From trade surplus to the dispute over the exchange rate quantitative analysis of RMB appreciation

• The repeated calculation of the foreign value-added return of exports and the domestic value-added return of imports in pro-cessing trade using the customs statistical method, accordin

the Exchange Rate

Quantitative Analysis

of RMB Appreciation

This page intentionally left blank

NEW JERSEY • LONDON • SINGAPORE • BEIJING • SHANGHAI • HONG KONG • TAIPEI • CHENNAI • TOKYO

Beijing Normal University, China

Library of Congress Cataloging-in-Publication Data

Names: Xin, Li, 1982– author | Dianqing, Xu, 1945– author.

Title: From trade surplus to the dispute over the exchange rate :

quantitative analysis of RMB appreciation / Xin Li (Beijing Normal University, China),

Dianqing Xu (Beijing Normal University, China).

Description: New Jersey : World Scientific, 2016.

Identifiers: LCCN 2015040538 | ISBN 9789814723954 (alk paper)

Subjects: LCSH: Balance of trade China | Foreign exchange rates China |

Currency question China | Monetary policy China | China Economic

policy 21st century | China Foreign economic relations.

Classification: LCC HF1014 X56 2016 | DDC 382/.170951 dc23

LC record available at http://lccn.loc.gov/2015040538

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Copyright © 2016 by World Scientific Publishing Co Pte Ltd

All rights reserved This book, or parts thereof, may not be reproduced in any form or by any means,

electronic or mechanical, including photocopying, recording or any information storage and retrieval

system now known or to be invented, without written permission from the publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance

Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA In this case permission to photocopy

is not required from the publisher.

In-house Editors: Suraj Kumar/Lixi Dong

Typeset by Stallion Press

Email: enquiries@stallionpress.com

Printed in Singapore

This book is the fruit of the cooperation between Chinese and foreign

scholars in various aspects

Since 2005, some Americans have been accusing China of

caus-ing the trade deficit, with one wave surgcaus-ing after another Some

well-known scholars have also accused China of manipulating the

exchange rate, arousing a considerable explosion of international

public opinion At the same time, many domestic scholars argue

against the RMB appreciation The two sides stand by their own

views and are passionate over their respective arguments The

rea-son for this situation is the lack of quantitative research or the lack

of elaboration of many extremely important indicators Some people

even hastily give their opinions without understanding the basic

def-initions and connotations of the key data In its takeoff stage, China

had several characteristics that other countries did not have

There-fore, committing the mistake of notching the boat to find the sword

is inevitable when applying the traditional trade theories Clearly,

providing a platform and enabling the two sides to exchange views

calmly is difficult without quantitative research; both sides could go

to extremes during the debate It is either this or that without giving

a convincing conclusion To be quantitative is to compromise; this is

the only way to find a win–win solution in international trade

In the summer of 2009, Professor Yao Yang, the Dean of the

National School of Development, Peking University, proposed to

conduct a quantitative research on the exchange rate of RMB and

Sino-US trade Undoubtedly, this subject is very important and

chal-lenging To conduct such a large-scale quantitative research, critical

technical difficulties must be resolved in advance

v

vi Preface

Conducting a quantitative research of the exchange rate of RMB

requires establishing a Computable General Equilibrium (CGE)

model with China as the core This model should include not only the

main export markets of China, such as the United States and Europe,

but also its real and potential competitors, that is, the Asian

eco-nomic giants such as Vietnam, Thailand, and India Furthermore, the

model must consider data not only on import and export trade but

also those on labor employment, income, output, and price of

differ-ent departmdiffer-ents Therefore, establishing such a large-scale database

is time-consuming and formidable, and dealing with a large amount

of data is difficult

In recent years, the proportion of processing trade in import and

export in China has been increasing, accounting for nearly half of

the foreign trade The customs statistics inevitably contains a large

amount of repeated calculation Moreover, a considerable part of the

property rights in the export commodities belongs to foreign

invest-ment The customs statistics objectively exaggerates the export

sur-plus of China, as it only contains logistics data through the said

agency and does not present the property ownership of the export

commodities All these are well-known facts, but establishing their

quantitative description is very difficult

Fortunately, after years of efforts, Prof Chen Xikang of the

Chi-nese Academy of Sciences and his team have developed a set of

non-competitive input–output tables that can effectively decompose the

data on import and export processing trade, which is a necessary

link in analyzing the current foreign trade of China Dr Wang Zhi

of the U.S International Trade Commission and Prof Wei Shangjin

of Columbia University have developed a set of models for

analyz-ing the value-added in processanalyz-ing trade Their contributions lay the

foundation for the present research

Organized by Prof Chen Xikang, Dr Wang Zhi, and Prof Yao

Yang and supported by the Major Programs of the

Fundamen-tal Research Funds for the Central Universities (#SKZZX2013009),

the National School of Development at Peking University, the

Sys-tem Science Institute of China Science Academy, the Department

of National Accounts National Bureau of Statistics of China, and

Preface vii

the U.S International Trade Commission have made a concerted

effort for three years (i.e., since 2009) to establish the continuous

global input–output database, including the time series data from

1995 to 2007 The input–output data distinguish between processing

and non-processing trades and ensure the endogenous balance of the

Social Accounting Matrix (SAM) after separating processing trade

from non-processing trade They realize a complete sector

classifi-cation in accordance with the requirements of international trade

research (48 sectors, including 27 manufacturing sectors and 18

ser-vice sectors)

