two groups of people, and found that wage and operational income were the cause of the gap. Zeng and Hu (2008), and Fan and Zhang (2011) employed the relative index of inequality (RII) to measure the urban-rural income gap. The results showed that three sources of income, except for operational income, caused an increase in the degree of inequality. Bai and Chen (2013) calculated the Gini coefficients in urban and rural areas from 2000 to 2011, and found that wages and transfers were conducive to narrowing the income gap between urban and rural residents. Li and Chi (2015) were along this line of thought to measure the income gap in east, central, and west China, and came to a similar conclusion. This category of studies explore the income gap based on the urban-rural divide, rather than classifying rural data into income quintiles (five quantiles) and urban data into septiles (seven quantiles). Therefore, these studies could only examine the difference between urban and rural residents, rather than between residents in general. The third category of studies take into account the income inequality between the country’s different regions, defining data for each region as an individual sample and measuring the contribution of each type of income. For example, Huang et  al. (2003) employed the General Electric (GE) index to measure and decompose residents’ income gap from 1993 to 2001. The finding showed that transfers exacerbated rather than reduced income inequality. Qu and Du (2010) used the regional income data to calculate the Gini coefficient in China, and found that regional income disparity in China still showed an upward trend, and the income concentration caused a further increase in the Gini coefficient. The studies in this category merely reflect the regional income gaps, rather than the Gini coefficient for the entire country.

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

63

In summary, although using microdata may improve the accuracy of the results, the size of the sample was limited, making it difficult to track changes in the Gini coefficient over a longer time period and at the same time undermining the analysis of the reasons behind the fall in the Gini coefficient. Macrodata, on the other hand, could avoid these even though the estimates based on them may not be as accurate. In the existing literature, macrodata are rarely used to examine income inequality from the perspective of income sources. This chapter would make full use of the income source data that are classified into different income levels in the statistical yearbooks, to measure urban, rural, and national Gini coefficients. It would then decompose the Gini coefficient by income source to examine the effects of wages, operational income, property income, and transfers on the overall Gini coefficient. The differences between the Gini coefficient in a given year and the average value during the period observed are decomposed into the inequality effect, share effect, and correlation effect of each income source, in order to examine the impact of different income sources on the decline in the Gini coefficient. Based on this, we may be able to answer whether and how each income source has led to a decline in the Gini coefficient. The main differences between this chapter and previous studies are as follows: First, we explore the causes for the fall in the Gini coefficient in China, while previous studies focused on the widening of income gaps. Second, the Gini coefficient is measured using data from all levels of urban and rural areas. Third, we examine not only the contribution of each type of income to the Gini coefficient in a given year, but also their contribution to the difference between the Gini coefficient in a given year and the Gini coefficient averaged over the years. This would help us explain why the Gini coefficient was falling in China. Fourth, we have found out the decline in the Gini coefficient of wage is the main cause for the decline in the Gini coefficient in China.

3   Research Methods and Statistical Notes 3.1  Research Methods In this chapter, the Gini coefficient based on income sources was decomposed in the following two ways: (1) The decomposition of the Gini coefficient for a given year; and (2) the decomposition of the difference between the Gini coefficient for a given year and the Gini coefficient averaged over the years.

64  

T. YANG AND Z. CAO

1. Calculation of the Gini Coefficient. The National Bureau of Statistics has not yet published microdata on household surveys; therefore, we could not directly decompose the Gini coefficients published by the National Bureau of Statistics. Given the availability of data, we could only employ the data of rural income quintiles and urban income septiles in the yearbooks. There are two ways to use urban and rural income data in the calculation of the national Gini coefficient. One is to regroup the mixed urban-rural population and assign a weight to each group (Sundrum 1990). This weighted grouping method is used by Chen and Zhou (2002), and Tian (2012). The other is to use some kind of income distribution function to fit the income distribution among urban and rural residents in China, and then use the income distribution function to calculate the national Gini coefficient. After comparison, we chose the first approach. The reason is that although the results of the income distribution function in theory can be more accurate, there are two flaws. First, it needs large amount of data, which made the calculation of the Gini coefficient impossible in certain years. Second, the choice of the income distribution function would greatly influence the final results. For example, Cheng (2007) used a logistic function; Chotikapanich et al. (2007) employed a generalized Beta distribution function; and Wang (2012) applied the maximum entropy distribution function. The resultant national Gini coefficient varied greatly in these studies. But the weighted grouping method does not have the above defects. The weighted grouping method may underestimate the Gini coefficient because it is difficult to satisfy the assumption that the urban-rural income distributions do not overlap. Although it was not possible to adjust the data, as indicated by Wang (2013), and difficult to meet the assumption of non-overlapping groups, the national Gini coefficient results we have obtained using the weighted grouping method are consistent with the trend in official data, and the peak value also appeared in roughly the same time (see Table 3.1). This shows that the method chosen in this chapter is reliable. We start with the calculation of the urban and rural Gini coefficients. The data format of the income groupings of urban and rural residents in the China Statistical Yearbook is inconsistent. The urban residents’ income prior to 1986 and the rural residents’ income before 2001 was grouped according to certain income ranges. The urban residents’ income after 1987 was divided into seven groups of non-equal level of income. The rural residents’ income after 2001 was divided into income quintiles. In order to ensure the consistency of the data, we used the formula of Gini

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

65

coefficient based on residents’ income groups. If the income and number of people in a group are known, the Gini coefficient can be calculated (Chen 2007).



n

n −1

i =1

i =0

G = ∑Fi Li + 2∑ (1 − Vi ) − 1

(3.1)

where F represents the proportion of each group in the total population, L represents the proportion of each group’s income, V represents the sum of L for each group, that is, V = L1 + L2 + … + Ln. After calculating the urban and rural Gini coefficients, we used the weighted grouping method to obtain the national Gini coefficient: G = p12

u1 u u −u G1 + p22 2 G2 + p1 p2 1 2 u u u

(3.2)

where P1 and P2 are the proportions of urban and rural populations, respectively; u1, u2, and u are the urban, rural, and national residents’ per capita income. G1 and G2 are the urban and rural Gini coefficients, respectively. 2. Decomposition of Gini Coefficient in a Given Year. Based on the national Gini coefficient calculated in this chapter, we decomposed the Gini coefficient by income source. Suppose i is the income source (i = 1, 2,…I), j is the individual (j = 1, 2, …N), and assuming that Xij is the i-th source of the j-th resident’s income, then



N

I

j =1

i =1

I

N

Xi = ∑Xij , X j = ∑Xij , X = ∑∑X ij i =1 j =1

(3.3)

Xi, Xj, and X are, respectively, the sum of every resident’s income from the i-th source, the sum of the j-th resident’s income from all sources, and the total income. Take a step further to obtain the proportions of income:

= Sij X= Xi / X , S j = X j / X ij / X , Si



(3.4)

Si, Sj, and Sij are, respectively, the proportion of the i-th income source in the total income, the proportion of the j-th resident’s income in the total income, and the proportion of the j-th resident’s i-th income source in the total income.

66  

T. YANG AND Z. CAO

Drawing on what Silber (1989) has done in decomposing income inequality, the Gini coefficient can be expressed as follows1:

I G = [ e′] G [ S ]



(3.5)

where IG is the Gini coefficient, e is the row vector consisting of the population’s proportion, G is an n×n matrix, all the diagonal elements are 0, all the matrix elements above the diagonal are −1, and those below the diagonal are 1. S is the column vector composed of the proportion of the income of each resident, sorted by the order of the total income of each resident from low to high. If each resident’s income is shown as the sum of his or her income from difference sources, the Gini coefficient can be written as follows:



 I  I G = [ e′] G ∑  Sij    i =1 

(3.6)

Let Vij = Xij/Xi, representing the j-th resident’s i-th income source as a proportion of all residents’ income from the source. Then Sij = Xij/X can be transformed into Sij = SiVij. Note that Sij is sorted according to the level of total income X, so is Vij that corresponds to it. Given the equation Vij = Xij/Xi and (3.6), the Gini coefficient can be expressed in the following form: I



I G = ∑ Si PGi i =1

(3.7)

where PGi  =  [e′]G [Vij]. PGi is also known as “pseudo Gini coefficient” (Shorrocks 1982), which mimics the total-income Gini coefficient. But the difference is that the pseudo Gini coefficient measures the inequality in income derived from a source, based on the sorting by total income X, rather than by income Xi from this source. For example, although high-­income 1  Fei et al. (1978) and Lerman and Yitzhaki (1985) also proposed a method of decomposing the Gini coefficient based on income sources. But as pointed out by Flückiger and Silber (1995), their methods are not only complex in computational procedures, but also have difficulty in decomposing the differences in Gini coefficients between different years and groups. Therefore, in this chapter, we drew on the experience of Silber (1989) and Flückiger and Silber (1995) in decomposing the Gini coefficient.

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

67

earners earn more than low-income earners, their income derived from a certain source may be less than the corresponding part of the latter’s income. Therefore, PGi is not the true Gini coefficient of the income from each source. The Gini coefficient proper should be expressed as AGi = [e′]G [yij], where yij = Xij/Xi. Although yij and Vij have the same expression, the former is sorted by income Xi from a source, representing the true gap between the rich and the poor in this income source group. Let Δi = PGi − AGi, denoting the difference between the pseudo Gini coefficient and the true Gini coefficient. We may thus obtain the total-­income Gini coefficient: I

I

i =1

i =1

I G = ∑ Si AGi + ∑ Si ∆ i



(3.8)

The Gini coefficient thus can be decomposed into two parts. One is the weighted sum of the Gini coefficients of different income source groups, and the other is the weighted sum of the difference between the pseudo Gini coefficient and the true Gini coefficient. The difference can be deemed as the correlation between the income sorted by the income of each source group and the income sorted by the total income. 3. Decomposition of Gini Coefficient Differences. While we may, using the above formulas, decompose the Gini coefficient for a given year by income source, we are more interested to know how each income source contributes to the Gini coefficient differences, as this will help us explain the fall in the Gini coefficient in China. For this reason, we calculated the difference between the Gini coefficient of the t year IG,t and the annual average value of the index IG,T. Drawing on the approach of Flückiger and Silber (1995), we decomposed the Gini coefficient difference into the following three parts2: I

Si ,t + Si ,T

( AGi,t − AGi,T ) 2 I S + Si ,T PGi ,t + PGi ,T +∑ Si ,t − Si ,T ) + ∑ i ,t ( ( ∆ i , t − ∆ i ,T ) 2 2 i =1 i =1

I G , t − I G ,T = ∑ i =1 I



(3.9)

2  Flückiger and Silber (1995) used this decomposition method to measure the impact of income sources on the rising Gini coefficient in Switzerland, but had never applied it to the analysis of a decline in the index in a country.

68  

T. YANG AND Z. CAO

Here, the factors that caused the Gini coefficient difference are subdivided into i sources of income, and each income source can be decomposed into three terms. The first term on the right side of Eq. (3.9) is the difference in the total-income Gini coefficient caused by the difference in the true Gini coefficient of each functional income. We called the term “inequality effect”. The second term on the right side is the value of the total-income Gini coefficient caused by the difference in the proportion of each functional income. We called the term “share effect”. The third term is the difference in the total-income Gini coefficient caused by the difference between the pseudo Gini coefficient and the true Gini coefficient. We called the term “correlation effect”. In the end, the difference in the total-­ income Gini coefficient can be decomposed into the inequality effect, the share effect, and the correlation effect of each income source. This method can be used to accurately divide the total-income Gini coefficient into the ascending and descending segments, and to decompose the rise (or fall) of the Gini coefficient of each year in each segment into the three effects of each income source. These effects explain the impact of each income source on the total-income Gini coefficient. It can be seen that this method is more accurate than simply decomposing the Gini coefficient. In addition, the empirical analysis later is based on Eq. (3.9). 3.2  Data Description According to the statistical standards of the National Bureau of Statistics, residents’ income can be divided into the following four categories: wages (or salaries), that is, what a worker received for his work or service; operational income, that is, income earned by an individual or a family by engaging in production and business activities; property income, that is, income derived from assets such as bank deposits, securities, and immovable property owned by the family, including interests, rents, and so on; transfers, including various transfer payments made by the state, organizations, or employers to a household or income transfers between households, such as pensions, housing funds, and donation, and so on. To this end, we need to collect the data on the per capita income of urban and rural areas and on the per capita functional income mentioned above. Unless otherwise indicated, the data come from the China Statistical Yearbook. The income source data of all strata of urban and rural residents are from the Survey Yearbook of Prices and Income and Expenditure of Urban and Rural Households in China (1996–2006) and the Yearbook of Urban Life and Prices in China

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

69

(2006–2012). As the officially released data on the rural population’s income sources are for years after 2002, and the Yearbook of Urban Life and Prices in China has, starting from 2011, ceased to publish the income source data of urban residents, our sample data cover the period from 2002 to 2011. Although this may not be deemed long, the peak values of the Gini coefficient and its subsequent trend of decline all occurred during the period. In general, it would not compromise the integrity of this study.

4   Measurement and Decomposition of the Gini Coefficient 4.1  Measurement and Decomposition of the Gini Coefficient for a Single Year According to the above methods, we first used Eq. (3.1) to calculate the urban Gini coefficient and the rural Gini coefficient, respectively, and then used Eq. (3.2) to merge. Finally, we got the Gini coefficient of national total income and the Gini coefficient of each income source from 2002 to 2011. Then, we decomposed them again with Eq. (3.8) to obtain the contribution of each income source to the Gini coefficient. The results are shown in Table 3.1. The Gini coefficient of the national total income was maintained at the range of 0.43–0.46, and in 2008 and 2009, reached the peak of about 0.453, and then showed a downward trend. The Gini coefficient shown in Table 3.1 is slightly lower than the official data as compared to the official Gini coefficient shown in Fig. 3.1. This is because the official data were based on microdata from household surveys, and all the total income data were sorted without distinguishing between urban and rural areas, resulting in a more accurate Gini coefficient. The demographic and other subgroups used in this chapter inevitably concealed the income gap within the group, in addition to the problem of overlap between urban and rural income distributions; therefore, the findings might be inconsistent with the official data. But overall, the Gini coefficient measured in Table 3.1 is consistent with the basic trend of the official Gini coefficient, and the peak appeared roughly at the same time, ensuring that the Gini coefficient measured in this chapter is credible. Table 3.1 shows that the Gini coefficient has indeed declined in recent years, which can be explained from the perspective of income sources. As shown in Table  3.1, wage inequality contributed the most to the Gini coefficient for national total income. The average contribution from

0.4376 0.4493 0.4486 0.4520 0.4530 0.4527 0.4531 0.4532 0.4429 0.4360

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

0.5658 0.5695 0.5725 0.5606 0.5482 0.5415 0.5336 0.5239 0.5075 0.4927

Gini coefficient 58.4% 60.3% 61.0% 61.4% 62.5% 62.5% 63.0% 62.5% 61.8% 61.6%

Contribution rate

National wage

0.4255 0.4154 0.4045 0.3768 0.3565 0.3543 0.2989 0.2955 0.2940 0.2739

Gini coefficient 20.3% 18.7% 18.1% 16.4% 15.2% 14.5% 12.8% 12.1% 12.3% 12.6%

Contribution rate

National operational income

Source: The authors’ computation based on collected data. The same as below

Gini coefficient of national total income

Year

Table 3.1  National Gini coefficient and shares of income sources

0.4224 0.4307 0.4319 0.4383 0.4788 0.5064 0.4962 0.4976 0.4939 0.5132

Gini coefficient 1.2% 1.5% 1.6% 1.7% 2.1% 2.6% 2.6% 2.7% 2.9% 3.3%

Contribution rate

National property income

0.6845 0.6755 0.6615 0.6449 0.6287 0.6166 0.5893 0.5822 0.5721 0.5545

Gini coefficient

20.1% 19.5% 19.3% 20.4% 20.3% 20.5% 21.6% 22.7% 22.9% 22.5%

Contribution rate

National transfer

70   T. YANG AND Z. CAO

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

71

2002 to 2011 was around 62%, and the Gini coefficient of wage had been slowly declining since 2004. It can be inferred that the decline in wage inequality was the main factor to alleviate the income inequality of residents. The Gini coefficient and the contribution rate of operational income had been declining throughout the sample period. This shows that the decline in operational income inequality also helped alleviate residents’ income inequality, though to a less extent due its lower share. Unlike the operational income and any other type of income, the Gini coefficient and the contribution rate of property income showed an upward trend throughout the sample period. This shows that property income inequality had become a major source of worsening income disparities. However, the contribution rate of property income inequality to the Gini coefficient was low (only about 3%). Therefore, it was not enough to reverse the decline in the Gini coefficient. The transfers showed a downward trend in the Gini coefficient and an increase in contribution rate. This trend shows that the decline in the inequality of transfer also helped alleviate residents’ income inequality, and the contribution rate of this factor was rising. This means a further decline in the inequality in transfers will help improve the national Gini coefficient. 4.2  Decomposition of the Difference in Gini Coefficient The above is the decomposition of the Gini coefficient for a single year. Although this decomposition can help us to estimate the contributions of income sources to the decline in the Gini coefficient, it cannot accurately measure the effect of each income source. In particular, from 2002 to 2011, the Gini coefficient in China first ascended and then descended. We can hardly distinguish which income sources affect the rise and decline in the national Gini coefficient by simply decomposing the Gini coefficient for a single year. Below we use Eq. (3.9) to decompose the difference between the Gini coefficient for a given year and the annual average value of the index to the inequality effect, the share effect, and the correlation effect of each income source, so that we can better understand the extent to which the Gini coefficient had been reduced by each income source. Tables 3.2, 3.3, 3.4, and 3.5 show the contribution of the inequality effect, the share effect, and the correlation effect of each income source to the Gini coefficient difference of the total income. The “total effect” column in each table indicates to which extent a certain income source led to the difference in the total-income Gini coefficient

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

0.00468 0.01489 0.01688 0.01086 0.00910 0.00679 −0.00345 −0.00886 −0.01945 −0.03166