On the basis of the database, a global multi-sector computable

general equilibrium model (GMCGE model), which is suitable for the

analysis of international economic cooperation, has been developed

The country samples contain 225 countries, including 29

Organiza-tion for Economic Co-operaOrganiza-tion and Development (OECD)

coun-tries and 10 major developing councoun-tries In terms of data size, the

GMCGE model is considerably broader than the databases controlled

by OECD, Global Trade, Assistance, and Production, and Basque

Institute of Statistics The static CGE model has 65,026 equations

and 74,620 variables, consisting of 38,376 exogenous variables and

36,244 endogenous variables The last chapter of this book briefly

summarizes the basic conclusions of this research

Dr Wang Zhi and I have been friends for many years He studied

and developed the model technology with perseverance and forged

ahead to the forefront of the study of the CGE quantitative model

in the world He cooperated with Prof Wei Shangjin and made

outstanding contributions to dealing with the import and export

value-added chain Dr Marinos E Tsigas of the U.S International

Trade Commission contributed substantially to the computer

pro-gram During the research, Li Xin and I went to Washington, and

Wang Zhi and Marinos E Tsigas went to Beijing twice We all

coop-erated to address all the difficulties we encountered

Professors Chen Xikang, Yang Cuihong, Zhu Kunfu, and Jiang

Xiumei of the System Science Institute of China Science Academy

as well as Dr Qi Shuchang of the National Bureau of Statistics

viii Preface

worked rigorously on the non-competitive input–output tables

With-out their pioneering work, decomposing the data of processing and

non-processing trade is nearly unthinkable

Three years have passed, and now I can finally offer the

prelim-inary results of this research to you Frankly, given such a difficult

and large-scale research, this research report is far from its maturity,

and further investigation and improvement are needed Nevertheless,

this study is a major step toward the development of quantitative

research on the exchange rate reform of RMB

Professors Yao Yang, Zhou Qiren, Lu Feng, Huo Deming, Li Ling,

and Wu Hemao of the China Center for Economic Research of Peking

University provided me with support and encouragement in various

aspects of this research I also want to express my sincerest

grat-itude to Wang Qishan, Zhou Xiaochuan, Yi Gang, Liu Guoguang,

Wu Jinglian, and Tang Ming for their help and guidance

Dr Li Xin and I would like to thank the Institute of National

Accounts, Beijing Normal University and the National School of

Development, Peking University for providing an excellent academic

environment for this research Objectively, every individual can use

his/her expertise and be encouraged to explore Moreover, I would

like to extend my appreciation to Huron College of the

Univer-sity of Western Ontario for providing me with exemplary research

conditions

Although I already expressed my gratitude to my wife Guan

Keqin, I will do so again I am still convinced that without her

logis-tics support, I would not have been able to write this book in such

a short time

In this book, we freely convey our own opinions in the hope

that they will generate more discussions Committing mistakes is

unavoidable, and thus I sincerely welcome the comments and advice

of friends

Xu Dianqing

May 4, 2012

1.1 Rapid Development of the Sino-U.S

Bilateral Trade 21.2 Why Trade Deficit in U.S Statistics is Higher

than that in Chinese Statistics? 51.3 Co-existence of Surplus and Deficit 9

1.4 Repeated Calculation of Processing Trade

in Customs Statistics 131.5 Theoretical Model of the Value-added Approach 17

1.6 Difference in Calculating China’s Trade between

the Customs Frontier Method and theValue-added Approach 221.7 Keeping Pace with the Times and Reforming

the Foreign Trade Statistical Method 33Chapter 2 On Trade Surplus from Property Rights 35

2.1 Distinguishing between Capital Flow and Logistics

in the International Trade 362.2 Competitive and Non-competitive

Input–Output Tables 402.3 Shares of the Domestic and Foreign Capitals in

Non-labor Compensation in Processing Trade 422.4 Distinguishing Capital Flow by Country 46

2.5 Sino-American Trade Surplus 55

ix

x Contents

Chapter 3 Interpreting the Economic Scale of China 59

3.1 Two Methods of Calculating the GDP 60

3.2 Limitations of Calculating the GDP using the Atlas Method 62

3.3 Advantages and Limitations of the PPP 67

3.4 Distorted Proportion of the Service Industry 70

3.5 Reasons for the Underestimation of the Service Industry 73

3.6 Estimate of the Proportion of the Service Industry in the GDP of China 78

3.7 Processing the Depth Coefficient Method for Estimating the GDP 91

3.8 Avoiding the Quality Difference 93

3.9 Value-Added in the Manufacturing Process 94

3.10 Virtual Output Value of Raw Materials 94

3.11 Processing Depth Coefficient 98

3.12 The Bottom Line and the Possible Range of the GDP of China 100

Chapter 4 Foreign Trade Dependence and Trade Weighted Method 103 4.1 Debate Caused by Foreign Trade Imbalance 104

4.2 Four Errors Affecting China’s Foreign Trade Statistics 107

4.3 Influence of Two GDP Statistical Methods on the FTD 109

4.4 Proportion of the Current Account in the GDP 113

4.5 International Comparison of Degree of Foreign Trade Imbalance 113

4.6 Adjusting the Current Account According to Property Rights 117

4.7 Adjusting the GDP according to the Proportion of the Service Industry 118

Contents xi

4.8 Double Correction of the Import and Export

Data 1204.9 International Comparison of China’s FTD 125

Chapter 5 The Standard in Judging the Exchange Rate 131

5.1 Different Opinions on the Exchange Rate 132

5.2 Basis for Judging the Exchange Rate 133

5.3 Calculating the Equilibrium Exchange Rate

According to Money Market and Importand Export Market Equilibrium 1365.4 Calculating the Equilibrium Exchange Rate

According to Trade Equilibrium 1385.5 No Nash Equilibrium in the Foreign Exchange

Market 1405.6 Different Conclusions based on Different Methods 142

5.7 Market: The Touchstone for Judging Distortion

of Exchange Rate 144Chapter 6 Stabilizing the Currency and Manipulating

6.1 Who Determines the Exchange Rate? 148

6.2 Selection of the Exchange Rate System 150

6.3 The Central Bank’s Duty is to Stabilize the Value

of Currency 1536.4 What is a Currency Manipulator? 159

6.5 The RMB is in the Appreciation Channel 165

6.6 Internal Contradiction between Price Stability

and Exchange Rate 1676.7 Excessive Liquidity and the Money Lake 174

Chapter 7 Origin and Development of the High

7.1 China’s High Savings Rate Causes Problems 186

7.2 China’s High Savings Rate 188

7.3 Savings and Economic Growth Stage 190

xii Contents

7.4 Three Parts of Total Savings: Household,

Enterprise, and Government Savings 1937.5 Why is the Enterprise Savings Rate Remarkably

High? 1977.6 Why is the Government Savings Rate

Remarkably High? 1997.7 Causes of the High Household Savings Rate 207

7.8 Is the Savings Rate Attributable to the High

Housing Price? 2197.9 The First-generation Effect of Savings 224

7.10 Empirical Study of the Savings Rate 229

7.11 Never be Misled by Others to Exceed

the Proper Limits in Correcting a Mistake 240

Chapter 8 Computable General Equilibrium Model

8.1 Compromise and Quantification 244

8.2 Five Stages of the Mathematical Model 246

8.3 Partial Equilibrium and General Equilibrium 248

8.4 CGE Model Structure 250

8.5 Avoiding the Double Account in Calculating

Processing Trade 2568.6 The Database Reflecting the Valued-added

Distribution in the Global Production Chain 2598.7 Use of the CGE Model to Study the Exchange Rate

Adjustment 260Chapter 9 Effect of the Exchange Rate on Employment 263

9.1 High Unemployment Rate: A Difficult Problem for

the U.S 2649.2 The Obama Government’s Countermeasures

to Increase Employment Opportunities 2669.3 Shifting the Target to Frame China 272

9.4 Main Reasons for the High Unemployment Rate

in the U.S 277

Contents xiii

9.5 Analysis of the Relationship between Exchange

Rate and Employment from the Perspective

of System Engineering 279

9.6 Unemployment and the Industrial Transfer 285

9.7 Different Effects on Different Industries 290

9.8 Effect of the Exchange Rate Adjustment on the Unemployment Rate of Different Industries 292

9.9 Effect of the Exchange Rate Adjustment on Employment in Processing Trade and General Trade 294

9.10 Effect of the Exchange Rate Adjustment on the Employment of Industrial Departments with Different Science and Technology Contents 298

9.11 Will the RMB Appreciation Reduce the U.S Unemployment Rate? 299

Chapter 10 The Impact of the Exchange Rate Adjustment on Import and Export 305 10.1 The RMB Exchange Rate and the U.S Trade Deficit 306

10.2 Effect of the RMB Appreciation on the Import and Export Volumes of Different Economies 309

10.3 Effect of the RMB Appreciation on General Trade and Processing Trade 315

10.4 Effect of the RMB Appreciation on the Manufacturing Industries of Different Science and Technology Contents 321

10.5 Three Groups of Competitors 328

10.6 Ways to Narrow the U.S Trade Deficit 342

10.7 Is the International Trade Really Distorted? 344

Chapter 11 Range and Path of the RMB Appreciation 349 11.1 Path Choice for the RMB Appreciation 350

xiv Contents

11.2 Dynamic Simulation of the Effect of Different

Exchange Rate Appreciation Ranges onChina’s Labor Market 35111.3 Dynamic Simulation of the Effect of Different