Total effect

13.4% 42.7% 52.4% 57.4% 128.9% 206.1% 19.6% 40.1% 57.3% 60.9%

Contribution rate 0.01388 0.01615 0.01791 0.01102 0.00385 −0.00003 −0.00463 −0.01027 −0.01977 −0.02823

Inequality effect 39.7% 46.3% 55.6% 58.2% 54.6% −1.0% 26.2% 46.5% 58.3% 54.3%

Contribution rate

Table 3.2  Impact of wage on Gini coefficient difference Contribution rate −26.3% −3.6% −3.2% −0.8% 74.3% 207.2% −6.7% −6.4% −0.9% 6.6%

Share effect −0.00920 −0.00125 −0.00104 −0.00016 0.00525 0.00682 0.00118 0.00141 0.00032 −0.00343

0 0 0 0 0 0 0 0 0 0

Correlation effect

0% 0% 0% 0% 0% 0% 0% 0% 0% 0%

Contribution rate

72   T. YANG AND Z. CAO

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

0.03242 0.02404 0.02033 0.00913 0.00052 −0.00351 −0.01480 −0.01892 −0.01926 −0.02034

Total effect

92.6% 69.0% 63.1% 48.2% 7.4% −106.5% 83.8% 85.7% 56.8% 39.1%

Contribution rate 0.01834 0.01544 0.01282 0.00619 0.00154 0.00103 −0.01099 −0.01143 −0.01173 −0.01640

Inequality effect 52.4% 44.3% 39.8% 32.7% 21.9% 31.4% 62.2% 51.8% 34.6% 31.5%

Contribution rate 0.01425 0.00859 0.00751 0.00291 −0.00102 −0.00456 −0.00384 −0.00751 −0.00756 −0.00397

Share effect

Table 3.3  Impact of operational income on Gini coefficient difference

40.7% 24.7% 23.3% 15.4% −14.5% −138.6% 21.7% 34.0% 22.3% 7.6%

Contribution rate

Contribution rate −0.5% 0.0% 0.0% 0.1% 0.1% 0.7% −0.1% −0.1% −0.1% 0.0%

Correlation effect −0.000164 0.000006 −0.000005 0.000025 0.000005 0.000024 0.000024 0.000023 0.000023 0.000024

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

73

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

0.00225 −0.00128 −0.00253 0.00081 −0.00233 −0.00218 −0.00118 0.00360 0.00191 −0.00417

Total effect

6.4% −3.7% −7.9% 4.3% −33.1% −66.3% 6.7% −16.3% −5.6% 8.0%

Contribution rate 0.01059 0.00900 0.00673 0.00408 0.00131 −0.00076 −0.00559 −0.00706 −0.00891 −0.01194

Inequality effect 30.3% 25.8% 20.9% 21.6% 18.6% −23.1% 31.7% 32.0% 26.3% 23.0%

Contribution rate

Table 3.4  Impact of transfers on Gini coefficient difference Contribution rate −23.8% −29.5% −28.7% −17.3% −51.7% −43.3% −25.0% −48.3% −31.9% −14.9%

Share effect −0.00834 −0.01028 −0.00925 −0.00327 −0.00365 −0.00143 0.00441 0.01066 0.01082 0.00778

0 0 0 0 0 0 0 0 0 0

Correlation effect

0% 0% 0% 0% 0% 0% 0% 0% 0% 0%

Contribution rate

74   T. YANG AND Z. CAO

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Contribution rate

−12.4% −8.0% −7.7% −9.9% −3.3% 66.7% −10.0% −9.5% −8.5% −8.0%

Total effect

−0.00435 −0.00280 −0.00248 −0.00187 −0.00023 0.00220 0.00178 0.00209 0.00289 0.00415

−0.00095 −0.00085 −0.00084 −0.00072 0.00018 0.00088 0.00062 0.00066 0.00059 0.00112

Inequality effect −2.7% −2.4% −2.6% −3.8% 2.6% 26.6% −3.5% −3.0% −1.7% −2.1%

Contribution rate −0.00340 −0.00195 −0.00164 −0.00115 −0.00041 0.00132 0.00116 0.00143 0.00230 0.00304

Share effect

Table 3.5  Impact of property income on Gini coefficient difference

−9.7% −5.6% −5.1% −6.1% −5.8% 40.1% −6.5% −6.5% −6.8% −5.8%

Contribution rate

0 0 0 0 0 0 0 0 0 0

Correlation effect

0% 0% 0% 0% 0% 0% 0% 0% 0% 0%

Contribution rate

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

75

76  

T. YANG AND Z. CAO

and the contribution rate of the source. The other columns indicate the extent of change caused by the three components (inequality effect, share effect, and correlation effect) and their contribution rate. The breakdown in each table shows that the main factors that influenced the Gini coefficient difference were the inequality effect and the share effect, but the contribution rate of the correlation effect was insignificant. Table 3.2 shows that wage led to a change in the Gini coefficient difference of the total income, which remained positive before 2007 and negative after 2007. This suggests that wage played a role in the rise in the total-income Gini coefficient before 2007 and in its decline after 2007. Therefore, the turning point was in 2007. Two important features are observed. First, the inequality effect and the total effect of wage moved in the same direction in most of the years, while the share effect and the total effect were opposite in most cases. This shows that the role of wage in reducing the total-income Gini coefficient after 2007 was mainly attributed to the inequality effect. According to the breakdown in Table 3.2, the inequality effect of wage has been negative since 2007 (i.e., driving the decline in the total-income Gini coefficient). The share effect of wage was positive from 2006 to 2010 (hindering the decline in the total-income Gini coefficient).3 While the inequality effect and the share effect of wage worked oppositely on the total-income Gini coefficient, overall, wage helped reduce the total-income Gini coefficient from 2008 to 2011. This suggests that the share effect of wage was offset by the inequality effect of wage. In other words, the decline in the inequality effect of wage, rather than the change in the share effect of wage, drove the decline in the total-­ income Gini coefficient. Second, the decline in wage inequality has played an increasingly important role in reducing the total-income Gini coefficient, with the contribution rate increasing from −1.5% in 2007 to 54.3% in 2011. This shows that the narrowing of the wage gaps had become an important reason for the decline in the total-income Gini coefficient. As can be seen from Table 3.3, operational income worsened the total-­ income Gini coefficient by 2006, but improved it after 2006. The inequality effect and the share effect of operational income changed in basically the same direction as the total effect. It is noteworthy that the contribution rate of the inequality effect after 2006 was much higher than that of the share effect. This shows that after 2006, the role of operational income 3  The impact of the share of wage income has improved in recent years. According to Table 3.2, the share effect of wage income had shifted from positive to negative in 2011.

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

77

in reducing the total-income Gini coefficient was mainly achieved by the decline in operational income inequality. While previous studies identify the share effect of operational income as the main factor in the curbing of income inequality (Fan and Zhang 2011), our study shows that the inequality effect of operational income is the main driving force. However, our results are not contradictory to those of Fan and Zhang (2011) because they used data from 2000 to 2008. The results in Table 3.3 show that in most of the years prior to 2008, the contribution rate of the share effect to the increase of the total-income Gini coefficient was much lower than that of the inequality effect, which is consistent with that of Fan and Zhang (2011). It is found that since the beginning of 2008, the impact of operational income on the total-income Gini coefficient has undergone some changes. Among them, the main driving force to curb the expansion of income inequality has shifted from the share effect to the inequality effect. The role of operational income in reducing the total-income Gini coefficient, whether it is the inequality effect or the share effect, shows a falling trend. Since 2009, the contribution of the two to a decrease in the total-­ income Gini coefficient has been lower than that of the inequality effect of wage. This means that wage has replaced operational income as the main force driving the decline in the total-income Gini coefficient. This confirms the results in Table 3.1 that the decline in the wage inequality was a major driver of the decline in the Gini coefficient nationwide. Table 3.4 shows that the effect of transfers on the total-income Gini coefficient was positive at times, and negative at other times, seemingly following no pattern at all. However, if the total effect is decomposed into the inequality effect and the share effect, the following characteristics can be seen. First, the inequality effect of transfers, which had contributed to the rise in the total-income Gini coefficient by 2007, caused it to fall after 2007. Second, the share effect of transfers, which had inhibited the increase in the total-income Gini coefficient by 2008, subsequently hindered the decrease in it. The two effects were opposite, and their contribution rates were close. These caused the total effect of transfers to fluctuate between positive and negative. This result shows that the shrinking inequality in transfers since 2007 has been the force driving the decline in the total-income Gini coefficient. However, marked inequality still exists in transfers, and thus an increase in the share effect of transfers would hinder the decline in the total-income Gini coefficient. In order for transfers to play an even bigger role in narrowing the income gaps, we

78  

T. YANG AND Z. CAO

need to further reduce transfer inequality, especially to narrow the social security gap between urban and rural areas and between different strata of society. Since 2009, both the inequality and share effects of transfers have declined. While the latter implies a reduced inequality in the income source, the decline in the inequality effect indicates that the decline in transfer inequality has slowed down. Further efforts are needed to reduce the level of transfer inequality. Table 3.5 shows a relatively insignificant impact of property income, whether in term of total effect, inequality effect, or share effect, and the income source’s contribution to the total-income Gini coefficient was small. Nonetheless, since the data in Table 3.1 show that the Gini coefficient of property income has been on the rise, we still need to examine the role of the income source in exacerbating the total-income Gini coefficient, which has commenced since 2008 according to Table  3.5. This explains two things. First, the Gini coefficient of property income was also rising before 2008, which, however, did not cause an increase in the total-­ income Gini coefficient due to the low share of the income source. Second, while the inequality in other income sources has generally been reduced after 2008, property income inequality has increased. Therefore, despite its insignificant share, its effect on the deterioration in the total-income Gini coefficient has become more pronounced. Based on the breakdown above, both the inequality effect and the share effect of property income have worsened the total-income Gini coefficient since 2008, with the impact of the share being greater. With the growth of household income, the level and share of personal investment and other forms of property income will continue to grow, which may magnify the role of property income in raising the total-income Gini coefficient. By comparing the effects of income sources on Gini coefficient difference, we have observed the following features: First, wage and operational income have helped reduce income inequality in recent years, with the former playing a bigger role. Transfers in some years reduced income inequality between residents, but property income has intensified the income inequality in recent years. We should thus, based on the different roles of the income sources, mitigate various income inequalities and optimize the income structure, so as to reduce the total-income Gini coefficient.

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

79

Second, a breakdown of the effect of each income source shows that the shrinking inequality in wages, operational income, and transfers, as well as the changes in the share of operational income in recent years, have played an important role in reducing income inequality. And the share effect of wages, transfers, and property income, in addition to the inequality effect of property income, has hindered the decline of income inequality in recent years. Third, by comparing the contribution of each effect of each income source with the difference in the total-income Gini coefficient, we found that in recent years, the inequality effect of wages has the greatest contribution to reducing the Gini coefficient difference, and the share effect of transfers has the greatest contribution to widening the difference. This shows that the decline in the Gini coefficient in recent years has been mainly due to the decline in wage inequality, while the increase in the share of transfers has been the main obstacle to the decline in the Gini coefficient. We need to narrow the wage inequality to further reduce the Gini coefficient. In addition, by reducing the inequality in transfers and adjusting the proportion of the income source, we can alleviate the negative effects of the increase in the share of transfers. Fourth, with the turning point occurring around 2008, the total-income Gini coefficient rose and fell for quite different reasons. In terms of the total effect of each income source, wage and operational income were the main reasons that reinforced income inequality during the rise of the total-income Gini coefficient. However, during the fall of the total-­ income Gini coefficient, they played an important role in reducing income inequality. Given their different roles, some of the conclusions in the literature seem to be untenable. For example, Zeng and Hu (2008), as well as Fan and Zhang (2011), based on the data before 2008, found that wage significantly widened income inequality, which we have found was not the case after 2008. Furthermore, based on the breakdown of the effect of each income source, the shrinking inequality in wages, operational income, and transfers, as well as the changes in the share of operational income, worsened income inequality during the rise of the total-income Gini coefficient, but reduced income inequality during the fall of the total-income Gini coefficient. On the contrary, the share effect of wages, transfers, and property income, as well as the inequality effect of property income, reduced the income inequality during the rise of the total-income Gini coefficient, but they worsened

80  

T. YANG AND Z. CAO

income inequality during the fall of the total-income Gini coefficient. This means that we cannot use the pre-2008 data to determine what helps narrow the income gap and what worsens it, but rather we should redefine targeted measures in the light of changing situations.

5   Conclusion In this chapter, we explore why the Gini coefficient has been falling in China in recent years based on the breakdown of income sources and of the effects of each income source (i.e., inequality effect, share effect, and correlation effect). The purpose of this chapter is to study how and to which extent each income source affects the Gini coefficient, and ultimately to boost the role of certain sources in reducing income inequality while controlling the role of others in exacerbating income inequality. The estimates in this chapter are different from those of previous studies on the decomposition of the national Gini coefficient (such as Li et al. 1998; Yao 1999; Liu 2013) or income inequality (such as Huang et al. 2003; Zeng and Hu 2008; Fan and Zhang 2011). This might be attributed to two causes, in addition to the difference in indicators and in the period selected. First, the purpose of this chapter is to explain a decline in income inequality (as expressed in the Gini coefficient), rather than an increase in it as in previous literature. Second, while previous studies mainly involved the breakdown of the Gini coefficient in a given year, this chapter decomposes the difference between it and the average Gini coefficient during the period observed, which would result in a more accurate estimate of how each income source contributes to the increase or decrease of the Gini coefficient. Reduced wage inequality is found to contribute the most to the decrease in the total-income Gini coefficient, while the rising share of transfers is a major impediment to the decrease. Consequently, the policy implications are as follows: First, wage inequality can be further reduced by raising the minimum wage, reducing taxes on low-income earners, and increasing employment. Second, the elimination of social security differences between urban and rural areas, regions, and occupations and the establishment of a national social security system will not only reduce the inequality in transfers, but also cause the effect of the rising share of transfers on the national Gini coefficient to shift from negative to positive. Third, we should properly control the share of transfers in the total income as the rising share of the income source militates against the decline in the Gini coefficient.

  WHY CHINA’S GINI COEFFICIENT HAS BEEN FALLING?   

81

In addition, the inequality effect and the share effect of operational income may facilitate the decrease of the national Gini coefficient. We should increase the proportions of the income source while at the same time decreasing the inequality. Since the inequality effect and the share effect of property income have worsened the national Gini coefficient, we should reduce the inequality and control its proportions. Because the reduced inequality in transfers has also contributed to the decline in the Gini coefficient, we should further reduce its degree of inequality. The share effect of wages has hindered the decline in the national Gini coefficient, but the effect has disappeared since 2011. We should continue to observe the effect in order to develop targeted policy. There are yet areas for improvements in this chapter. Due to limited availability of data, this chapter only studied the influence of income sources on the income inequality among different strata of urban and rural residents from 2002 to 2011. The data basis of this chapter is the macrodata published by the National Bureau of Statistics, rather than the household survey data that the National Bureau of Statistics used to calculate the Gini coefficient. With data availability increased over time, our study of the decline in China’s Gini coefficient will improve as well.

References Bai, Suxia & Chen, Jingan, On Income Gap between Urban and Rural China from the Perspective of Income Resources. Social Science Research, No. 1, 2013. Chen, Changbing, Gini Coefficient Calculation and Non-parametric Econometric Model Analysis of Residents’ Income in Various Regions. The Journal of Quantitative &Technical Economics, No. 1, 2007. Chen, Zongsheng & Zhou, Yunbo, On Income Distribution in Reform and Development. Economic Science Press, 2002. Cheng, Yonghong, The Evolution of National Gini Coefficient and Urban-Rural Decomposition since the Reform. Social Sciences in China, No. 4, 2007. Fan, Congliang & Zhang, Zhongjin, Income Inequality Effects and Income Structure Optimization Itemized. Journal of Financial Research, No. 1, 2011. Gu, Haibing & Wang, Yahong, Decomposition of Income Gap between Rural and Urban China Residents: 1985–2007. Economist, No. 6, 2008. Huang, Zuhui, Wang, Min & Wan, Guanghua, Income Inequality of Residents in China: From the Perspective of Transfer. Management World, No. 3, 2003. Li, Shi, Zhao, Renwei & Zhang, Ping, China’s Economic Transformation and Changes in Income Distribution. Economic Research Journal, No. 4, 1998. Liu, Xuke, Impact of Income Sources on Income distribution among Rural Households in China. Journal of Graduate School of Chinese Academy of Social Sciences, No. 6, 2013.

82  

T. YANG AND Z. CAO

Li, Qiyun & Chi, Cheng, Macro-Decomposition and Regional Disparities of Urban-Rural Income Gap—from the Perspective of Income Sources. Reform of Economic System, No. 6, 2015. Qu, Xiaobo & Du, Yang, Income Gap and Compositional Changes among Rural Areas: 1995–2008. Economic Theory and Business Management, No. 7, 2010. Tian, Weimin, Gini Coefficient Calculation of Provincial Residents’ Income and Analysis of Changing Trends. Economic Science Press, No. 2, 2012. Wang, Yafeng, Income Distribution Estimation of the Residents in Urban and Rural China, 1985–2009. The Journal of Quantitative & Technical Economics, No. 6, 2012. Wang, Pingpin, On Problems with Gini Coefficient Calculation of Residents’ Income. China Information News, 1st edition, 5 February2013. Zeng, Guoan & Hu, Jingjing, Causes of Formation and Expansion of Urban-­ Rural Residents’ Income Gap Since 2000: An Analysis from the Perspective of Income Sources. Finance & Trade Economics, No. 4, 2008. Chotikapanich, D., Rao, P. & Tang, K., Estimating Income Inequality In China Using Grouped Data and the Generalized Beta Distribution. Review of Income and Wealth, Vol. 53, No. 1, 2007, pp. 127–147. Fei, J. H., Ranis, G., Shirley, W. & Kuo, Y., Growth and the Family Distribution of Income by Factor Components. Quarterly Journal of Economics, Vol. 92, No. 1, 1978, pp. 17–53. Flückiger, Y. & Silber, J., Income Inequality Decomposition by Income Source and the Breakdown of Inequality Differences between Two Population Subgroups. Swiss Journal of Economics & Statistics, Vol. 131, 1995, pp. 599–615. Lerman, R. I. & Yitzhaki, S., Income Inequality Effects by Income Source: A New Approach and Applications to the United States. Review of Economics & Statistics, Vol. 67, No. 1, 1985, pp.151–156. Shorrocks, A.  F., Inequality Decomposition by Factor Components. Vol. 50, 1982, pp. 193–211. Silber, J., Factor Components, Population Subgroups and the Computation of the Gini Index of Inequality. Review of Economics & Statistics, Vol. 71, No. 1, 1989, pp. 107–115. Sundrum, R. M., Income Distribution in Less Development Countries. London and New York: Routledge, 1990, p. 50. Yao, S., On the Decomposition of Gini Coefficients by Population Class and Income Source: A Spreadsheet Approach and Application. Applied Economics, Vol. 31, No. 10, 1999, pp. 1249–1264.