Ranges of Appreciation on China’sImports and Exports 35511.4 Dynamic Simulation of the Effect of Different

Ranges of Appreciation on China’s DomesticProduction 356Chapter 12 Economic Sanctions and Free Trade 361

12.1 Origin and Evolution of the Trade War 363

12.2 Means to Impose Economic Sanctions 377

12.3 Probability of Success of Economic Sanctions 379

12.4 Six Commandments of Trade War 382

12.5 Cost of a Trade War 394

12.6 Different Logics on Dealing with a Trade War 396

12.7 CGE Model Simulating a Trade War 399

12.8 Scenarios of a Trade War 401

12.9 Impact of a Trade War on Global Trade 402

12.10 Impact of a Trade War on GDP

and Employment 41012.11 A Trade War is Not Good for Anyone 413

12.12 Possibility of a Trade War between the U.S

and China 41712.13 China–U.S Free Trade Agreement 423

13.1 Focus of Dispute on the RMB Exchange Rate 429

13.2 Fictitious Foreign Trade Imbalance 430

13.3 Stabilizing the Currency and Manipulating

the Exchange Rate 43413.4 Origin, Advantages, and Disadvantages of High

Savings Rate 436

Contents xv

13.5 Impact of RMB Appreciation on Employment and

Foreign Trade 43813.6 A Win–Win Free Trade Agreement is Better than a

Lose–Lose Trade War 442

This page intentionally left blank

Chapter 1

Overestimated Trade Surplus

• The overestimated U.S trade deficit in Sino-U.S trade can

be explained by five main reasons: (i) issues on re-exportation

(Chinese export commodities must be transferred through a third

port, resulting in the statistical discrepancy); (ii) two parties offer

different quotations; (iii) the caliber of the statistical district is

different; (iv) customs statistics causes the repeated calculation of

processing trade; and (v) a large part of the property rights in the

processing trade does not belong to China

• In the new century, globalization promotes the intra-industry

divi-sion of labor, and the production chain crosses the borders for

con-tinuous expansion The value-added products comes from different

countries in the production chain, inevitably causing the repeated

calculation of customs statistics with the geographical concept as

the basis To discuss trade surplus in China in accordance with the

traditional customs method is similar to notching the boat to find

the sword

• The repeated calculation of the foreign value-added return of

exports and the domestic value-added return of imports in

pro-cessing trade using the customs statistical method, according to

the customs statistics, caused the total trade volume of China to

be overestimated by 16.5% in 2002 and 25.8% in 2007 and the trade

surplus in China to be overestimated by 25.5% in 2002 and 33.2%

in 2007 With the development of processing trade, the

statisti-cal errors caused by the repeated statisti-calculation become increasingly