CHAPTER 4

China’s Leverage Ratio and Systemic Financial Risk Prevention Jiantang Ma, Xiaojun Dong, Hongxiu Shi, Jie Xu, and Xiaofang Ma

Abstract  The leverage ratio of non-financial businesses and the government sector in China has been on the rise since the outbreak of the global financial crisis. China’s leverage ratios are not high compared with those in developed countries but substantial implicit liabilities and fast growth of debt indicate considerable potential risks. The rapidly rising leverage ratio is closely connected with such factors as overreliance on indirect financing, low efficiency in the use of funds, large amounts of funds occupied ineffectively because of overcapacity and overuse of monetary and credit policies as instruments. Now the top priority for financial regulators is to resolve the debt issue and prevent systemic financial risks. Keywords  Leverage ratio • Implicit liability • Non-financial business • Systemic financial risk

CLC: F83; Document code: A; Article ID: 1002—8102(2016)01—0005—17 J. Ma (*) Chinese Academy of Governance, Beijing, China X. Dong • H. Shi • J. Xu • X. Ma Economics Department, Chinese Academy of Governance, Beijing, China © The Author(s) 2019 D. He, C. Wang (eds.), A New Era, https://doi.org/10.1007/978-981-10-8357-0_4

83

84  

J. MA ET AL.

Since 2008, the debt-to-GDP ratio of China’s non-financial businesses and government sector has been on the rise. Although China’s leverage ratio is not high compared with that in developed countries, substantial implicit liabilities and fast growth of debt indicate considerable potential risks. Currently, the top priority for Chinese financial regulators is to gradually resolve the debt issue and defuse systemic financial risks without compromising the steady growth of the economy in the long run.

1   China’s Leverage Ratio: Reality and Estimations 1.1  Definition and Scope of Leverage Ratio Leverage ratio requirement was introduced to the international regulatory framework for banks in the wake of the global financial crisis. Typically, the leverage ratio is calculated by dividing Tier 1 capital by a commercial bank’s average total consolidated assets and certain off-balance sheet exposures. In economic studies, leverage ratio is analyzed from two perspectives. On the microeconomic level, the leverage ratio is the relationship between an economic entity’s assets and liabilities. The total amount and structure of liabilities indicate how much debt a business needs to pay back at present and in the future as well as the urgency and pressure of debt repayment. The total amount and structure of assets indicate how much economic resources a business possesses at a time as well as how the resources are distributed and its earnings power. By looking at both liabilities and assets, one can assess a company’s performance, financial flexibility and security, as well as solvency and operational stability. Statistically, the leverage ratio can be measured by the ratio of liabilities to shareholders’ equity, the ratio of assets to shareholders’ equity, the assets-to-liabilities ratio or their reciprocals. On the macroeconomic level, the leverage ratio can be calculated by dividing a country’s total liabilities by total assets or total debt by GDP. In this chapter, the leverage ratio is the debt-to-GDP ratio. In estimation of the total leverage ratio, total debt is the sum of debts of household, business, government and financial sectors. The optimal level of leverage ratio is yet to be defined. Ceccheti et  al. (2011) studied the fund flow statements of 18 OECD member countries for 1980–2010 and concluded that the threshold level of debt was 85% in the government and household sectors and 90% in the business sector.

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

85

1.2  Leverage Ratios by Sector and Total Leverage Ratio The data for 1996–2014 are used in the estimation of leverage ratios in China by sector. For the period 2013–2014, the leverage ratio of the household sector is calculated based on data provided by the Statistics and Analysis Department of the People’s Bank of China, while the leverage ratios of the other three sectors are calculated based on their average debt growth rates. See Table 4.1 for the results of calculation. First, leverage ratio of the household sector. It is calculated by dividing the amount of loans by GDP. According to the table about credit funds of financial institutions in RMB and foreign currencies by sector compiled by the People’s Bank of China, as of the end of 2014, the outstanding loans of the household sector stood at RMB 23.1 trillion. Table 4.1 shows that the leverage ratio of China’s household sector had been on the rise, growing from 3% in 1996 to 36% in 2014. The 18-year period can be divided into three phases. In the first phase (1996–2002), the leverage ratio increased from 3% to 14%; in the second phase (2003–2008), the ratio stayed around 18%; in the third phase, the ratio went up from 24% in 2009 to 36% in 2014. Table 4.1  Leverage ratios in China by sector (%) Year

Leverage ratio, household sector

Leverage ratio, government sector

Leverage ratio, non-financial business sector

Leverage ratio, financial sector

Total

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2012 2013 2014

3 3 4 5 7 10 14 18 18 17 18 19 18 24 28 31 34 36

23 25 28 33 33 35 38 39 40 41 40 43 41 49 49 53 56 62

84 93 99 100 94 93 102 109 103 95 100 97 98 99 101 113 114 121

4 5 6 7 7 8 8 9 9 11 12 13 13 15 13 18 19 22

113 125 138 145 142 147 162 174 170 163 171 172 170 187 192 215 223 241

86  

J. MA ET AL.

The leverage ratio of China’s household sector is far below 68.25%, the average level in developed countries. For example, the household debt-toGDP ratio is 77%, 86%, 92% and 73%, respectively, in the United States, the United Kingdom, Canada and Spain, a clear proof of debt-fueled spending. Second, leverage ratio of the government sector. As shown in Table 4.1, government debt is estimated to be RMB 33 trillion and RMB 39.6 trillion and take up 56% and 62.2% of GDP, respectively, in 2013 and 2014 based on the average annual growth rate of government debt from 1996 to 2012. Table 4.1 also shows that the government sector’s leverage ratio began to rise steadily following the 1997 Asian financial crisis, apparently under external influence. Government debt made up only 23% of GDP in 1996 but the percentage reached 33% in 1999, presenting an increase of ten percentage points over three years. Then it grew by eight percentage points to 41% in 2008. After the global financial crisis, the leverage ratio of China’s government sector rose by eight percentage points in 2009 alone and stayed above 50% in the four years that followed. The ratio of government debt to GDP in China was far below 117.5%, the average level in developed countries. It was 234%, 139% and 132%, respectively, in Japan, Italy and Spain. It should be noted that government debt includes debts of both the central and local governments. Local governments in China were not legally allowed to be indebted before the amended Budget Law came into force, so the final accounts of previous years only included central government debt. If local governments are taken into consideration, it would be much more complicated. Figure 4.1 shows a comparison between local government growth rates and economic growth rates. Compared with developed countries, government debt in China has three characteristics: First, the central government has the obligation to offer guarantee. Under the political system in China, there is no legal basis for over-indebted local governments to file for bankruptcy. Defaults or bad debts on the part of local governments, embodied as local fiscal deficits or non-performing loans of banks, have to be financed, either with budgetary resources or through the central bank, or written off by the central government. Therefore, government debt in the case of China should include the debts of local governments. In developed countries, ­however, the central government does not take responsibility for the debts of local governments which can file for bankruptcy in the case of

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

87

Fig. 4.1  Growth Rates of Local Government Debt and the Economy. Notes: (1) The local government debt in the figure only includes the debts that local governments have the obligation to repay. Contingent liabilities and debts the government has rescue responsibility for are complicated, so it is not desirable to do simple addition. There will be a detailed analysis in later sections. (2) The growth rate of local government debt for 2002 is the average annual growth rate in the period 1998–2002, the figure for 2007 is the average annual growth rate in the period 2002–2007 and that for 2013 is the average annual growth rate in the period 2010–2012. The data for 2012 only cover the debts of 36 local governments audited by the National Audit Office of China. Sources: (1) GDP growth rates: National Bureau of Statistics of China, China Statistical Yearbook 2014. (2) Local debt growth rates: Audit Results of National Audit Office (No.35 of 2011, No.24 of 2013, and No.32 of 2013)

insolvency. For that, the calculation of government debt in these countries does not cover the debts of local governments. It is also true with the IMF statistics on government debt. Second, there are no effective constraints on government debt under the system of state-owned land property rights system and the state-owned banking system. Financial institutions “dare to” grant loans to local governments only because of the special institutional arrangement that local governments can exercise their state-owned land property rights. The basic financing model for local governments in China is to offer state-owned land rights as security for commercial bank loans. Under such a model, land prices determine the financing scale and debt risks of local governments. A sharp fall in land prices would incur significant insolvency risks.

88  

J. MA ET AL.

Third, implicit liabilities of the government deserve attention. Different types of government debt incur risks to varying degrees, so simple addition is not desirable. According to the fiscal risk matrix developed by Polackova (1998), to accurately demonstrate the debt risk levels of the government sector, the probability or level contingent and implicit liabilities should also be calculated, apart from explicit liabilities. This concerns the types of government debt and conversion factor for estimation. For example, in the 2013 audit of local government debt, the National Audit Office divided local government debt into three categories, that is, liabilities that local governments have the obligation to pay back, contingent liabilities and liabilities that local governments may have rescue responsibility for, and set the conversion factors for the latter two as 19.13% and 14.64% according to the actual repayments since 2007. As a result, the leverage ratio of local governments in China was estimated to be 39.43% at the end of 2012. To present a full picture of government debt in China, apart from the liabilities of the central and local governments, consideration should also be given to at least the following actors: (1) the financial gap in the Social Security Fund; (2) the outstanding debt of four asset management companies stripped off from state-owned banks; (3) domestic and foreign debt issued by policy banks (estimated as sovereign debt according to IMF standards); (4) net losses of central and local state-owned enterprises (SOEs) and (5) legacy from previous reforms, such as the compensation gap as a result of wage reform at public institutions, unsettled accounts of local governments as a result of the reform of grain marketing system. Therefore, both explicit and implicit liabilities as described above (despite the different conversion factors) should be taken into account in the estimation of government debt in China. Given that, the total leverage ratio of China’s government sector is probably larger than 39.43%. Third, leverage ratio of the non-financial business sector. Based on the analytical framework and data on total social financing provided by Li et  al. (2013),1 we estimated the outstanding debt of the non-financial business sector and its proportion in GDP in 2013 and 2014 (Table 4.1). 1  It should be noted that the estimated debt of the non-financial business sector includes to a considerable degree the liabilities of local government financing platforms (LGFPs). To avoid repeated estimation, the liabilities of LGFPs estimated earlier need to be taken out here.

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

89

Apparently, the leverage ratio of China’s non-financial business sector had been on the rise during the period 1996–2014. The sector’s debt rose from RMB 6 trillion at the beginning of 1996 to RMB 77 trillion in 2014 and its proportion in GDP increased by 37 percentage points from 84% in 1996 to 121% in 2014. The leverage ratio of this sector is closely connected with the economic cycle. Along with the cyclical fluctuations of the economy, the leverage ratio of China’s non-financial business sector also experienced four periods of fluctuation. From 1996 to 1999, the leverage ratio of this sector grew by 16 percentage points from 84% to 100%; from 2000 to 2003, the ratio rose by 15 percentage points from 94% to 109%; between 2003 and 2007, the ratio went down by 12 percentage points from 109% to 97%; between 2008 and 2014, the ratio went up again from 98% to 121%, presenting an increase of 23 percentage points. Internationally speaking, the leverage ratio of China’s non-financial business sector was far higher than that of other countries, which is 74% in the United Kingdom, 67% in the United States and 60% in Canada. In 2014, the leverage ratio of China’s business sector reached 121%, higher than the average level of developed countries—82.8% as well as the threshold level of OECD member countries—90%. Therefore, it is an alarmingly high ratio. Fourth, leverage ratio of the financial sector. Financial enterprises are special in that they sell financial products based on a high level of liabilities and their liabilities are the assets of non-financial sector. To avoid repeated calculation, circulating medium and deposits are considered not as the financial sector’s debt but as various bonds issued by the sector. On the basis of calculations done by Li et al. (2013), we estimated the total liabilities and debt-to-GDP ratio of China’s financial sector in 2013 and 2014 (Table 4.1). The debt-to-GDP ratio of China’s financial sector showed a clear upward trend during the period 1996–2014. The total liabilities of this sector increased by 48.9 times from RMB 286.2 billion in 1996 to RMB 14 trillion in 2014 and their proportion in GDP rose by 18 percentage points from 4% in 1996 to 22% in 2014. The background to the sharp increase in the financial sector’s leverage ratio is the fall in banks’ capital adequacy ratio caused by the large-scale lending to support high-speed economic growth under a financing model where direct financing is the main approach. Commercial banks have issued a large amount of subordinated debt to raise capital. China Development Bank and other policy financial institutions have also issued financial bonds on a large scale. Internationally speaking, however, the

90  

J. MA ET AL.

leverage ratio of China’s financial sector is far below the average level of developed countries—91%. For example, the ratio is 255% and 112% in the United Kingdom and Japan, respectively. Fifth, total debt and total leverage ratio. China’s total debt was RMB 153.7  trillion and total leverage ratio was 241% in 2014 (Table  4.1). Apparently, China’s debt-to-GDP ratio has experienced increases and decreases since 1996 but the general trend is upward. The leverage ratio began to grow rapidly after the Asian financial crisis, from 113% in 1996 to 174% in 2003. It fell to 163% in 2005 and picked up again after the global financial crisis in 2007, rising from 170% in 2008 to 241% in 2014. China’s debt-to-GDP ratio was below the average level of developed countries—359.5%. In 2014, the total leverage ratio exceeded 500% in Japan and the United Kingdom, 300% in Spain, France and Italy and 270% in the United States, Germany and Canada (Table 4.2). The analysis above leads to two basic conclusions. One is that China’s debt-to-GDP ratio is on the whole lower than those in most developed countries and at a moderate and controllable level but it has been growing at an alarmingly high rate in recent years. The other is that the leverage ratios of the household and financial sectors are in a proper range but Table 4.2  China’s total leverage ratio in comparison with that of other countries, 2014 (%)

United States Japan United Kingdom Canada France Germany Italy Spain Average China

Leverage ratio, household sector

Leverage ratio, government sector

Leverage ratio, non-financial business sector

Leverage ratio, financial sector

77

89

67

46

279

65 86

234 92

101 74

112 255

512 507

92 56 54 43 73 68.25 36

70 104 80 139 132 117.5 62

60 121 54 77 108 82.75 121

54 65 91 55 50 91 22

Total leverage ratio

276 346 279 314 363 359.5 241

Sources: World Economic Outlook, IMF; BIS; Haver Analytics; National Central Banks; McKinsey Global Institute Analysis. The figures for other countries are data for Q2 of 2014, while those for China are estimates for the end of 2015. The ratios of the financial sector are estimated

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

91

Table 4.3  Leverage ratios in China, 2014 (%)

China Average level of developed countries Comparison and conclusion

Leverage ratio, household sector

Leverage ratio, government sector

Leverage ratio, non-financial business sector

Debt-toGDP ratio, financial sector

Total debt-to-GDP ratio

36 68.3

62 117.5

121 82.8

22 91

241 359.5

The leverage ratio of China’s household sector is rather low, about half the level of developed countries

The leverage ratio is not high for explicit liabilities but implicit liabilities grow fast and the repayment pressure is enormous

Alarmingly higher than the average level of developed countries—82.75%— and the threshold level of OECD member countries—90%

Lower than the average level of developed countries

China’s total leverage ratio is lower than the average level of developed countries and at a moderate and controllable level but it grows fast

those of the business and government sectors have been growing very fast, the former in particular, and the government sector has heavy implicit liabilities and enormous repayment pressure (Table 4.3).