serious

• The distortion of the statistics of the trade surplus using the

customs method varies with the industry The customs

statis-tical method overestimates the trade surplus in the industries

1

2 From Trade Surplus to the Dispute over the Exchange Rate

of office and computer manufacturing, electronic equipment

manufacturing, instrument and meter, motor vehicle and its

equipment manufacturing, general and special equipment

man-ufacturing, and electromechanical manufacturing Conversely, it

underestimates the trade deficit in the electronic equipment

man-ufacturing industry In 2007, the trade surplus in the office and

computer manufacturing industry was overestimated by 78.6%,

whereas the trade deficit in the electronic equipment

manufactur-ing industry was underestimated by 24.8%

• The repeated calculation caused by using the customs method

overestimates the trade surplus of China and exaggerates the trade

deficit of developed countries, resulting in the misjudgment of the

trade dependence among different countries Similarly, it

height-ens the trade frictions between the developed and the developing

countries, which is not conducive to the integration of world

econ-omy Therefore, the import and export statistical system should

be reformed urgently

1.1 Rapid Development of the Sino-U.S.

Bilateral Trade

In 2014, the foreign goods’ import and export value of China

amounted to $4.3 trillion, that is, up to 3.6% over the previous year

In this statistics, goods exports reached $2.3 trillion and the goods

imports reached $2.0 trillion The trade surplus of goods was $382.4

billion The trade surplus of China mainly comes from the processing

trade, and the general trade has already had a deficit Trade surplus

has been declining for three consecutive years, with its proportion in

the GDP increasing from 3.1% in 2010 to 3.6% in 2014

Some scholars take the proportion of the current account

sur-plus in the GDP as the indicator of the degree of trade imbalance

According to the World Economic Outlook database, the shares of

the current account imbalances of China in the global imbalances

increased from 6.8% in 2003 to 24.3% in 2008 It dropped slightly to

24.1% in 2009 and to around 19% recently Some researchers assert

that the current account imbalance is a major cause of the world

Overestimated Trade Surplus 3

financial crisis and that China is the first object to be criticized and

attached to it

In recent years, the Sino-U.S bilateral trade has been developing

rapidly, with more than 180 times growth from 1979 to now

Accord-ing to U.S statistics, the Sino-U.S trade volume was only $102 billion

in 1999 However, it grew to $302 billion in 2005 and $649 billion in

2014, which is an average annual growth of 13.6% The proportion

of goods and service exported by China in the total imports of the

U.S increased from 4.9% in 1999 to 12.1% in 2014 The proportion

of exports by the U.S to China in its total exports was only 1.2% in

1999, but it increased to 4.8% in 2014 The proportion of Sino-U.S

trade in the total trade volume of the U.S increased from 3.2% in

1999 to 8.8% in 2014

In the past decade, the U.S exports to China surged to

an aggregate of 468% In 2014, Sino-U.S bilateral trade totaled

$649.3 billion (Table 1.1) China’s imports from the U.S were $166.6

Table 1.1: Proportion of U.S imports from China in Its total imports and

the proportion of U.S exports to China in its total exports.

Proportion of U.S.

imports from China

in its total imports

Proportion of U.S.

exports to China in its total exports

Proportion of Sino-U.S trade in U.S trade

4 From Trade Surplus to the Dispute over the Exchange Rate

Table 1.2: Top 10 Trade Partners of the United States.

Source: The U.S Census Bureau, http://www.census.gov/foreign-trade/statistics.

billion According to the U.S Chamber of Commerce statistics, the

U.S exports to China have created more than three million job

opportunities for the American people According to the

investiga-tive report of Morgan Stanley, the commodities imported from China

save about $100 for U.S consumers, and U.S enterprises earn $600

billion every year.1

From 1997 to the Global Financial Crisis in 2008, the U.S exports

to the world increased by 53.7%, and its exports to China increased

by 443.4%, which is 8.26 times faster than the former During this

period, the growth of U.S exports to China was 1.19 times as high

as the growth of its imports from China However, the U.S trade to

the world shows the reverse situation in the same period: the growth

of its imports from the world was 1.47 times higher than the growth

of its exports to the world Among the trade partners of the United

States, China’s ranking is rising year by year In 1990, China ranked

No 8 in the U.S imports and No 18 in the U.S exports, but in

2014, China ranked No 1 in the U.S imports and No 3 in the U.S

exports (Table 1.2)

1Quoted fromThe Retrospect and Prospect of RMB Exchange Rate Formation

Mech-anism Reform Process, the People’s Bank of China, October 2011.

Overestimated Trade Surplus 5

1.2 Why Trade Deficit in U.S Statistics is Higher

than that in Chinese Statistics?

The U.S trade deficit progressively increases with the annual increase

in bilateral trade

According to U.S statistics, in 2000, the U.S trade deficit with

China, Japan, Canada, and the Organization of Petroleum Exporting

Countries (OPEC) accounted for 19.2%, 18.7%, 1.9%, and 11%, of its

total deficit, respectively In the past decade, U.S trade deficits with

Japan and Canada declined gradually, whereas its trade deficits with

the OPEC and China increased gradually In particular, its trade

deficit with China increased rapidly According to the statistics of the

U.S Trade Department, in 2014, the U.S trade deficit with China

accounted for more than 56% of its total trade deficits, exceeding the

total deficits of the U.S with Germany, France, Canada, and Mexico

(Table 1.3)

In the Sino-U.S bilateral trade, how much exactly is the U.S

trade deficit? The statistics of China and the U.S vary greatly The

statistical data of the U.S is much higher than those of China As

a result, both sides keep arguing about it In 2014, according to

Chinese statistics, the U.S trade deficit was $237.0 billion, whereas

according to U.S statistics, its trade deficit was $342.9 billion, with

a difference of as much as $105.9 billion (Table 1.4)

The 20th China–U.S Joint Commission on Commerce and Trade

held in Hangzhou in October 2009 released The Report on the

Sta-tistical Discrepancy of Merchandise Trade between the United States

and China, which established the official explanation of the

statisti-cal discrepancy of merchandise trade between the U.S and China

According to the research scope and methods agreed upon by both

sides, the Chinese team selected the Sino-U.S trade statistics in 2000,

2004, and 2006 and compared the data according to the grouping

method of eastbound and westbound trades The so-called

west-bound trade refers to the logistics to the west; that is, China is

the importer and the U.S is the exporter Conversely, the so-called

eastbound trade refers to the logistics to the east; i.e., the U.S is

the importer and China is the exporter The statistical figures of the

Table 1.3: The trade deficits of different economies to the U.S (billion $).

Brazil Canada China France Germany India Italy Japan South Mexico Arabia Kingdom

Definition of Indicator: “—” means the U.S trade surplus with the UK in 2011.

Source: The U.S Department of Commerce, Bureau of the Census, Foreign Trade.

Overestimated Trade Surplus 7

0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 400.0

1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Chinese Statistics U.S Statistics

Figure 1.1: U.S Trade deficit with China (billion $).