2   Causes of the Increase in China’s Leverage Ratio The increase in China’s leverage ratio may have something to do with the macroeconomic changes brought by the development of the banking system and financial market as the country makes the shift from a planned economy to a market economy. The excessively high growth rate of the leverage ratio, however, is closely connected to the following factors. 2.1  Overreliance on Indirect Financing Over the years, indirect financing has been the dominant financing approach in China. Direct financing takes up a very small share, and a

92  

J. MA ET AL.

Table 4.4  Financing structure of businesses Debts instruments

2006 2007 2008 2009 2010 2011 2012 2013 2014

Bank loans

Equity financing

Amount (RMB 100 million)

Share (%)

Amount (RMB 100 million)

Share (%)

Amount (RMB 100 million)

Share (%)

1891 1953 5331 15,468 15,162 22,582 31,860 69,651 55,116

5.38 4.23 10.59 13.24 14.50 21.69 32.14 40.30 34.31

30,595 36,406 41,704 96,290 79,511 74,700 62,990 94,748 97,800

86.99 78.84 82.82 82.45 76.06 71.76 63.55 54.80 60.88

2684 7816 3318 5033 9865 6821 4264 4074 7718

7.63 16.93 6.59 4.31 9.44 6.55 4.30 1.30 4.80

Source: Monetary policy reports published by the People’s Bank of China from 2006 to 2014

market-based equity funding mechanism for businesses is absent. In the period 2006–2014, equity financing made up less than 10% of total social financing, registering only 1.3% in 2013 (Table 4.4). The overreliance on bank loans has undoubtedly contributed to the high leverage ratio of nonfinancial businesses. The overreliance on bank loans and the strict regulation of the banking industry have led to very high lending rates in China. The large amount of principal plus high interest rate means heavier debt burden and higher leverage ratio for the non-financial business sector. For example, the benchmark one-year lending rate is 6% in China, which is higher than the 2.25% in the United States, 3.79% in Canada, 1% in Japan, 3.5% in Germany and 3.79% in Italy (Table 4.5). 2.2  Low Efficiency in the Use of Funds If a company is innovative and uses funds efficiently, liabilities would not pose any risks because it has the capacity to repay all the debts. If a company operates at high costs and with low efficiency and suffers heavy losses, however, it will not be able to repay debts and would need new loans to pay back old ones, thus leading to growing liabilities and decreasing effective assets. It is no different from increasing the company’s debt burden and leverage ratio. This is a very important cause of the high leverage ratio of Chinese enterprises but due attention is yet to be paid. A number of finan-

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

93

Table 4.5  Benchmark interest rates in major countries, 2014 Central bank’s benchmark interest rate Developed countries  United Federal funds target States rate—0.25%  Canada Target for the overnight rate—1.0%  Japan Uncollateralized overnight call rate—0.1%  Germany ECB benchmark interest rate—0.15%  Italy ECB benchmark interest rate—0.15% Emerging markets  China Benchmark interest rate for one-year loans—6.0%  India  Russia  Brazil  Mexico  South Korea

8% 8% 11% 3.0% 2.5%

Lending rate for businesses

One-year lending rate—2.25% Three-year lending rate—3.79% One-year lending rate—1% One-year lending rate—3.5% One-year lending rate—3.79%

Benchmark one-year lending rate is 6.0%. For banks, there are various fees so the actual one-year lending rate is higher than 8% Benchmark lending rate—10.25% Less than 12% Prime one-year lending rate—around 15.7% One-year lending rate—around 7% One-year lending rate—4.58%

Source: Data released by the media

cial indicators can be used to measure efficiency in the use of funds. The top three are funds turnover, amount of funds involved in receivables and selling expenses ratio. First, funds turnover. Funds turnover is the number of times a company’s funds are utilized during the reporting period and the measure of such turnover can be working capital turnover ratio or asset turnover ratio. According to Fig.  4.2, the asset turnover ratio of industrial enterprises above a designated size in China grew steadily between 2003 and 2008, from 0.83 to 1.15 times per year, dropped to 1.09 in 2009, rose to 1.22 in 2011, and then fell gradually to 1.17 in 2013. The same trend can be found in the working capital turnover ratio of industrial enterprises. It fell sharply from 2.67 in 2009 to 2.43 in 2010, rose to 2.62 in 2011, and then dropped back to the range of 2.5 to 2.6. In particular, the ratio stayed at an even lower level among state-owned holding companies (Fig. 4.3).

94  

J. MA ET AL.

Fig. 4.2  Funds turnover of industrial enterprises above a designated size, 2003–2013. Note: Asset turnover ratio  =  sales revenue/total assets. Source: Database of the National Bureau of Statistics

Fig. 4.3  Working capital turnover ratio of industrial enterprises above a designated size, 2003–2013. Source: Database of the National Bureau of Statistics

Second, amount of funds involved in receivables. The change in the funds involved in receivables depends on whether the company can collect payments in time after products are sold. The default in payments suggests efficiency in the use of funds in the entire market during a period. A measure in this regard is receivables turnover ratio. The higher the ratio, the smaller the amount of funds involved in receivables, the higher the degree of assets liquidity and the greater the solvency. In addition, a high receivables turnover ratio helps to reduce collection expenses and losses incurred by bad debts and thus increase the profitability of the company’s current

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

95

Fig. 4.4  Receivables turnover ratio of industrial enterprises above a designated size, 2003–2013. Note: Receivables turnover ratio = revenue from main business/ average balance of accounts receivable. The dotted line in the figure indicates the growth trend. Source: Database of the National Bureau of Statistics

assets. According to Fig. 4.4, the receivables turnover ratio of industrial enterprises above a designated size grew steadily during the period 2003–2008 from 7.8 to 11.4, fell to 10.6% in 2009, increased again to 11.9 in 2011, and then dropped back to 10.7 in 2013—a level close to that in 2009. Third, selling expenses ratio. The ratio can be calculated by dividing the sum of selling and financial expenses by sales revenue. The change in the selling expenses ratio shows the change of efficiency in the use of funds among industrial enterprises above a designated size. According to Fig. 4.5, the ratio fell quickly during the period 2003–2007 from 5.1% to 3.8%, experienced no major decreases from 2008 on, fell to 3.5% in 2011 and then rose again to the range of 3.7% to 3.8%. 2.3  Large Amounts of Funds Occupied Ineffectively Because of Overcapacity Overcapacity has become an outstanding problem in China as the government seeks to adjust the growth rate of the economy. For enterprises, overcapacity means a lot of funds, fixed capital and even working capital, are not put into effective use, and a considerable portion of them are borrowed from banks. Therefore, overcapacity is synonymous with the ineffective occupation of funds and liabilities, and aggravated overcapacity means more serious ineffective occupation of funds and “zombie” debts.

96  

J. MA ET AL.

Fig. 4.5  Selling expenses ratio of industrial enterprises above a designated size, 2003–2013. Note: Selling expenses ratio = (selling expenses + financial expenses)/ sales. Source: Database of the National Bureau of Statistics

Now about 25% of the production facilities in China’s industrial sector are left unused. Overcapacity is most serious in five industries including steel, cement, aluminum electrolytic capacitors, sheet glass and shipbuilding. According to international standards, it is normal that 79–83% of production facilities are used; over 90% being put into use indicates lack of capacity and less than 79% being put into use suggests overcapacity. As of the end of 2012, China contributed to 46% of the global crude steel production capacity and 70.8% of its capacity was used; the two figures were 42% and about 72% for aluminum electrolytic capacitors, 60% and 73% for cement and 50% and 68% for sheet glass; about 50% of the shipbuilding industry’s production capacity was used. In short, nearly one-third of production capacity was left unused in the five industries.2 It means that one-third of the enterprises’ fixed liabilities are occupied ineffectively. It should also be noted that quite a few local governments encourage production activities of overcapacity industries to create false prosperity but the result is that the enterprises are not able to sell their products and recover investment and have to borrow more from banks, hence a higher leverage ratio. 2.4  Overuse of Monetary and Credit Policies as Instruments In response to the drastic changes in domestic and international economic landscapes, the Chinese government began to adopt a monetary policy in 2

 Reference: Wang (2013).

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

97

Table 4.6  Credit granting in China, 2008–2014 Year

RMB-denominated loans, social financing (RMB 100 million)

Bank loans (RMB 100 million)

M2/ GDP

2008 2009 2010 2011 2012 2013 2014

49,041 95,942 79,451 74,715 82,038 88,916 97,813

320,040.87 425,590.66 509,220.60 581,890.25 672,870.46 766,320.66 864,120.66

1.50 1.75 1.77 1.76 1.82 1.88 1.93

Source: CEIC, National Bureau of Statistics and People’s Bank of China

2008 to boost the real economy. Since then, the amount of bank loans and the M2-to-GDP ratio have been on the rise in the context of steady economic growth (Table 4.6). During the six years from 2008 to 2014, the amount of RMB-denominated loans as part of social financing grew from RMB 4.9041 trillion to RMB 9.7813 trillion, presenting an average annual growth of 12%, which was also the GDP growth rate during the same period. The fast growth of RMB-denominated loans and other types of financing led to the continued increase in the M2-to-GDP ratio, from 1.5 in 2008 to 1.93 in 2014—a fairly high level compared with other countries.

3   Risks Incurred by a High Leverage Ratio Every financial crisis is a credit crisis and to defuse it is to deleverage, as evidenced by the Great Depression of 1929, the 2008 financial crisis that started in the Wall Street, as well as the violent fluctuations in China’s stock market (A-shares) in 2007 and recent years. A high leverage ratio would incur risks in four aspects for the Chinese economy. 3.1  Liquidity and Solvency Risks Liquidity risk is the possibility of failure to pay debts due for lack of liquidity, while solvency risk is the possibility that the debtor becomes insolvent. The former would provoke a liquidity crisis, while the latter would cause a solvency crisis. The leverage ratios of the business, government and household sectors show that the business sector faces considerable liquidity and solvency risks. Overcapacity industries, in particular, face huge liquidity risk, and with less than 70%, or even 50%, of the equipment put into

98  

J. MA ET AL.

operation, replacement cost is sure to fall sharply and the value of assets on the balance sheet will decline. That will eventually lead to a solvency crisis. The government and household sectors in China do not face any solvency risk but run a high liquidity risk. Under the unique property rights system regarding urban land and mineral resources in China, the massive size of assets owned by local governments, apart from SOE assets, may not be found anywhere else in the world. Even if the falling prices of land and mineral resources in recent years are factored in, the debt-to-asset ratio in the government sector might be the lowest in the world. However, liquidity risks are grave in this sector. Large amounts of loans granted by commercial banks go to public utility and infrastructure projects of local governments whose returns are low and service life is as long as decades and even centuries, thus resulting in serious maturity mismatches. A substantial part of social financing in China is indirect financing of the banking industry. The risks incurred by local government debt are surely concentrated in the banking industry. 3.2  Financial Imbalance Risk Liquidity and solvency risks can be understood as risks on the microeconomic level. As the financial system develops, the capital market, the ­banking system and the money market will become ever more closely connected, and leverage-incurred risks will spread to macroeconomic mechanisms. First, the quality of assets in China’s banking industry has declined due to the shrinking demand and economic slowdown in other countries. The spread of risk varies from bank to bank given the differences in the size and allocation of assets. According to the China Financial Stability Report 2015 published by the People’s Bank of China, as of the end of December 2014, the interbank risk exposures of 28 commercial banks with over RMB 400  billion in total assets are as follows: (1) Once credit default occurs among major commercial banks, risks will spread to five banks and the spread of risks will go in two rounds, consuming about 9% of the 28 banks’ total capital. Twenty postal savings banks are likely to be impacted by interbank credit default. (2) Once interbank credit default occurs among joint stock commercial banks, risks will spread to 12 banks and the spread of risks will go in three rounds, consuming about 24% of the 28 banks’ total capital. (3) Once interbank credit default occurs among city

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

99

commercial banks, risks will spread to four banks and the spread of risks will go in one round, consuming about 7% of the 28 banks’ total capital. The overall impact is not significant. (4) Once interbank credit default occurs among rural commercial banks, no risks will spread to other banks. In short, once primary default risk arises, joint stock banks in China face the greatest financial risks. Second, close attention should be paid to the leverage risk of China’s capital market. Since June 2015, substantial risks have built up in China’s securities market due to various mismatches. In the context of positive expectations for the stock market, “innovative” derivatives such as securities margin trading, umbrella trusts and off-exchange trading have been introduced and large amounts of household deposits and bank funds flow into the stock market. Structural and maturity mismatches become increasingly serious. The growing leverage leads to much more risks in the capital market and the deleveraging process is sure to witness rapid risk exposure. The regulators have started with strictly controlling illegal off-exchange financing activities to control the increase of leverage. It means that participants that have increased leverage would need to adjust their balance sheets. The butterfly effect of active individual activities has caused stock market plunges. The lesson learned here is clear and bitter. 3.3  Risk of Real Economy Collapse A high leverage ratio may also lead to the collapse of the real economy and the process is similar to that in which risks spread within the financial system. When a highly leveraged business plunges into financial distress, it will be unable to pay the debts due without external assistance or capital injection and thus run the risk of going bankrupt. Moreover, as economic climate changes, asset prices of market participants will fluctuate widely, and the financial accelerator mechanism will arise. Under such a mechanism, when the economy is in a boom period, the high net asset value of companies has positive impact on investment and companies can offer more collateral to reduce financing costs; when the economy suffers a downturn, the fall in net asset value will amplify the impact on investment. After the subprime crisis broke out in the United States, Richard Koo, Chief Economist at the Nomura Research Institute, put forward the idea of “balance sheet recession” in his book The Holy Grail of Macroeconomics: Lessons from Japan’s Great Recession. According to the book, the Japanese economy failed to recover soon after the bubble burst because businesses’

100  

J. MA ET AL.

motivation for borrowing money kept falling. In the era of bubble economy, Japanese companies invested bank loans in land but the sharp decline in land prices led to the serious shrinking of these companies’ assets. They had to shift the focus from “maximizing profit” to “minimizing debt”, giving up investment and using available cash flow to repay debt. That results in a fall in demand and interest rate as well as deflation. The sluggish economy caused further decrease in asset prices, so the companies became insolvent and went bankrupt. The interplay between the real economy and the financial system plunged the economy into balance sheet recession. Similar situation can also be seen in China in the last two decades. Market-oriented reforms, demographic dividend and China’s accession to the WTO have led to the rapid growth of the Chinese economy as well as the continued rise of asset prices. Market participants tend to borrow money for better development, so the economy’s reliance on debt keeps growing. The leverage ratios of the household, business and government sectors have been on the rise, posing more risks to the financial system and even the economy as a whole. Two concrete examples are the increasing amount of government debt and the difficulty of companies in overcapacity industries in paying debt. 3.4  Social Instability Risk Social instability risk is the possibility of social relations deterioration caused by changes in the relations between assets and liabilities and the adjustment of interests brought by repair effort. The level of leverage ratio reflects, to some extent, the role of the financial system in the economy as well as the status of various players in primary income distribution. The adjustment of leverage ratio is a process of balance sheet repair on the microeconomic level and a process of pricebased asset restructuring on the macroeconomic level. If some enterprises stop normal operations for deleveraging, it may influence other enterprises. The deleveraging efforts in some industries may lead to massive adjustments across sectors. In a nightmare scenario, large-scale business failures and unemployment may occur. For the government sector, to deleverage means, apart from capital injection, to reduce other expenses and use the money saved to repay debt or to encourage the central bank to issue more notes and thus reduce debt. Reducing public expenditure, especially the spending on welfare and

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

101

people’s lives, will provoke public discontent. A typical example in this regard is the social and political upheavals caused by reduced welfare spending in Greece. In the latter case, issuing notes to settle government debt (deficit), recover SOE losses and write off bad debt of financial institutions or repaying past debts in social security is no different from an overdraft of sovereign credit, which will cause serious inflation and thus provoke social instability. For the business sector, there are in theory three ways to repair the balance sheet: (1) introducing new shareholders, (2) selling assets to pay the debt and (3) filing for bankruptcy. For a heavily indebted company, it would be very difficult to attract new investors except that an appealing debt restructuring plan is drawn up. There is a high chance that the company will have to face either of the latter two scenarios that involve the scaling down of production and lay-off of workers. For the household sector, there are two scenarios in which a high leverage ratio would cause social instability. If residents borrow money with family property, land management right or homestead as collateral, the loss of asset in case of failure to repay a loan would influence their lives and thus cause social problems. If residents are creditors, especially in the case of private lending where no collateral or pledge is provided, borrowers’ failure to repay the debt would influence the lenders’ lives and thus cause social problems. According to the China Household Finance Survey (CHFS), 22.3% of the households in China were involved in informal finance in 2013. About 420,000 households involved in loan sharking did not rely on their own money but adopted the practice of borrowing money at a low interest rate and lending at a higher rate.3 A considerable proportion of those speculating in the stock market are from the household sector. Individual investors in the stock market pose huge risks. In the second quarter of 2015, about 8.8% of the households in the country, or 37 million households, invested in the stock market.4 Some households suffered serious losses due to the violent fluctuations of stock prices. In addition, some stock investors are not fully aware of the risks. Stock market crashes would lead to petitions and other mass incidents and thus undermine social stability. 3  Survey and Research Center for China Household Finance. China Informal Finance Development Report. January 2014. 4  The figures are the result of a survey conducted by the Survey and Research Center for China Household Finance among 5000 representative households across the country from June 15 to July 2, 2015.

102  

J. MA ET AL.

China is a major country. If not well managed, individual risks may evolve into industry risks, financial risks may spread to other sectors of the economy, partial risks may become systemic risks and economic risks may eventually translate into social and even political risks.

4   Recommended Measures for Defusing China’s Systemic Financial Risks In the context of economic downturn and asset depreciation, the risks incurred by high leverage ratios in the stage of high-speed economic growth will gradually be exposed. In addition, economic downturn makes it difficult to deleverage. When government revenue decreases, reduced expenditure will lead to further decline in total demand. To address this challenge would require a balance between the effort to stabilize growth and that to prevent risks. 4.1  Restructuring the Debt: To Ease the Pressure in Short-Term Debt Repayment There are four ways of debt restructuring for the business and government sectors. First, to adjust the debt maturity structure, namely to extend debt maturity or pay short-term debt with long-term debt. Debt extension is an agreement reached between the debtor and the creditor, which is easy and convenient to execute. When a company is having financial difficulty, debt extension will win it more time for financial adjustment so as to avoid bankruptcy. A local government under enormous debt repayment pressure can also ease the pressure through debt extension. The quotas set by the Ministry of Finance for local debt replacement on three occasions since 2015 amount to RMB 3.2  trillion. Debt replacement is a type of flexible debt extension. It should be noted, however, that the effort to ease the pressure of local governments and their financing platform companies in paying debts should go with stricter debt constraints to reduce moral hazard. As for paying short-term debt with long-term debt, companies can issue bonds to raise long-term funding or work with policy banks to obtain mid- or long-term loans. Second, to adjust the interest rate structure, namely to replace lowinterest-rate liabilities with high-interest-rate ones. Interest rate varies between long-term and short-term debts. As a rule, the rates on long-

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

103

term debt are relatively lower, while those on short-term debt are higher. For example, the interest rates on bonds are lower, while those on bank loans are higher. Currently, companies should choose financing approaches allowing lower interest rates. (a) To lower interest rate levels in the market by reducing the interest rates on government bonds. The interest rate on ten-year government bonds is only 2.2% in the United States and 0.21% in Japan—a record low in the country and negative in the euro zone. In China, however, the rate has been on the rise, reaching 3.64%. As long as the interest rate on ten-year government bonds does not fall, there will be no declines in the interest rates on long-term loans and longterm corporate bonds, hence the high financing costs of companies. The interest rates on long-term government bonds are the basis of interest rates on all long-term liabilities. Therefore, it is advisable to reduce the interest rates on long-term government bonds to bring about a decrease in interest rate levels in the market as a whole. (b) To reduce companies’ financing costs by issuing corporate bonds. Interest rates in the securities market are lower than those on loans granted by commercial banks. Major enterprises should be encouraged to issue bonds. On the one hand, the development of the securities market helps major businesses to obtain funding at lower costs; on the other, it gives small and medium-sized enterprises (SMEs) more opportunities to get loans. For example, a real estate company in Xicheng District of Beijing raises a loan from China Development Bank (CDB) at a rate of about 5% with land as collateral; it also raises funds from the securities market at a rate of about 4%. Third, to adjust the loan structure, namely to replace foreign debt with domestic debt. The costs of indirect financing in China are higher than those in developed countries. One-year lending rate in China stands at 5.1% but companies usually get loans at a premium of 20% or even 100%. In the United States, however, the interest rate on dollar-denominated loans is only around 3% and the rates on dollar bonds are usually not high, either. Now Chinese companies have four ways of financing in a foreign country. The first one is to issue corporate bonds. The dollar bonds issued by China National Offshore Oil Corp (CNOOC), Bank of China, Vanke and Greentown are very popular with foreign investors. In April 2013, CNOOC issued bonds in the United States totaling USD 2 billion, including USD

104  

J. MA ET AL.

750 million in three-year bonds, USD 500 million in five-year bonds and USD 750 million in ten-year bonds whose coupon interest rates were set, respectively, as 1.45%, 1.95% and 3.4%. The second one is to ask for loans from commercial banks. Some foreign banks require loan guarantees in which case domestic banks can offer guarantees for applicant enterprises. The third one is to extend the scope of pilot free trade zones, increasing the number of free trade accounts and expanding business scope, so as to further facilitate Chinese companies’ financing activities in foreign countries. The fourth one is to set up special-purpose entities for raising funds overseas. Chinese enterprises can leverage their domestic assets or equity to directly set up companies overseas or gain a controlling stake in foreign companies for the purpose of equity financing. Fourth, to offer discounts on some debts so that companies have more time for making financial adjustments. When necessary, the creditor can offer a discount on the liabilities that the debtor has difficulty in repaying. This can help commercial banks to recover part of the loans and also give the companies in difficulty a chance to rise again by removing their heavy debt burden. In 2013, Wuxi Suntech Power Co., Ltd. declared bankruptcy as a result of a serious debt crisis. Jiangsu Shunfeng Photovoltaic Technology Co., Ltd. promised to offer RMB 3  billion in cash which would be used to pay the debt together with Suntech’s receivables. It means that Suntech only needed to pay 30% of its total debt. If the bankruptcy proceedings had been commenced, the liquidity ratio would be probably less than 17% since Suntech’s net assets amounted to only RMB 2.28 billion and a series of expenses would be incurred, including severance pay for employees, transfer fee, registration fee and other fees charged in the bankruptcy process. Debt restructuring usually goes with asset restructuring. Most companies in overcapacity industries have one thing in common: they are asset heavy. For them, there are two approaches to asset restructuring. (a) To reduce the proportion of fixed assets by means of leasing. Enterprises can entrust financial leasing companies to buy the equipment or production lines and then lease them for use. This helps to reduce the amount of money paid in a lump sum and lower the leverage ratio. For example, quite a few airlines lease airplanes from aircraft leasing companies and in some cases airplanes and flight crew are leased together. In this way, they do not need to bear the high cost of buying airplanes.

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

105

(b) To choose an asset-light business model. Enterprises in overcapacity industries can run business in collaboration with other companies to leverage funds and make full use of intangible assets, thus making the shift to an asset-light model. For example, real estate companies traditionally have an asset-heavy development model. If an assetlight model is adopted, most part of the financing responsibility lies with real estate funds, while developers can focus on putting into use their expertise in design, construction, investment attraction, management and so on. 4.2  Rebuilding the Financial System: To Establish a Multi-tier Financial Service System First, gather speed in developing a multi-tier direct financing system. The proportion of direct financing should be increased so that both indirect and direct financing approaches will play their due roles. (a) To replace the approval system for main board IPO with a registration system. Registration helps to reduce approval costs, streamline the approval process and develop a mature market. Securities registration has been a practice long adopted in the United States. The Securities and Exchange Commission (SEC) only checks the quality of information exposure in the stage of securities issuance, while the pricing, placement, rights and listing are all left to the market mechanism. China and the United States differ in market environment and the basis for the rule of law, so detailed rules for the adoption of a registration system should be laid down. It is advisable to pass the draft amendment to the Securities Law as soon as possible and make early preparations for that. The procedures concerning information disclosure should be improved first and then sound legislation and judicial procedures concerning illegal activities such as false disclosure should be put in place. (b) To improve the National Equities Exchange and Quotations (NEEQ), also known as the New Third Board, and connect it with equity markets. The New Third Board, local equity markets and property rights markets in China run separately with their respective trading systems and platforms. It means waste of resources in the development of systems and the inconsistent standards of the systems make it hard to make specific enquiries and compare compa-

106  

J. MA ET AL.

nies of the same category, causing confusion among market participants. In light of the problems in off-exchange trading markets, it is advisable to establish a national equity markets alliance under the guidance of the China Securities Regulatory Commission (CSRC). The alliance should be responsible for developing uniform standards and trading system to achieve connectivity among local equity markets. It should also take on a regulatory role as the industry association for off-exchange trading. The main responsibility of the CSRC is to regulate national off-exchange trading market, the national equity markets alliance enforces discipline over local equity markets and local equity markets exercise self-discipline. (c) To improve the model of equity-based crowdfunding. To develop equity-based crowdfunding, no capital pool should be created and funds should be entrusted to a third-party bank; no guarantee should be provided, either. In addition, with no revisions made to application laws, the requirement that the number of shareholders shall be no more than 200 must be met, according to the Company Law and Securities Law. Requirements on the number of investors in equity-based crowdfunding and the development stage of crowdfunding projects may be relaxed in the future to enable more flexible ways of financing. (d) To establish board upgrade mechanisms. Capital markets in China are divided into the main board (first board), growth enterprises market (second board), NEEQ (third board), regional equity market (fourth board) and equity-based crowdfunding market (fifth board). The financing capacity of the growth enterprises market is enormous. Overfunding often occurs and enterprises listed on this market are seldom motivated to shift to the SME board or main board. The main board and SME board are rather alike, so enterprises listed on them are not motivated to shift, either. Therefore, connection mechanisms should be established between the boards to allow for upgrades from the NEEQ to growth enterprises market and SME board, and from equity-based crowdfunding market to the NEEQ or even growth enterprises market so that enterprises can choose a market most suitable for their stage of development. It is advisable that the upgrade experimentation starts with the NEEQ by introducing fast-track service platforms for upgrades from the NEEQ to the main board and growth enterprises market. Eligible companies established via equity-based crowdfunding may apply for listing on the NEEQ.

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

107

Second, build a multi-tier indirect financing system. In China’s indirect financing system, state-owned commercial banks have always played a central role. The participation of small and mid-sized banks and private banks is very limited. For that, micro and small enterprises find it very hard and expensive to get a loan. A multi-tier indirect financing system will satisfy the financing needs of enterprises of different sizes, in different industries and with different types of ownership. (a) To lower the access threshold for private banks. The General Office of the State Council has recently issued the Guidelines on Promoting the Development of Private Banks. According to the Guidelines, the requirements on private banks are generally quite strict and should be relaxed moderately. The cases of the five established private banks show that the registered capital requirement is rather demanding— no less than RMB 500 million and no more than RMB 1 billion. In the United States, the capital requirement for establishing a commercial bank is USD 1 million. Therefore, China should lower the requirement to a proper level so that a large number of small banks can be established. (b) To relax control on interest rates, which is of great significance to the better allocation of resources. Upon the approval of the State Council, the People’s Bank of China decided to remove the ceiling on deposit interest rate fluctuations for commercial banks and rural cooperative financial institutions as of October 24, 2015. It marks that China has made a key step forward in the market-based reform of interest rate regime—a milestone in the history of China’s financial reform—and shows China’s confidence and resolve to push forward the reform. Market-based interest rates will help channel financial resources to industries and enterprises that need them most and have the brightest prospects. The market will play a decisive role in resource allocation so that allocation efficiency will become higher. Since China is in a crucial period where new engines of growth are being created, relaxing the control over interest rates will create more room for financial institutions to support industries and enterprises in a market-based way, which will help stabilize growth, adjust structure and benefit the people, thus promoting the healthy and sustainable growth of the economy. (c) To better implement the Deposit Insurance Regulations. China introduced a deposit insurance system in the first half of 2015. The system is intended to help better protect depositors’ interests,

108  

J. MA ET AL.

improve the financial safety net and promote financial stability in the long run. It will also be of great significance to the healthy development of the banking industry, the improvement of the industry’s competitiveness and the increase of its capacity to serve the real economy. Relaxed access threshold in the banking industry, relaxed control over interest rates and deposit insurance system will lay a solid foundation for the establishment of a multi-tier indirect financing system in China. Third, deleverage by converting debt to preferred stock. Compared with traditional ways of deleveraging and converting debt to common stock, the conversion of debt to preferred stock has its unique advantage. Traditional ways of deleveraging such as recovering bank loans and stop renewing loans can easily cause fund chain breaks for enterprises. The drawback of converting debt to common stock is that banks are directly involved in business operation so that the expertise of commercial banks cannot be brought into full play. The advantage of converting debt to preferred stock is that it does not cause fund chain breaks of indebted companies and ensures their right to business operation and management. Fourth, develop asset securitization. One of the reasons why some heavily indebted companies find it very hard to change their situation is the poor liquidity of assets. Where possible, such companies should sell, lease or securitize some of their assets to increase liquidity and repay debt with the cash earnings. They may securitize receivables, entrust property to real estate fund, lease fixed assets and so on. Fifth, promote mezzanine financing—a way of financing that has the elements of both debt and equity financing. Representative financial vehicles include equity warrants and convertible corporate bonds. The term of financing is long and the way of paying debt is flexible. Companies can choose between debt and equity according to their future funding needs and business performance. 4.3  Deepening Economic Reform: To Inject Greater Vitality into the Real Economy The causes of a company’s high leverage ratio include low efficiency in the use funds, lack of vitality and low level of competitiveness. Therefore, in the process of restructuring debt and rebuilding the financial system, intensified efforts should also be made to deepen economic reform and inject vitality into the real economy,

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

109

First, take all measures to reduce overcapacity. Long-term mechanisms for preventing and reducing overcapacity should be put in place, incorporating such approaches as elimination, transfer, integration and phase-out. (a) To shut down some overcapacity enterprises. Some local governments overstress the importance of employment, social stability, officials’ performance and other factors, which has resulted in overcapacity for a long time and the waste of substantial financial resources. It is advisable to shut down businesses not up to environmental standards and those excessively concentrated in a region. Stricter environmental requirements should be imposed on steel and aluminum electrolytic capacitor companies to achieve business shutdown. Loan granting should be strictly controlled to prevent overcapacity companies from further expansion. (b) To encourage mergers, acquisitions and bankruptcy. The bankruptcy proceedings in the United States help companies that have failed in competition to exit properly so that financial risks will not build up. Although China has Enterprise Bankruptcy Law, its implementation confronts resistance from local governments and various stakeholders. Since a sound M&A market is absent in China, some enterprises find it hard to exit through M&A. This undermines efficiency in the economy as a whole and poses risks to loan security of commercial banks. The government should avoid excessive intervention in M&A activities, allowing market participants to act on their own initiative. (c) To encourage overcapacity enterprises to go global. The Belt and Road Initiative and “Go Global” strategy should be leveraged to promote international cooperation on production capacity. It can start with infrastructure and then move to manufacturing. There is a huge international market for Chinese manufacturing enterprises. Many labor-intensive industries can be transferred to Africa where labor costs are low. The shoe manufacturer Huajian’s facility in Ethiopia runs very well. Latin America is suitable for the high-end equipment manufacturers. Zoomlion and China National Nuclear Corporation have invested in Argentina and high-speed rail is also expected to enter the market. In the United States and Europe, low- and midend Chinese manufacturers have some investment opportunities. Fuyao Glass has established operations in the United States where it enjoys transparent and efficient services and a composite tax rate of 43%, which is three percentage points lower than the 46% in China.

110  

J. MA ET AL.

(d) To accelerate upgrading and transformation in key industries. Construction projects in industries with serious overcapacity problems should be dealt with a systematic approach. Mechanisms should be introduced to encourage enterprises in such industries to introduce advanced technology and key equipment, increase their innovation capacity, adjust structures and make the shift to strategic new industries, advanced manufacturing, modern services and public utilities. Second, revitalize enterprises and enhance efficiency in the use of funds. As a rule, dynamic businesses are efficient in using funds. If a company can take effective measures to improve efficiency in the use of funds, new vitality will be injected into its business. In this respect, the following should be done: (a) To lead SOEs lacking in vitality out of competitive fields. Private enterprises usually have more flexible mechanisms and they are more motivated to enhance efficiency and control costs. SOEs should play their due role in stabilizing economic growth and providing public goods, so they should concentrate in industries that concern the national economy and people’s lives. About 80,000 SOEs are still concentrated in competitive fields such as hotels, real estate, catering and shopping malls. Such enterprises should adopt mixed ownership and the experience of pilot program should be promoted. They may also get listed on the NEEQ or regional equity market to gradually reduce shares held by the government. (b) To support the growth of innovative enterprises that are the future of industries. Innovative enterprises offer new products and services or adopt new business models, so their development potential and profitability levels are very high. Besides, most innovative enterprises are asset light and their leverage ratios are low. Promoting the development of such enterprises will help to revitalize the economy and boost efficiency in the use of funds across the board. Alibaba’s IPO in the United States raised over USD 20 billion. It did not apply to any Chinese commercial bank for a loan but attracted venture capital and equity investment to support some new projects. The general policies of the central government should be combined with distinctive local policies in supporting innovative enterprises. China should continue to implement the preferential policies that the State Council introduced for supporting micro and small enterprises and innovative enterprises at its executive meeting in February 2015. Local governments should

  CHINA’S LEVERAGE RATIO AND SYSTEMIC FINANCIAL RISK PREVENTION   

111

encourage enterprises to establish strategic alliances with universities and research institutes for developing innovative technologies, and build key labs, R&D centers and other research facilities so as to keep increasing their innovation capacity. Venture capitalists should be actively engaged to foster innovation. (c) To improve budgeting for higher efficiency in the use of funds. First, exercise strict control over purchasing. The funding for supplies purchasing has direct bearing on a company’s reserve fund turnover ratio. Companies should establish long-term strategic partnerships with eligible suppliers and pay for supplies in installments or with bank acceptance draft to reduce the use of cash. For import supplies, companies should settle accounts via interest arbitrage by targeting forward exchange rates so as to reduce financial expenses. Second, reduce the size of receivables. Bad debts should be avoided by using receivables tracking system, aging schedules, collection rate analysis system, bad debt preparedness system and credit management system, and improving product quality. Third, manage inventory properly. Keeping inventory at a proper level helps a company to operate with greater flexibility but too much inventory means higher storage costs and price fall risks for the company. Therefore, enterprises should control inventory properly and strive to reduce inventory costs and accelerate capital turnover.

References Yang Li et al. China’s National Balance Sheet 2013. China Social Sciences Press. 2013. Ceccheiti et al. The Real Effects of Debt. BIS Working Paper. No 352, 2011. Hana Polackova. Contingent Government Liabilities: A Hidden Risk for Fiscal Stabilities. World Bank Policy Research Working Paper, No. 1989. 1998. IMF. World Economic Outlook. Second Quarter. 201-1.