Source: The Chinese statistics are from CEInet Database and the U.S Statistics

are from the U.S Bureau of Economic Analysis.

westbound trade of both sides were very close, with a difference of

only $4 billion in 2006 However, the statistical figures of the

east-bound trade of both parties varied greatly, that is, as high as $84.3

billion (Fig 1.1)

The reasons for this discrepancy are as follows

First, Chinese export commodities must be transferred through a

third port, resulting in a statistical discrepancy A large number of

the commodities that China exports to the U.S are initially

trans-ported to Hong Kong, Taiwan, Busan in South Korea, and Mexico

Some commodities are re-packed, and others are shipped to the U.S

after simple re-processing In this process, the prices of the

commodi-ties increase, generating a part of the value-added Chinese exporters

do not necessarily know that the commodities are exported to the

U.S in the end Chinese Customs have to classify these commodities

as exports to South Korea and Hong Kong according to the customs

declaration of the exporters However, the U.S classifies these

com-modities as imports from China according to the principle of origin of

goods The resulting difference was $44.1 billion in 2006, accounting

for 52% of the difference in the eastbound trade

8 From Trade Surplus to the Dispute over the Exchange Rate

Table 1.4: The U.S trade deficit with China (billion $).

Chinese The U.S Statistical discrepancy statistics statistics between China and the U.S.

Source: The Chinese statistics are from CEInet database, and the

U.S statistics are from the U.S Bureau of Economic Analysis.

Second, different quotations cause statistical discrepancy Among

the commodities exported from China to the U.S., 60% are from

pro-cessing trade Substantial differences exist between the commodity

declaration for exportation by the Chinese exporters and the customs

clearance by the U.S importers As the Chinese exporters are only

responsible for receiving orders and producing products and do not

control the design and sales links, they do not have a clear idea of the

final price set by the U.S importers As a result, the U.S statistics

of imports from China are greater than China’s statistics of exports

to the U.S

Third, the different calibers used by China and the U.S in

pro-cessing the customs data also lead to statistical discrepancy For

example, under the statistical area, the U.S regards Puerto Rico and

the Virgin Islands as the customs, and its trade statistics with China

includes the data of these two areas However, China regards these

Overestimated Trade Surplus 9

two areas as separate administrative regions and does not include

the trade with these two regions in the trade volume with the U.S.,

resulting in a statistical discrepancy Similarly, the cross-year

trans-portation eventually causes the statistical discrepancy

On March 31, 2011, Deming (2010), the former Minister of

Com-merce of the People’s Republic of China, pointed out in a signed

arti-cle entitled Promote the Sound Development of the Sino-U.S

Eco-nomic and Trade Relations Constructively that the degree of the U.S.

trade deficit with China was obviously overestimated based on the

Report on the Statistical Discrepancy of Merchandise Trade between

the U.S and China According to the calculation of the Ministry of

Commerce, in 2009, the figure released by the U.S exceeded nearly

$60 billion than what should be the actual U.S trade deficit with

China After the adjustment, the figures of both sides were close to

each other Moreover, the U.S trade deficit with China was

approx-imately $150 billion

Aside from three points discussed above, two important factors

also contribute to the overestimated U.S trade deficit, namely, the

repeated calculation of processing trade by the customs statistics and

the property ownership in the processing trade With the progress

of the world economic integration, the intra-industry division of

labor develops rapidly As a result, processing trade accounts for an

increasing proportion in China’s exports For example, it accounts for

nearly half of China’s exports The new manner of trade has brought

many new challenges to the international trade theory and statistics

Chapter 2 of this book will discuss the phenomenon of separation

of cash flow from logistics in international trade caused by

prop-erty ownership This chapter focuses on the statistical discrepancy

caused by the repeated calculation of processing trade in customs

statistics

1.3 Co-existence of Surplus and Deficit

Despite the different import and export statistics between China and

the U.S., the latter has an extremely high trade deficit, and the figure

is very high

10 From Trade Surplus to the Dispute over the Exchange Rate

Table 1.5: Proportion of the service and merchandise trade surplus/deficit in

the GDP in the major countries of the world.

Definition of Indicator: Negative numbers represent the service and trade deficit, and

the positive numbers represent service and trade surplus.

Source: WDI, World Bank.

Notably, China’s trade surplus has been moving on a downtrend

trend after successively increasing for 15 years The proportion of

China’s trade surplus in the GDP reached a peak of 8.8% in 2007

Afterward, it began to decline year after year The proportion was

3.8% in 2010, which was lower than that of Germany (5.2%) and the

same as that of Russia (8.5%) Since 2010 China’s trade surplus has

been declining more significantly In 2014, the trade surplus in the

entire year accumulated to $382.5 billion, which is 2.6% of the GDP

(Table 1.5)