CHAPTER 5

A Study of the Monetary Policy Framework with Mutually Supportive Priceand Quantity-Based Tools Xiandong Yan and Yantao Zhang Abstract  After the world financial crisis in 2008, the use of quantity- and price-based tools for regulative purposes under certain circumstances has been a choice by many central banks. Even after the monetary policy reverts to the normal state, some countries have started to lay emphasis upon the combination of quantity- and price-based tools. This chapter, by building the Dynamic Stochastic General Equilibrium (DSGE) model encompassing residents, non-financial enterprises and the government, compares and analyzes the effect of monetary policy tools in six

NSSF Project: Project supported by National Social Science Foundation, “A Study on the Transformation of Monetary Policies in the Period of the New Normal”. Our thanks go to the two anonymous reviewers for their constructive feedback. The authors shall be responsible for their chapter. CLC: F820.3; Document code: A; Article ID: 1002—8102(2016)10—0059—13 X. Yan (*) People’s Bank of China Financial Survey and Statistics Department, Beijing, China Y. Zhang Changsha Branch of the People’s Bank of China, Changsha, China © The Author(s) 2019 D. He, C. Wang (eds.), A New Era, https://doi.org/10.1007/978-981-10-8357-0_5

113

114  

X. YAN AND Y. ZHANG

c­ ombinations of price- and quantity-based tools. The empirical research has shown that mixed policy tools are better than single policy tools. If more attention is paid to gross domestic product (GDP) targets, regulation using primarily quantity-based tools supplemented by price-based tools (QP) is more effective; if more attention is paid to the level of inflation, then middle- and long-term regulation using primarily price-based tools supplemented by quantity-based tools (LPQ) is more effective; if the focus is on employment targets, short-term regulation using primarily price-based tools (SPQ) supplemented by quantity-based tools is more effective. Keywords  Price-based tools • Quantity-based tools • Mixed policy tools • Monetary policy framework

1   Introduction Before the outbreak of the international financial crisis in 2008, the monetary policy framework using primarily price-based tools supplemented by quantity-based tools has become prevalent in the world. However, the issue of liquidity shortage caused to countries by the outbreak of the crisis cannot be effectively dealt with simply with traditional price-based tools such as interest rate cuts. Thus, major central banks started to adjust their monetary policy frameworks and adopted comprehensive regulation models combining quantity- and price-based tools. When countries fell into the “liquidity trap”, they went further to adopt a portfolio of policy tools “placing price-based tools before quantity-based ones”. After the outbreak of the subprime crisis in August 2007, the Federal Reserve began adjusting its monetary policy framework. On the one hand, it shifted its ultimate goal from “price stability and economic growth” before the crisis to “full employment and price stability”; on the other, to match the new ultimate goal, it started transforming its approach to monetary policy regulation, using quantity- and price-based tools comprehensively according to specific circumstances in different periods for regulative purpose. From September 2007 to September 2008, it mainly employed price-based tools such as interest rate cuts (ten successive rate cuts), supplemented by innovative and quantity tools provided to the financial market, such as term auction facility (TAF). From October 2008 to November 2015, it transitioned to a combination consisting mainly of quantity-based tools supplemented by price-based ones. On the basis of basically constructing

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

115

an “interest rate corridor”, it adjusted the goal of policy interest rate from a specific value to an interval of 0–0.25%, which lasted seven years. The interval regulation of interest rate in fact had the effect of quantity-based tools. Moreover, it implemented in succession the generalized quantitative easing aimed at repairing the credit market and the quantitative easing monetary policy involving bulk asset purchase. Since December 2015, the Federal Reserve has gradually reverted to the regulation model using primarily price-based tools supplemented by quantity-based ones, but with more attention paid to quantity indicators in the macro-prudential framework such as general credit. On December 16, 2015, the Federal Reserve upped the target interest rate from 0–0.25% to 0.25–0.5% and, on December 21, officially published the countercyclical capital buffer (CcyB) framework, incorporating general credit targets into it so as to smooth credit growth, giving the framework some features of quantity-based regulation. Since its establishment, the European Central Bank has attached great importance to the combined use of price- and quantity-based tools. Before 2008, it employed mainly price-based tools, in coordination with quantity-­ based ones. From the crisis to 2011, it began to combine closely the two categories of tools. While greatly releasing liquidity with quantity-based tools, it employed price-based ones such as interest rate cuts to strengthen the regulation of short-term interest rates on the monetary market. The years 2008 and 2009 saw a total of eight interest rate cuts. The interest rate of main refinancing operations (MRO) was lowered from 4.25% to 1%. After 2011, while mainly employing quantity-based tools including very-long-term refinancing operations (VLTRO), reduction of deposit-­ reserve ratio, asset purchase programs and targeted long-term refinancing operations (TLTRO), it started monitoring a quantity indicator, namely credit-to-gross domestic product (GDP) gap of its member states. This measure has also been supplemented by price-based tools such as reducing MRO rates and negative deposit facility interest rates. It is noteworthy that some emerging economies have consistently stressed the combined use of price- and quantity-based tools. Brazil’s monetary policy framework has undergone multiple changes across different stages of economic development. The ultimate goal of its monetary policies has experienced relapses, from dual objective to single objective and then to dual objective again. However, as far as the use of policy tools is concerned, it has always emphasized the combination of quantity- and price-based tools. To curb high inflation, in the 18  months after July 1994, the Brazilian central bank had adjusted various deposit-reserve ratios 53 times and covered bank-issued bonds. After an interest rate

116  

X. YAN AND Y. ZHANG

corridor was basically constructed, it adjusted key policy interest rates for 28 times. In early 1999, less than half a year after the adoption of a totally freely floating exchange rate system, Brazil’s central bank began to implement inflation targeting in July 1999, which combined the interest rate corridor and completely freely floating exchange rates to give play to the dominant role of price-based tools. During this stage, it adjusted key policy interest rates for 63 times. The deposit-reserve ratio, mainly used to regulate liquidity, was adjusted for ten times only. Meanwhile, the capital provision mechanism was introduced and associated with the gap between credit growth and GDP growth as a major reference indicator for regulating domestic credit. Since October 2008, the interest rate corridor has been used as the main regulative means (changed 29 times), whereas the reserve as a supplementary means (adjusted six times). Besides, Brazil’s central bank also introduced macro-prudential tools such as countercyclical buffer capital, with attention paid to general credit. All this has shown that whether in abnormal or normal periods, whether the monetary policy framework is multi-purposed or single-purposed, countries or regions have, to varying extents, used policy tools, including price- and quantity-based tools, comprehensively. Particularly in the post-­ crisis era, central banks of countries and regions where price-based tools predominate have come to lay increasing emphasis upon quantitative indicators and introduce new macro-prudential indicators such as general credit into their monetary policies. In the meantime, the practice in some countries has revealed that when price-based tools are not so effective, quantity-based tools may be adopted in due advance. Interest rates do not have to near the zero lower limit, but the “quantity-price” tool set should be chosen according to the actual circumstances including the country’s financial structure and monetary transmission mechanism, ultimate goal of monetary policies and so on. A single target needs the combination of price- and quantity-based tools, and multiple targets need it more. Over a long period, China has adopted a monetary policy framework using primarily quantity-based tools supplemented by price-based tools, which has created a good monetary policy environment for rapid, sustained economic growth and accomplished the ultimate goal with fairly good effects. However, in recent years, as the market-oriented reform of the interest rate system is basically completed and monetary policy tools increasingly diversify, as financial innovation poses more challenges in statistical monitoring of money supply and final index association, there has been an increasingly strong call for the transition of China’s monetary

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

117

policy framework from quantity-based tools to price-based ones. Therefore, in the current period when China is experienced in the use of quantity-­ based tools and freshly moving toward price-based ones, particularly in the context that the market reform of the interest rate system is basically completed but an interest rate regulatory system has not been entirely built, it is of greater practical significance to study the monetary policy framework where price- and quantity-based tools support each other.

2   Literature Review Existing literature has seen studies exploring monetary policy tools from many perspectives. Theoretical studies are mostly based on partial or general equilibrium theory. Poole builds the LS-LM model to conclude that if the variance of money supply disturbance is small, quantity-based tools are more effective than price-based ones, and vice versa. Sargent and Wallace find that an uncertainty equilibrium if price-based tools are used for regulation, and a unitary equilibrium if quantity-based tools are used; and that the latter are more effective than price-based tools in macroeconomic regulation. Later, two classic rules on monetary policies are derived: first, the McCallum Rule (McCallum Rule 1984) for quantity-based tools; and second, the Taylor Rule (Taylor Rule 1993) for price-based tools. There have been lots of foreign empirical studies evolving around these two rules. Some scholars have introduced the two rules into the Dynamic Stochastic General Equilibrium (DSGE) model with microscopic foundation. Chinese studies on monetary policy tools have centered on the above-mentioned classic rules on monetary policies, but with considerably different conclusions. Some scholars believe that quantity-based tools are better than price-based ones. Hu thinks that quantity tools are more effective than interest rates in credit regulation; Liu et al. (2014) investigate the robustness of the two monetary policy tools in China under conditions of uncertainty by building a DSGE model. The results show that if the ultimate goal of monetary policies focuses on price stability, the two types of policy tools are not distinctly different in effectiveness; that if it focuses instead on output, then quantity-based tools are better than price-based ones. Some scholars believe that price-based tools can better convey information related to government behavior to the market, better control inflation and thus are better than quantity-based tools in macroeconomic regulation in China. A study by Zhang with the DSGE model leads to the finding that

118  

X. YAN AND Y. ZHANG

price-based tools are better than quantity-based tools in macroeconomic regulation in China; and when interest rates are more responsive to inflation, the range of economic fluctuation will be smaller. A study conducted by Wang et al. (2013) with the Structural Var (SVAR) and DSGE models shows that price-­based tools can more effectively regulate output and inflation and are less prone to incur economic fluctuation. Bian Zhicun and Hu (2015) find in a study employing the DSGE model that price-based tools are more powerful regulatory tools than quantity-based ones and can better flatten economic fluctuations. Only a few scholars hold that regulation combining quantity- and price-based tools is more suitable for China. Liu and Zhang think that the hybrid rule formed by incorporating money supply into the Taylor Rule generates better effect than a single quantity rule or interest rate rule. Yue and Niu (2014), by building a DSGE model, compare the McCallum Rule, the Taylor Rule and Hybrid Taylor Rule including money supply (i.e. the Hybrid Rule). The results show that the McCallum Rule generally explains China’s monetary policies over the past two decades. The Hybrid Rule incorporating monetary factors enjoys overwhelming advantage over single quantity rule or interest rate rule when used to interpret data. To sum up, it can be seen that there have been extensive studies on relevant questions of regulation with the two types of monetary policy tools from both theoretical and empirical perspectives. Especially in recent years, many scholars have incorporated the two types of monetary policy tools into the DSGE with micro foundation and derived some meaningful conclusions. However, these studies focus on which rule is better. Only in a few cases are hybrid rules considered (incorporating monetary factors into the Taylor Rule, forming a model dominated by price-based tools and supplemented by quantity-based tools). None of them think of the possibilities of a model dominated by price-based tools and supplemented by quantity-based ones or one dominated by price-based tools supplemented by quantity-based tools for long-term regulation. Besides, when proxy variables are chosen for price-based tools in existing literature, interest rate variables have been limited to interbank offered rate or benchmark deposit and lending rates. This chapter, however, chooses different central bank policy interest rates in different periods according to China’s actual ­situations, to better reflect the response of interest rates to major macroeconomic indicators. Therefore, this chapter, in the DSGE model with price stickiness, investigates respectively six monetary policy rules including regulation using only quantity-based tools (Q), regulation using primarily

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

119

quantity-based tools supplemented by price-based tools (QP), short-term regulation employing only price-based tools (SP), short-term regulation using primarily price-based tools supplemented by quantity-based tools (SPQ), long-term regulation using solely price-based tools (LP) and midand long-term regulation using primarily price-based tools supplemented by quantity-based tools (LPQ); compares their effects in a detailed way; draws conclusions accordingly and gives policy advice, with the hope to provide empirical reference for the transformation of China’s monetary policy regulation.

3   Model Building with Mutually Supportive Price- and Quantity-Based Tools in China Generally speaking, when monetary policy tools are introduced into the DSGE model, it is common to suppose that the central bank follows certain rules and introduces proxy variables. Current studies on China’s monetary policy tools can be classified into two types: the first type investigates price-based tools following the Taylor Rule, with the main difference lying in the choice of different proxy variables—for instance, adopting interbank offered rate and one-year deposit rate as proxy variables of price-­ based tools—and thus arrives at different conclusions; the other types study quantity-based tools following the McCallum Rule, with the main difference lying in the form of the McCallum Rule—in this case, proxy variables are, in most cases, money supply M2 or M1. Only a few scholars have added monetary factors on the basis of interest-rate rules, which in fact derive the Hybrid Rule using primarily price-based tools supplemented by quantity-based tools. 3.1  Theoretical and Practical Basis of Hybrid Rule for Monetary Policy Tools and Its Expression  heoretical and Practical Basis for Hybrid Rule for Monetary Policy T Tools According to the theory of macroeconomics, in an ideal world, when the central bank uses money supply as the main tool and the resulting changes of interest rates can bring about the equilibrium of the money market, then the equilibrium of the money market determines new interest rates, whose fluctuations change corporate marginal costs, which then impact

120  

X. YAN AND Y. ZHANG

macroeconomic indicators such as prices, investment and output, and the economy reaches a new equilibrium; when the central bank uses interest rates as the main tool, the adjustment of interest rates changes corporate marginal costs and impacts macroeconomic indicators such as prices, investment and output to reach a new equilibrium. Therefore, from the perspective of the simplicity of model building, the equation of money demand is in fact redundant; interest-rate rules alone can close the model. That is the main reason why most DSGE models use only interest-rate rules to describe monetary policies of the central bank. That is, under certain circumstances, when the targets of monetary policies are not different, price- and quantity-based tools have the same policy significance and can transform into each other, except that policies are different in effectiveness. This can be derived from the influence mechanism on the ultimate goal in the DSGE model. Take the impact of price- and quantitybased tools on the final GDP target in the DSGE model for example. It can be learned from the utility function of consumption, that is, Uc, t = βEt(Uc, t + 1 × It/πt + 1) that interest rate cuts can reduce the utility of consumption in the t + 1 period on the right of the equation; families will increase consumption in period t and reduce consumption in period t + 1, that is, advance consumption happens, which results in the increase of the current total demand. When the factor, money, is introduced, the abovementioned interest rate channel remains effective. Meanwhile, the optimum condition of currency holding Uc, t = Um, t + βEt(Uc, t + 1 × It/πt + 1) shows that increasing money supply will increase inflation, thus reducing the effect in the t + 1 period. As a result, families will increase consumption in period t and reduce consumption in period t + 1, that is, advance consumption happens, which results in the increase of the current total demand. When the two policy tools are used together, they are likely to augment the influence on the total demand. From this perspective, whether price tools, quantity tools or mixed policy tools will eventually influence the total demand. Therefore, they contain the same policy significance, though policy tools vary in effectiveness. Especially when the transmission channel of monetary policies is somewhat impeded, for example, financial markets of developing countries are not mature enough, under certain circumstances, financial markets of developed countries may malfunction or be subjected to the influence of the lower limit of interest rates, using a specific type of monetary policy tools alone cannot achieve the ultimate goal of monetary policies; and the two types of policy tools may generate more different results. So far, the overall environment of

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

121

China’s monetary policy regulation is rather complex, in face not only of an international environment where monetary policies of major developed economies diverge but also of a domestic environment of the economic New Normal tasked with “steadying growth and restructuring”. The market channel of transmitting monetary policies is yet to be improved. Some scholars thus construct the Monetary Policy Rule containing interest rates and money supply at once, termed in this chapter as the Mixed Monetary Policy Rule. Relevant research has supported this argument. Yao et al. find by building an open economy model containing capital flow disturbance that China also needs to construct a monetary policy framework with dual objectives and dual tools. Wu and Liu (2015) believe that influenced by factors such as economic restructuring and capital account convertibility, China needs to build a monetary policy framework with multiple objectives and employing multiple tools. Ostray et al. with IMF Research Department think that emerging economies have a weaker tolerance of currency mismatch and therefore need to construct a monetary policy framework with dual objectives and dual tools.  xpression of Hybrid Monetary Policy Tools E First, the Hybrid Rule employing primarily quantity-based tools supplemented by price-based tools. Traditional quantity rules have generally been expanded from the McCallum Rule. Their core content is: the monetary base growth rate changes to the deviation between nominal GDP growth rate (or level value) and the set objective. Some scholars find that the original McCallum Rule does not factor in prospective behaviors and thus cannot portray the situation in China. On the basis of the Quantity Rule with Prospective Behavior, this chapter introduces price-based policy tools and use interest rates, which have relatively great impact on money supply, as a proxy variable. Some empirical research has shown that the central bank bill issuance rate is an important factor influencing the amount of money supply in China (Zheng and Mo 2011). According to China’s actual circumstances, we have chosen different interest rates for different periods. For the period from the first quarter of 1996 to the first quarter of 2003, we use three-month deposit rate; for the period from the second quarter of 2004 to the second quarter of 2013, we use the central bank bill issuance rate; for the period from the third quarter of 2013 to the second quarter of 2015, we adopt the central bank reverse repurchase rate. The Hybrid Rule using primarily quantity-based tools supplemented by price-based tools which we create is as follows:

122  



X. YAN AND Y. ZHANG

vˆt = ρ1 vˆt −1 − ρ2 Et πˆ t +1 − ρ3 Yˆt − ρ 4 Rˆ t + ε M ,t

(5.1)



In the equation, vˆt stands for the deviation of the growth rate of money supply from its long-term trend; ≠ˆ t is the inflation rate; Yˆt is the output gap; Rˆ t is the market interest rate with a fixed term in different periods; and εM, t is process AR(1). In (5.1), if the variable, interest rate, is not considered, it can be simplified into a single quantitative rule. Second, the Hybrid Rule using primarily price tools supplemented by quantity tools. Considering the mode combining short-, middle- and long-term regulation in actual operation in recent years, if the short-term liquidity regulation tool (SLO), middle-term lending facility (MLF) and pledged supplementary lending (PSL) are respectively employed to regulate short- middle- and long-term liquidity in combination with quantity policy tools such as targeted RRR cuts, this chapter, in accordance with the Taylor Rule, constructs respectively two forms of rules with quantity tools: the first is Taylor Rule stressing short-term regulation; and the other, Taylor Rule stressing middle- and long-term regulation, with the factor of currency incorporated. Thus, two hybrid rules using primarily price tools supplemented by quantity tools are constructed. The expressions are respectively as follows:



 Rˆ t = λ1 Rˆ t −1 + (1 − λ1 ) λ2 ( Et πˆ t +1 − πˆ t ) + λ3 πˆ t + λ4 Yˆt + λ5 mˆ t  + ε R ,t    ε = ρ ε + η R ,t R R , t −1 R ,t 

(

(



)

 Rˆ t = λr Rˆ t −1 + (1 − λr ) πˆ t∗ + γ π πˆ t +1 − π t∗ + λy Yˆt + λm mˆ t      + γ dπ (πˆ t − πˆ t −1 ) + γ dy Yˆt − Yˆt −1 + η R ,t   π t∗ = ρπ t∗ + ηπ ,t  



(5.2)

)

(5.3)

In the equations, λ1 stands for the interest rate smoothing factor—the greater the value, the more consistent policies are; Et πˆ t +1 is the expected inflation rate, which shows the central bank’s prospective thinking of inflation; εR, t is the impact of monetary policies (or interest rates) and complies with the first-order auto-regressive process; π t∗ stands for the inflation target and complies with the first-order auto-regressive process, expressing the impact of monetary policies (interest rates); ηR, t stands for the impact of monetary policies (inflation target) and Yˆt Rˆ t is different from above.