People usually focus on China’s huge trade surplus and ignore

another aspect: Although China has a trade surplus with some

economies, its trade deficit with other economies is remarkably high

In 2013, China’s trade surplus amounted to $259.7 billion,

includ-ing $215.9 billion with the U.S and $118.9 billion with the European

Overestimated Trade Surplus 11

Union (EU 28) Therefore, China’s trade surplus with Europe and

the U.S was as high as $334.8 billion At the same time, China had a

huge trade deficit of $91.9 billion with South Korea, Taiwan ($116

bil-lion), Australia ($61.3), and Brazil ($17.9) After balancing between

the trade surplus with Europe and the U.S and the trade deficit with

South Korea, Taiwan, Australia, and Brazil, China’s trade surplus

with other regions was only $109.5 billion

China has a trade surplus with Europe and the U.S and a trade

deficit with four Asian economic giants, reflecting the operation

char-acteristics of the global production chain Many multinational

com-panies (mainly from Europe and the U.S.) invest in China They

import raw materials and spare parts from countries around the

world (Japan, South Korea, and Taiwan account for a large

propor-tion), produce and assemble products in China, and finally export the

products A large number of products made in China are exported to

European and U.S markets However, Europe and the U.S provide

a small amount of raw materials and spare parts, and the customs

statistics must reflect the trade deficit Japan, South Korea, and

Tai-wan provide a large number of spare parts; thus, they have a huge

trade surplus with China China has a very high trade deficit with

Thailand, the Philippines, Brazil, Australia, and Russia

The logistics figure of the customs hardly reflects which country

makes money from the trade or whether the business can be

con-tinued In terms of production and trade, property ownership is the

most important Property ownership determines who makes money

Without a doubt, most profits of the production chain are earned

by multinational countries Otherwise, why would they cross oceans

to invest in setting up factories in China? Among the multinational

companies, European and U.S companies account for a high

propor-tion A larger trade deficit of Europe and the U.S corresponds to a

greater profit earned by the multinational companies The

multina-tional companies, in which Europe and the U.S have major shares,

take money out of the left pocket and put it in the right pocket

This kind of trade has continued Therefore, when discussing trade

sustainability, we should focus on how many property rights in the

12 From Trade Surplus to the Dispute over the Exchange Rate

export commodities of China belong to the foreign-invested

compa-nies

According to the analysis of capital flow (ownership), one-third

of the property rights of the balance between imports and exports

included in China’s trade surplus do not belong to China That

is, one-third of the profits are earned by the multinational

com-panies According to our estimation by non-competition of Input–

Output Table, in 2007, China’s imports from the U.S amounted to

$69.39 billion, and its exports to the U.S totaled to $232.68

bil-lion, with a trade surplus of $163.29 billion After excluding $72.073

billion whose property rights belonged to the U.S., China’s trade

surplus should be adjusted to $91.217 billion The adjustment range

reached 44.14% In other words, almost 44% of the property rights

of China’s trade surplus belonged to U.S companies However, the

customs statistics included this part of income into China’s trade

surplus according to the geographical concept, which is completely

unreasonable If we conclude that China’s trade surplus is

remark-ably large according to customs statistics, then the degree of trade

imbalances must be exaggerated.2

From the level of enterprise, since 2004, the imports and exports

of foreign enterprises have always been the most important source of

China’s trade surplus The trade surplus of state-owned enterprises

and that of the collective enterprises has been declining year after

year They have had a trade deficit since 2003 (Fig 1.2)

2See Chapter 2 of this book for the analysis of the trade capital flow Capital flow does

not necessarily coincide with property rights According to the data of2010 Monitoring

Report on China’s Cross-Border Capital Flows by the State Administration of Foreign

Exchange, in the past 10 years, the accumulated profits remittance of the foreign-invested

enterprises amounted to $261.7 billion, accounting for only 22% of the direct investment

stock in China Especially after the outbreak of the global financial crisis in 2008, the

profit remittance of foreign-invested enterprises grew slowly from an average annual

growth of 36% in 2002 and 2007 to 6% in 2008 and 2010 According to the joint annual

inspection of foreign-invested enterprises, the retained profits of the foreign-invested

enterprises by the foreign party amounted to $170.8 billion A considerable part of the

export returns of foreign-invested processing enterprises in China is taken as the retained

profits or is used for reinvestment instead of being remitted outside the Chinese market,

which is regarded as direct foreign investment.

Overestimated Trade Surplus 13

−300.0

−200.0

−100.0 0.0 100.0 200.0

Source: CEInet database.

1.4 Repeated Calculation of Processing Trade

in Customs Statistics

The traditional international trade statistics is based on general

trade The commodities exported through customs are completely

produced by the country, which is often referred to as the customs

statistics For a long time, people have calculated the trade surplus

or deficit according to the customs statistics, and both sides have no

objection

The customs statistics is applicable to general trade, but it

can-not reflect the value-added process of different links and regions in

the global processing and production chain After entering the new

millennium, multinational companies have become the main body of

world trade The rise of multinational companies caused the division

of production links of various products to the maximum, changing

the optimized distribution of global elements For economies with a

large proportion of processing trade, the difference began to appear

between the trade volume according to the customs statistics and

that according to the production chain value-added statistics With

the development of processing trade, the difference has increased

progressively

14 From Trade Surplus to the Dispute over the Exchange Rate

Profit is calculated based on the value-added of each link in the

production process If the value output in each link is taken into

account, it will likely to lead to repeated calculation As early as

1999, the research of Feenstra et al (1999) pointed out that the

processing products exported from China to the U.S should not be

included entirely into the U.S trade deficit with China Since then,

many studies have explored establishing a new statistical method

that decomposes the total export value into the value-added of each

production link and discussed the trade surplus or deficit according

to the distribution of the value-added.3

Hummels, Ishii, and Yi (hereinafter referred to as HIY, 2001)