123

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

The two interest-rate rules both stress the response of interest rates to inflation and output gap; the difference is that the former stresses the prospective thinking of the monetary authorities and shows that the central bank prioritizes short-term macro-control; the latter stresses the central bank’s consideration for inflation targets and shows that the central bank eyes middle- and long-term macro-control. Equations (5.2) and (5.3) will regress into corresponding single quantitative rules if variable mˆ t or amount of money supply is not factored in. 3.2  Construction of the DSGE Model with Price Stickiness Because the main purpose of this chapter is not to discuss the features and implications of the DSGE model, the general form of the DSGE model with Price Stickiness in the existing literature has been adopted. As for the hypothesis and economic meaning of the model and the process of its derivation, refer to the relevant literature (Zhuang and Cui 2013). This chapter directly sets forth first-order log-linear equations. For any variable Xt, this chapter makes Xˆ t = ln X t − ln X , which here shows the percentage deviation of variable Xt from its steady-state value. The log-linear equations are as follows: 1. Consumption Equation



1 1− b b ˆ  Rˆ − E πˆ + E εˆ Cˆ t = − εˆ  Ct −1 + Et Cˆ t +1 − 1+ b 1+ b (1 + b ) σ c  t t t +1 t c,t +1 c,t 



The variable with ^ stands for the percentage deviation from the steady-­ state value; Cˆ t stands for the consumption of the family in period t; b and σc are respectively elastic parameters for the development of family consumption habits and inter-temporal substitution of consumption; Rˆ t is the short-term interest rate in period t and Et πˆ t +1 is the family’s expectation in period t of the inflation in period t + 1. εˆc ,t is the impact of consumption preferences. 2. Investment Equation



Iˆt =

1 ˆ β ϕ ˆ 1  β Et εˆi ,t +1 − εˆi ,t  I t −1 + Et Iˆt +1 + Qt + 1+ β 1+ β 1+ β 1+ β 

β is the family’s subjective discount factor; Iˆt is investment in period t; φ is the costs of investment adjustment and εˆc ,t is the impact of investment.

124  

X. YAN AND Y. ZHANG

3. Tobin’s Q Equation 1−δ r Qˆ t =  Et πˆ t +1 − Rˆ t  + Et Qˆ t +1 + Et rˆt +1 + ηq ,t 1−δ + r 1−δ + r





Qˆ t is the asset price of period t, Et rˆt +1 is the expectation in period t for the marginal return on capital in period t + 1, and r is the steady-state return on capital. 4. Capital Accumulation Equation Kˆ t = (1 − δ ) Kˆ t −1 + δ Iˆt





Kˆ t is the capital deposit in period t; Iˆt is investment in period t; and δ is the depreciation rate of capital. 5. Resource Constraints C I G KY Yˆt = Cˆ t + Iˆt + Gt + uˆt Y Y Y Y



Yˆt is the total output in period t; C , I , G, K are respectively steady-state consumption, investment, government spending and capital stock. 6. Aggregate Production Function Yˆt = φ  Zˆ t + α Kˆ t −1 + (1 − α ) Lˆ t 



Yˆt is the total output of period t; ϕ is intermediate product manufacturers’ aggregate of production and the adjustment coefficient of the total output. 7. Wage Inflation Equation wˆ t =

1 + βγ w β 1 Et ( wˆ t +1 + πˆ t +1 ) − πˆ t + ( wˆ t −1 + γ wπˆ t −1 ) 1+ β 1+ β 1+ β (1 − ξ w ) (1 − βξ w ) σ   wˆ t − σ l Lˆ t − c Cˆ t − bCˆ t −1 + εˆl ,t − η w,t  −  1− b  σ (1 + λw )    (1 + β ) ξ w  1 + l  λw  

(

)

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

125

wˆ t is the actual wage of period t; σl stands for the replacement elasticity of consumption and labor; 1 + λw stands for the replacement elasticity of different labor; λw and ξw are, respectively, Calvo pricing and price adjustment probability parameter and iL is the family’s supply of labor in period t. 8. Labor Demand Lˆ t = rˆt + Kˆ t −1 − wˆ t



Lˆ t is labor demand in period t; wˆ t is the actual wage of period t; rˆt is the rate of return on capital and Kt − 1 stands for the capital stock in the previous period. 9. Price Inflation Equation (New Keynesian Phillips Curve, NKPC):

πˆ t =

γp 1 + βγ p +



πˆ t −1 +

β Et πˆ t +1 1 + βγ p

(1 − ξ ) (1 − βξ ) α rˆ + 1 − α wˆ ( ) (1 + βγ ) ξ  p

p

t

p

p

t

− Zˆ t + η p,t 

γp and ξp are respectively the Calvo index pricing of manufacturer pricing and price adjustment probability parameter.

4   Data Processing and Model-Based Estimation 4.1  Data, Parameter Calibration and Estimation The model is built with the following parameters: money supply (M), nominal interest rate (R), GDP, investment in fixed assets (I), consumption (C), consumer price index (P) and employment (H); and data are obtained for the time period from the first quarter of 1996 to the second quarter of 2015, mainly from WIND.  All data except those for interest rate and employment have been converted based on fixed-base CPI. Drawing reference from other Chinese scholars, we use year-on-year CPI growth to measure inflation. As for parameters to keep the model in a steady state, we do the calibration based on relevant data of the Chinese economy and existing studies home and abroad (Table 5.1). The two additional parameters are estimations

126  

X. YAN AND Y. ZHANG

Table 5.1  Calibration of basic parameters Parameter Value

β

a

δ

b

gy

σc

σm

σL

λ1

λ2

λ3

0.99

0.4

0.03

0.7

0.2

1.2

3.13

2.5

0.75

2.6

1

based on regression. Specifically, when adding interest rate into the quantity rule, we get ρ4 = 0.66; and when adding monetary factor into the interestrate rule, we get λ5 = λm = 0.07. For other parameters, we mainly refer to Zhang, Zhuang and Cui and so on. These parameters are estimated with the Bayesian method before which a function of prior distribution is constructed based on the theoretical meaning and range of each parameter as well as the conclusions of relevant studies from home and abroad. Specifically, for parameters ranging between 0 and 1, the prior distribution is set to be beta distribution; for those always greater than 0, a normal distribution is adopted; for the standard deviation of exogenous impact, invert gamma distribution is adopted. See Table 5.2 for details. 4.2  Comparison Between Six Combinations of  Monetary Policy Tools  ffectiveness of Different Regulation Measures Based on Pulse Response E In the DSGE model of price stickiness, we test six different combinations of monetary policy tools. Two combinations based on the basic McCallum Rule, namely a simply quantity-based regulation (Q) and quantity-based plus price-based regulation (QP). And the other four are combinations based on the basic Taylor Rule, namely simply price-based short-term regulation (SP), simply price-based long-term regulation, price-based plus quantity-based short-term regulation (SPQ) and price-based plus quantity-­based long-term regulation (LPQ). The effectiveness of different combinations of policy tools for regulation mainly depends on how much impact they have on major economic variables toward the ultimate goal set by the central bank and how much time it takes for the economy to get back to equilibrium. For example, when comparing the effectiveness of Q and QP, if the monetary policy becomes more influential on major economic variables after interest rate is taken into the picture, then QP is more effective than Q. Using DYNARE, we present estimations of the impact of the six different combinations of monetary policies on the ultimate goal of the central bank (see Figs. 5.1, 5.2 and 5.3). Figure 5.1 shows the impact on GDP. In

Price index Wage index Wage stickiness Price stickiness Adjusted investment cost Cost of capital utilization Technical impact coefficient Preference impact coefficient Investment impact coefficient Labor impact coefficient Government impact coefficient Policy smoothing coefficient 2 Inflation target coefficient Expected inflation coefficient Inflation coefficient

Parameter Q

QP

SP

Beta Beta Beta Beta Beta Beta Beta Beta Norm 1.00 2.60

ρz

ρc

ρi

ρl

ρg

λ1

ρv

ρπ

λ2/γπ

0.2

0.5

0.20 0.80

0.20 0.75

0.20 0.50

0.20 0.50

0.20 0.50

0.20 0.50

0.20 0.50

Norm 0.10 0.20

0.50 0.50 0.50 0.50 6.00

ψ

0.20 0.20 0.20 0.20 2.00

Beta Beta Beta Beta Norm

0.08 0.04 0.06 0.05 0.45

0.89 0.73 0.31 0.45 8.00

0.04 0.03 0.04 0.04 0.40

0.68 0.64 0.57 0.69 6.54

0.004 0.003 0.003 0.002 0.032

LPQ

0.94 0.45 0.08 0.70 8.50

0.02 0.01 0.02 0.01 0.16

0.64 0.75 0.73 0.56 7.12

0.04 0.02 0.02 0.02 0.51

0.85 0.35 0.30 0.82 6.07

0.02 0.04 0.03 0.01 0.53

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

– 0.82 0.06 0.18 0.04

–

0.72 0.01 0.75 0.03 0.62 0.03

(continued)

2.75 0.014 0.27 0.15 2.65 0.03 1.33 0.03

–

–

0.97 0.01

0.01 0.01 0.03 0.03 –

–

0.78 0.04 0.96 0.00 0.74 0.002 0.79 0.01 0.94 0.02 0.87 0.02

0.83 0.03 0.84 0.03 0.82 0.004 0.84 0.02 0.75 0.03 0.72 0.02

0.25 0.04 0.12 0.02 0.98 0.003 0.99 0.00 0.68 0.02 0.91 0.03

0.84 0.04 0.95 0.01 0.78 0.009 0.48 0.01 0.82 0.03 0.92 0.04

0.13 0.04 0.98 0.01 0.92 0.006 0.74 0.01 0.54 0.03 0.99 0.01

0.01 0.01 0.01 0.01 0.27 0.001 0.14 0.00 0.03 0.02 0.19 0.01

0.82 0.81 0.50 0.52 7.75

LP

Mode STD Mode STD Mode STD

SPQ

Posterior distribution

STD Average Mode STD Mode STD Mode STD

γp γw ζw ζP φ

Type

Prior distribution

Table 5.2  Prior distribution and posterior distribution of modified models

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

127

Q

QP

SP

Norm 0.10 0.10 Inv. g Inf. Inv. g Inf. Inv. g Inf. Inv. g Inv. g Inv. g Inv. g Inv. g Inf.

γY

σπ

σz

σc

σi σl σp σg

σm/σr

Inf. Inf. Inf. Inf.

Norm 0.20 0.60 Norm 0.20 0.63 Norm 0.20 0.30

λ4/ρ3 λ5/ρ4 γdπ

3

1 5 1 3

1

3

14

Norm 1.00 3.00

λ3

0.2

Norm 1

LPQ

–

– –

– –

– –

– –

–

0.07 0.64 0.03 0.05

0.76 2.74 0.30 0.47

0.08 0.62 0.03 0.04

0.49 2.96 0.29 0.55

0.05 0.05 0.02 0.03

0.57 0.05 0.42 0.04 0.12 0.00

0.52 2.59 0.30 0.44

0.31 0.03 0.38 0.08 0.66 0.06

0.36 0.02 0.35 0.02 0.35 0.02

–

–

–

–

0.82 0.06 0.18 0.04

0.06 0.01 0.18 0.00

0.12 0.28 0.07 0.04

0.46 4.62 0.47 0.12

0.17 0.28 0.04 0.00

0.30 4.28 0.48 0.12

0.12 0.23 0.04 0.00 0.01 0.00 0.25 0.26 0.26 0.02

3.14 4.35 0.56 0.52

0.33 0.11 0.42 0.08 0.35 0.03

0.39 0.04 0.35 0.02 0.35 0.02

–

–

0.98 0.03 0.14 0.02 0.52 0.004 0.13 0.01 0.17 0.01 0.03 0.01 – – 0.74 0.03 – – 0.66 0.02 – – 0.28 0.02 – – – – – – – – 0.01 0.02 0.07 0.01

1.45 0.07 1.08 0.03 3.15 0.009 1.28 0.06 1.65 0.03 1.33 0.03

1.45 0.07 1.08 0.03

LP

Mode STD Mode STD Mode STD

SPQ

Posterior distribution

STD Average Mode STD Mode STD Mode STD

ρ2

Type

Prior distribution

Notes: (1) STD means standard deviation; (2) when an equation follows the quantity rule and the price rule, the policy smoothing coefficient is respectively the smoothing coefficient for money supply and interest rate, while the policy response coefficient represents the response with respectively the interest rate and the monetary factor taken into account; (3) STD value stands for the degree of freedom for invert gamma distributions

STD of price impact STD of government impact STD of policy impact

Output gap coefficient Policy response coefficient 3 Current period inflation coefficient Current period output coefficient STD of inflation target STD of technical impact STD of preference impact STD of investment impact STD of labor impact

Parameter

Table 5.2 (continued)

128   X. YAN AND Y. ZHANG

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

129

Fig. 5.1  Impact of the six combinations of policy tools on GDP

Fig. 5.2  Impact of the six combinations on inflation

comparison, mixed measures are more effective than single ones. Specifically, with the QP model, every one percentage point of increase in money supply pushes GDP growth up by 0.27 percentage point, 0.05 percentage point higher than with the Q model. With the LPQ model, every one percentage point of decrease in interest rate pushes GDP growth up by 0.28 percentage point, 0.12 percentage point higher than with the LP model. And with the SPQ model, the impact of adjustment to interest rate is also slightly stronger than with the SP model. Meanwhile, combinations mainly based on price are generally more effective than mainly

130  

X. YAN AND Y. ZHANG

Fig. 5.3  Impact of the six combinations on employment

quantity-­based combinations. The SP model sees the GDP growth up by 0.25 percentage point for every one percentage point of decrease in interest rate, higher than the Q model by 0.03 percentage point. Among all mixed models, LPQ has the strongest impact on GDP growth, followed successively by QP and SPQ.  In terms of the time it takes to return to equilibrium, mixed short-term models are the best. SPQ needs three quarters, and QP four quarters, one quarter shorter than the corresponding single-tool models. However, for the mid-to-long term, LPQ needs ten quarters, while LP requires only four. Figure 5.2 shows the impact of the six combinations on inflation. As we can see, they do not differ much. When the monetary policy changes by one percentage point, inflation changes by 0.01–0.06 percentage point. Figure 5.3 shows the impact of the six combinations on employment and the results are similar to the impact on GDP growth with difference seen in the following: QP is the most effective in that it sees employment growth up by 0.36 percentage point with every one point of increase in money supply, much higher than any other model.  ffectiveness of Regulations Based on Impulse Response E To have a better understanding of how much each type of monetary policy contributes to the overall effect when major economic variables are subject to the impact of multiple factors, we carry out variance decomposition on major economic variables and do some comparison (see Tables 5.3, 5.4 and 5.5).

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

131

Table 5.3  GDP variance decomposition in different models %

Consumption impact Investment impact Labor supply impact Government purchase impact Price markup impact Technical impact Monetary impact Interest rate impact Inflation expectation impact

Q

QP

SP

SPQ

LP

LPQ

0.27 37.46 0.74 37.54 0.92 3.7 19.37 – –

0.09 22.13 0.59 33.97 0.82 18.86 23.53 – –

0.01 17.53 1.19 15.81 7.25 47.25 – 10.95 –

0.01 0.18 0.58 8.82 11.34 70.77 – 8.29 –

0.04 0.43 0.88 59.35 8.26 3.45 – 7.74 19.85

0.02 0.43 0.99 67.62 6.59 12.6 – 10.42 1.33

Table 5.4  Variance decomposition for inflation in different models (%)

Consumption impact Investment impact Labor supply impact Government purchase impact Price markup impact Technical impact Monetary impact Interest rate impact Inflation expectation impact

Q

QP

SP

SPQ

LP

LPQ

0.11 13.78 0.92 2.82 16.16 61.8 4.41 – –

0.13 13.32 0.6 9.29 15.95 54.81 5.89 – –

0.03 9.51 0.57 10.4 23.6 53.07 – 2.81 –

0.01 0.35 0.51 10.78 16 57.7 – 14.66 –

0.08 0.82 2.62 54.78 3.08 16.41 – 10.84 11.37

0.04 0.88 2.86 41.87 6.93 15.48 – 28.38 3.56

Table 5.5  Variance decomposition for employment in different models (%)

Consumption impact Investment impact Labor supply impact Government purchase impact Price markup impact Technical impact Monetary impact Interest rate impact Inflation expectation impact

Q

QP

SP

SPQ

LP

LPQ

0.09 12.51 1.78 15.18 0.35 61.55 8.54 – –

0.04 13.27 1.88 25.09 0.64 39.24 19.84 – –

0.01 36.33 2.98 7.59 12.59 18.16 – 22.33 –

0.02 0.66 2.03 43.95 8.42 15.75 – 29.17 –

0.14 0.84 9.6 11.35 10.61 9.18 – 16.76 41.52

0.05 0.97 5.54 26.56 3.56 34.4 – 25.47 3.45

132  

X. YAN AND Y. ZHANG

The results for GDP (Table 5.3) show that monetary policies generally contribute more in mixed models. Specifically, in the QP model, money supply explains 23.53% of GDP fluctuations, higher by 4.16 percentage points than in the Q model; in the LPQ model, interest rate explains 10.42% of GDP fluctuations, 2.68 percentage points higher than in the LP model. Also, quantity-based models see a higher contribution from monetary policy than price-based models. For example, in the Q model, money supply explains 19.37% of GDP fluctuations, higher than in the SP, SPQ, LP and LPQ models by respectively 8.42, 11.08, 11.63 and 8.95 percentage points. Meanwhile, in all mixed models, the contribution of technical impact to GDP fluctuations is higher. In the SPQ model, changes in technology (total factor productivity) make a contribution of 70.77% to GDP, 23.52 percentage points higher than in the SP model. The contribution also grows sharply in the other two mixed models. Therefore, it is fair to say that compared with single-tool models, mixed models do better in promoting the contribution of the supply-side reform to GDP. According to the results of variance decomposition for the impact of inflation (Table 5.4), monetary policies have greater explaining power in mixed regulation models. This shows that combinations of quantity-based and price-based tools are more helpful for the realization of the goals set for inflation control. Specifically, in the QP model, the contribution of monetary impact is 1.48 percentage points higher than in the Q model; in the SPQ model, the impact of interest rate is 5.2 times that in the SP model; and interest rate changes are the most effective in the LQP model, making a contribution of 28.38%, much higher than in any other model. Therefore, if the central bank can bring inflation expectation and monetary factors into the picture while adopting the Taylor Rule, it will be able to better realize its mid- and long-term goal of inflation control by imposing mainly interest rate-based regulation measures. Variance decomposition of the impact on employment (Table  5.5) shows that in mixed regulation models, monetary policies have greater explaining power for employment conditions. This means that by combining quantity-based and price-based tools, we can do a better job in attaining our employment goals. Specifically, in the QP model, the contribution of money supply changes to employment is 2.3 times that in the Q model; in the SPQ model, interest rate makes a contribution of 29.17% to employment, 8.84 percentage points higher than in the SP model. However, it is worth noting here that in both the QP and the SPQ model, technical impact (supply-side reform) makes smaller contribution.

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

133

5   Conclusions and Policy Recommendations This study examined some mutually supportive price-based and quantity-­ based monetary policy tools adopted by the central bank before and after the crisis, built six different combinations of such policy tools on the basis of the Taylor Rule and the McCallum Rule and put them into the DSGE model. The empirical results show that for China, the different monetary policy tools and their combinations do differ greatly in how effective they are for the attainment of our ultimate goals. First of all, if the central bank focuses its regulation more on GDP-related goals, the QP model is a better choice. Second, if the goal of regulation focuses more on inflation, it is better to take the LPQ model. Third, if the focus is more on employment, SPQ is a better choice. In this model, with monetary factors brought into the picture, interest rate changes have a greater explaining power, but the explaining power of technical impact (supply-side reform) is reduced. In comparison, the LPQ model, which takes into consideration both the inflation expectation and monetary factors, allows both interest rate and technical impact (supply-side reform) to explain a much larger share of employment conditions. Therefore, we believe SPQ is more effective than LPQ in terms of the explaining power of interest rate for employment conditions. The reason here may be as follows: In China, currently in the early stage of building a market-based interest rate system, market players on the microscopic level are not very sensitive to interest rate changes and the impact of monetary policies is exerted to a large extent through bank loans. However, we need to do more to improve the transmission mechanism from a key interest rate policy issued by the central bank to benchmark interest rate in the market. Therefore, we still need more time to move from a mainly quantity-based regulation model to a mainly price-­ based one. Considering the nature of the DSGE model, the transmission of the impact toward the ultimate goal may be enhanced when some price-­ based policy tools are adopted in addition to quantity-based ones and thus, in a transitional period, monetary policies may be more effective if we apply different combinations of quantity- and price-based policy tools. These conclusions prove to us that the adoption of different monetary policies must be based firmly on specific national conditions and targeted precisely at the ultimate goal of the regulation. For this purpose, we hereby raise the following policy recommendations. First, during the transitional period before complete financial liberation is achieved by completing the reforms of interest rate and foreign exchange

134  

X. YAN AND Y. ZHANG

systems, the transformation of our monetary policy is not a simple shift of the focus from quantity-based to price-based tools. Instead, we should keep an eye on the progress toward fully market-based interest rate system and be careful about the timing for the adoption of price-based regulation. The QP model should be applied to facilitate the forming of a better interest rate regulation system. According to what the Chinese central bank says about its ultimate goals, the focus is more on economic growth and based on what we have found out in this empirical study, we believe that in the early stage of market-based interest rate system, we are not yet ready for a price-based regulation model and quantity-based measures are still needed to play their role during the transitional period. In the meantime, the central bank should move faster toward a set key policy interest rate and a clearly defined rather than implicit interest rate corridor. Second, after financial liberation is fully achieved, with a better interest rate regulation system in place, our monetary policy should shift to model dominated by price-based tools and supplemented by quantity-based ones. Which specific model is the best choice depends on what our ultimate goal is. As the decrease in the potential economic growth rate slows down, the central bank may shift its goal in the future. If employment becomes the top priority, the SPQ model will be the best choice; if inflation is the key, the LPQ model shall be picked, taking into consideration not only quantity-based indicators such as money supply, credit and reserves, but also various expectations; if the focus is to balance between multiple goals, one of the mixed models should be the choice, supported by appropriate structural monetary or public financial policies. Third, we should combine expectation management better with quantity- and price-based tools. According to international experience, in normal circumstances, expectation management promotes the effective transmission of monetary policies and its promotional power may be even stronger in some special circumstances. For example, during crises, the Federal Reserve and the central banks of Europe and Japan all strengthened expectation management continuously while applying in combination quantity- and price-based tools, greatly enhancing the effective transmission of monetary policy tools. In the meantime, some central banks, including the Federal Reserve and the European Central Bank, also paid more attention to expectation management for CcyB, broad credit/ GDP or its gap. The empirical results of this study also show that when inflation expectation is added into the picture, the explaining power of various factors for economic growth, employment and inflation all rose.

  A STUDY OF THE MONETARY POLICY FRAMEWORK WITH MUTUALLY…   

135

Therefore, for China, regardless of what stage we are in, it is always advisable to improve the coordination and combination between expectation management and quantity- and price-based policy tools. When applying a monetary policy, we must take timely measures to explain to the general public our policy framework, how our specific measures will take effect and how to make flexible use of these policies as the prospect changes. This will help make our regulation more effective.

References Bian Z. and Hu H.  Selection of Monetary Policy Tools in China: Quantity or Price? A DSGE-based Analysis. Studies of International Finance, 2015 (6). Liu X., Li L. and Gao M.  Comparative Study on the Robustness of China’s Monetary Policy Tool Rules under Uncertain Conditions. Studies of International Finance, 2014 (7). Wang J., Guan X. and Wang Y.  Tool Selection, Macro Effectiveness, and Rule Design for China’s Monetary Policies. Journal of Financial Research, 2013 (8). Wu G. and Liu K. Explore the Rule System of China’s Monetary Policies: Multiple Goals and Multiple Tools. Studies of International Finance, 2015 (1). Yue C. and Niu L.  Estimation and Comparison of China’s Monetary Policies. Journal of Quantitative and Technical Economics, 2014 (3). Zheng Z. and Mo T.  Can Policy Interest Rate Guide Market Interest Rate: A Study on the Bilateral Interaction between the Policy Interest Rate and Market Interest Rate of Central Bank Bills. Finance and Trade Economics, 2011. Zhuang Z. and Cui X.  Instability, Macroeconomic Fluctuations, and Monetary Policy Selection. Working paper of Peking University, 2013.

CHAPTER 6

Does the Upgrading of Industrial Structure Mean Service Sector Is More Important? The Influence of the Interactive Development of Manufacturing and Service Sectors on China’s Economic Growth Shenning Qu and Tie Lyu

Abstract  With China’s service industry gradually overtaking manufacturing to become the largest sector of the economy, there are growing calls for stepped-up efforts to develop this “pillar industry.” However, this view is largely based on the traditional economic thought on industrial structure. In the context of global industrial integration today, the view is debatable. The American strategy of “re-industrialization” indicates that manufacturing in the national economy plays a role that is not replaceable by the service sector. We construct in this chapter a dynamic stochastic general equilibrium (DSGE) model for the manufacturing and service sectors, and analyze from the perspective of the two-sector integration, the ­economic impact of the CLC No.: F062.9 Document Code: A Paper No.: 1002-8102(2016)03_0138_10 S. Qu (*) • T. Lyu Institute of Industrial Economics of Chinese Academy of Social Sciences, Beijing, China © The Author(s) 2019 D. He, C. Wang (eds.), A New Era, https://doi.org/10.1007/978-981-10-8357-0_6

137

138  

S. QU AND T. LYU

transformation of industrial structure. A numerical simulation shows that in terms of the potential of economic growth, as the contribution of the service sector to GDP grows, technological innovations in manufacturing will have a reduced impact on GDP growth, while the impact of innovations in the service sector will not change much. The service sector acts as an “economic stabilizer,” but the externality of its technological progress on the overall economic system is weaker than that of the manufacturing sector. A balance needs to be achieved in developing the manufacturing and service sectors in the future process of “upgrading the industrial structure.” Keywords  Industrial integration • Re-industrialization • Two-sector DSGE

Industrial

structure

•

1   Introduction After the launch of the reform and opening up, China’s manufacturing and service sectors have shown rapid growth momentum, but their growth rates varied significantly in different periods. Between 1987 and 1998, the added value of manufacturing grew at an annual rate of 3.6 percentage points higher than that of the service sector. Between 1999 and 2013, the average annual growth rates of the two sectors became quite similar, both surpassing the growth rate of GDP. The service sector, however, showed signs of faster growth during the period. Thanks to the government’s industrialization strategy, China’s manufacturing had grown faster than the service sector for a long period of time. The factors of production had been skewed to manufacturing, while the fiscal expenditure had been mainly used in the construction of industrial infrastructure. The service sector, which is closely related to people’s livelihood, had been poorly funded. The development of the two sectors had been grossly unbalanced, causing to some extent the crowding out effect in both resources and factors of production. With industrialization advanced, the growth rate of manufacturing began to slow down, while the service industry gathered momentum, and by 2013, it had surpassed manufacturing to become China’s largest industry. In this context, a growing number of people call for efforts to further expedite the development of the service sector, believing as a “pillar industry,” it is more important than manufacturing (He and Yao 2008).

  DOES THE UPGRADING OF INDUSTRIAL STRUCTURE MEAN SERVICE…   

139

The dynamic change in the proportions of the two sectors is also a process of adjustment and evolution in the industrial structure. In general, when an economy grows, its industrial structure will also change accordingly. Economic growth and structural change are complementary to each other and affect each other. In the process of economic growth, the change of industrial structure mainly manifests in two aspects. First, due to their huge differences in technological requirements and absorptive capacity, the pace of technological advancement varies in different sectors, so does the growth rate in the output produced by each of them. The country’s industrial structure is thus adjusted dynamically. Second, in different stages of development, a country needs different pillar industries to support economic growth. With the transformation of national strategies for economic development, new industrial policies will be developed to guide the replacement of pillar industries, which will have a profound impact on the production and consumption in various supporting industries, and fundamentally change the industrial structure (Kuznets 1971). Evidently, the change of industrial structure is not only related to the characteristics of an industry itself, but also to the policy orientation of the country. When a country, through national policies, stimulates the development of industries that are deemed strategically important and capable of fueling the growth of the entire economy, a profound impact will be produced on the industrial structure adjustment. The belief that the service sector is more important than manufacturing is largely based on what the traditional school of industrial structure has advocated. It argues that the evolution of industrial structure is a process of economic growth by the absorption of technological innovation and the subsequent transformation of the dominant sectors (Rostow 1962). A country’s industrial structure has experienced a shift, from agriculture-­oriented, to industry-oriented, and then to service-oriented. In this process, the factors of production gradually flow from the lowproductivity sector to the high-productivity sector; thus, the productivity of the whole society rises. The productivity dividend generated by this structural change has been the driving force behind economic growth, which is the main reason for the shift in industrial structure to promote economic growth. Therefore, the traditional school of industrial structure believed that a country should first focus on the development of highly productive industries, making them engines of economic growth (Chenery 1989). In China, the rise in the service sector in recent years has also led some scholars to believe that it is the main driver of economic growth

140  

S. QU AND T. LYU

(Pang and Deng 2014; Yang and Tan 2014). However, they usually assume that the various sectors are independent of each other, while ignoring the interaction between them and the spillover effect. Given the trend of industrial integration in the world today, such view is debatable. In recent years, the developed countries, where the service sector has long been a mainstay of the economy, have begun to place weight on manufacturing once again. In the United States, for example, in 2012, the government launched the strategy of “re-industrialization” to revive the “Made in America” Movement. By fostering awareness and launching a “Select America” program, the American and foreign manufacturers are encouraged to rethink “Made in America.” However, based on the trend of industrial structure development, manufacturing in the United States has been declining. In 1950, manufacturing accounted for 27% of US gross domestic product (GDP), which accounted for 31% of total US jobs, but by 2010, the two ratios fell to 12% and 9%, respectively. Evidently, the reason why the United States wanted to revive its manufacturing industry is not due to its contribution to GDP, but the fact that high-tech manufacturing, accounting for 70% of the country’s R&D activities, has a great spillover effect and is capable of stimulating the productivity gains in the service sector and the overall economy. Note that the idea is about productivity growth in the entire economy rather than in one sector alone. From the perspective of industrial integration, the manufacturing industry has a significant positive externality and is critical to the innovation and competitiveness of a country. The painful lessons of the international financial crisis have proved that it is dangerous for people to think a country without manufacturing can become a global powerhouse in innovation as well and that the United States should abandon the “de-­industrialization” test as prompt as possible. America’s “re-industrialization” strategy raises the alarm for people who emphasize the overriding importance of the service sector in our country. Manufacturing in the national economy plays a role that can hardly be replaced by the service sector. Manufacturing can not only aggregate R&D and innovation gains and create real economic needs but also provide the jobs needed for the producer services. In the current industrial structure, although the service sector gradually becomes dominant, it does not mean that the importance of manufacturing is reduced. Manufacturing has a significant externality for the service sector, and the scale economy brought by this kind of industrial integration cannot be neglected. Some Chinese scholars have discovered this. Huang et  al.

  DOES THE UPGRADING OF INDUSTRIAL STRUCTURE MEAN SERVICE…   

141

(2013) argue that the reason for the United States’ “re-industrialization” is that “de-industrialization” restricted the development of advanced manufacturing that depends on advanced producer services, and in turn made it difficult for advanced manufacturing to fuel the development of the service sector; this thus caused a slowdown in economic growth of the United States. Yuan and Gao (2015) find that every 1% increase in manufacturing jobs in cities would lead to about 0.435% more jobs in services, particularly consumer services. The increase in manufacturing jobs will contribute to the growth of employment in the service sector. Li (2013) also believes that in view of the huge externalities of manufacturing, the driving force of China’s economic growth until 2020 is still the secondary sector, and it is difficult for the tertiary sector to become the driving force. During the third session of the 12th National People’s Congress in 2015, Premier Li Keqiang made clear the need to upgrade industrial structure. According to the Premier, “manufacturing is our dominant industry,” and the main way to upgrade the industrial structure is “to promote the deep integration of industrialization and informatization, formulate the ‘Internet+’ action plans, as well as facilitate the integration of mobile Internet, cloud computing, big data, and Internet of things with modern manufacturing.” Therefore, in the future, the integration of the manufacturing and service sectors will become the key of “maintaining growth” and “adjusting structure” of the economy under “new normal.” In this case, the study of the integration of the two sectors will have significant practical value for the debate on the importance of the manufacturing and service sectors and for the exploration of industrial structure upgrading. With the updating of the research methods, more and more scholars began to use the structural econometric modeling and dynamic general equilibrium modeling to explore, from the dynamic and long-term perspective, the macroeconomic impact of the transformation and upgrading of industrial structure, which is represented by dynamic adjustment between the manufacturing and service sectors. Ngai and Pissarides considered, through theoretical analysis, that a country’s structural transformation has a significant impact on economic growth rate. Herrendorf and Valentinyi constructed a dynamic stochastic general equilibrium (DSGE) model, and found that structural transformation had an effect on total factor productivity (TFP). Carvalho studied at the industrial level and pointed out that a country’s economic structure has a greater impact on the output fluctuations of various sectors. Because of the particularity of China’s industrial structure and economic devel-

142  

S. QU AND T. LYU

opment, the evolution course in China was not consistent with foreign economic rules and practices, and needs to be measured and analyzed according to the actual situations in China. Sachs argued that the successful transformation of China’s industrial structure is the main reason for its rapid economic growth. Fan further pointed out that the dynamic changes in China’s industrial structure accelerated the mobility of labor force among sectors and regions and increased TFP. However, Lyu (2002), by measuring the change of TFP of manufacturing in the process of industrial structure transformation, found that the transformation of industrial structure had limited the TFP, and the “structural dividend” was not significant. Liu and Zhang (2008) supported this statement, pointing out that over time, this marginal benefits brought about by “structural dividend” would gradually reduce, but they affirmed that the industrial structure transformation was the driving force behind economic growth. Gan et al. had carried out econometric tests on the correlation between China’s economic growth and industrial structure, and found both the rationalization of the industrial structure and the industrial upgrading would produce an impact on economic growth, except that the relationship between structural rationalization and economic growth is more stable while that between structural upgrading and economic growth is less so. In the studies of industrial structure mentioned above, the effect of industrial integration on industrial structure had been neglected, and the impact of the externalities of the two sectors on economy has not been examined. At the same time, the studies of the impact of the adjustment of industrial structure largely focus on economic growth, and the impact of economic fluctuation is obviously inadequately researched. Therefore, in this chapter, we seek to make up for these deficiencies in the existing research, from the following several aspects: (1) To construct a two-sector DSGE model of endogenous industrial integration, examine the intermediate input of the manufacturing and service sectors, and analyze the influence of industrial structure transformation on economy from the perspective of the integration of the manufacturing and service sectors. (2) In the study of the impact of industrial restructuring on economic growth, we will delve into the economic fluctuation and the characteristics of the volatility of China’s GDP growth rate in the process of industrial upgrading in recent years. What growth rate should China’s manufacturing and service sectors maintain, in order to be more conducive to the goals of “maintaining

  DOES THE UPGRADING OF INDUSTRIAL STRUCTURE MEAN SERVICE…   

143

growth” and “adjusting structure,” and promote the upgrading of the industrial structure? This article attempts to answer this question and, from the perspective of industrial integration and industrial transformation, analyze the different roles of the manufacturing and service sectors in China’s long-term economic growth. The main structure of this chapter is as follows: The second part is on the basic model; the third part, the data processing; the fourth part, the result of modeling; and finally, the conclusion and the policy implication.

2   Basic Model We built a two-sector DSGE model that includes the integration of the manufacturing and service sectors. The model consisted of three parts: the production sector, the household sector, and the market equilibrium. 2.1  Household Sector Assume that the economic system is composed of countless households, evenly distributed between [0,1]. Each household in the economic system has a preference between the two sectors, and each period has a unit of labor endowment, then the optimized utility function is: ∞



max U = E0 ∑β ′  log ( ct ) + ϕ log (1 − nt )  i =0

(6.1)

where E is the expected operator and β is the discount factor. 1− ρ ct = cmρ ,t ( cs ,t + s ) , cm is per capita industrial product consumption level. cm is per capita service consumption level. n is labor. parameter s̅ is home production services, which means that the elasticity of service consumption should be greater than 1, and other parameters meet s > 0, ρ 1, ϕ 0 . In addition to consumption, households also make investment decisions. We drew on the setting in Kongsamut et al., assuming that during the time t, a household inputs It industrial products into new capital v

1  It  kt , and kt satisfies kt =   kt , where kt the initial capital accumulation ω  kt  during the t period, with the parameters ω > 0 and 0