pro-posed a method for calculating the direct and indirect value-added

of the export of a country Prof Ping (2005) of Peking University

referred to the HIY method and used the China input–output tables

in 1992, 1997, and 2000 to calculate the degree of the processing with

materials in the Sino-U.S trade Based on this previous study, the

vertical specialization of China’s exports to the U.S totaled to 22.9%

In 2003, the U.S imports from China amounted to $92.6 billion

In the U.S production and exports, nearly $1.3 trillion of the value

had direct or indirect relations with China Assuming that this part

was approximately 5%, U.S companies earned at least $60 billion

of profits from China Although the HIY method can be applied to

analyze the position of a country in the vertical integration

produc-tion network, the domestic value-added calculated by the standard

HIY method requires two assumptions First, the domestic

produc-tion can be classified into two: One is for export and the other is

for meeting the domestic demand These two parts must share equal

import input degree Second, the intermediate input of all imports is

the foreign value-added The first assumption is unsuitable for

devel-oping countries based on the exports of processing trade, and the

second assumption is unsuitable for developed countries that need to

transport products through a third port and whose imports contain

great value-added shares of their own

3Refer to Koopmanet al (2008) for the discussion on the added value in the processing

trade.

Overestimated Trade Surplus 15

To solve the deficiencies of the HIY method, Daudin, Rifflart,

and Schweisguth (hereafter referred to as DRS, 2010) pointed out

that the import commodities contain the domestic value-added that

is processed abroad and then returned, that is, the domestic

value-added returned in the export commodities Koopman, Power, Wang,

and Wei (hereinafter referred to as KPWW, 2010) proposed the

KPWW method that integrates the trade statistical method and

the customs statistical method in the accounting system of national

accounts based on the HIY and DRS methods By building the input–

output database of multiple departments in the world, it expands the

domestic value-added statistics from a single country to regions and

even to the world to calculate the domestic and foreign value-added in

the trade of a country According to the calculation results of KPWW

(2010), 37.5% of China’s trade surplus calculated based on the

tra-ditional customs statistical method belongs to overseas value-added

On the surface, China’s trade surplus is very large However, in

reality, China’s general trade has a deficit, and its processing trade

has a surplus According to the Chinese customs statistics,

process-ing trade in China’s exports to the U.S was $175.64 billion in 2011,

accounting for 54.1% of China’s exports to the U.S The

process-ing trade surplus was $153.75 billion, accountprocess-ing for 75% of China’s

trade surplus with the U.S.4 The increasing proportion of processing

trade has caused the difference between the trade surplus according

to the customs statistics and that according to the production chain

value-added statistics to progressively increase The traditional

cus-toms statistics cannot properly reflect the characteristic that China’s

processing trade takes over the idea of trade surplus

For example, we assume that a multinational company produces

an electronic product in China (e.g., iPod by Apple) The total value

of the general parts, such as screw and battery, produced by the

Chinese enterprises is A When the parts are exported overseas, a

part A(1 − α) is used for final consumption and another part αA is

used for producing hardware, audio decoder, and other parts with

the processing value-added of B The total value of importing these

4Source: The People’s Daily, March 23, 2012.

16 From Trade Surplus to the Dispute over the Exchange Rate

+A

+C

+B

(1 − α)A A

Figure 1.3: Import and export statistics in processing trade.

Table 1.6: Difference between the two statistical methods.

Value-added Difference between

Ex+ Im (1 +αβ + α)A + (1 + β)B + C A + B + C α(1 + β)A + βB

Ex − I m (1 +αβ − α)A − (1 − β)B + C A − B + C βB − α(1 − β)A

Definition of Indicator: The total trade volume and trade surplus in the table both

refer to those of China.

parts by China is (αA + B) A part of these parts, that is, (1 − β)

(αA+B), is used for final consumption, and another part β(αA+B)

is used for producing iPod products with the value-added ofC Then,

the export volume of the product is β(αA + B) + C (Fig 1.3).

According to the customs statistics, China’s total trade volume

is (1 +αβ + α)A + (1 + β)B + C, and its trade surplus is (1 +

αβ − α)A − (1 − β)B + C However, according to the value-added

statistics, China’s total trade volume is A + B + C, and its trade

surplus is A − B + C The conclusions drawn by the two methods

have a difference of βB − α(1 − β)A (Table 1.6).

Clearly, if B

A > α(1−β) β , China’s trade surplus is likely to be timated Thus, when comparing the overseas value-addedB with the

overes-domestic value-addedA, a higher value of the former corresponds to

a greater possibility that China’s value-added will be overestimated

In the high-tech industries, such as in the production of iPods, the

value of a computer chip produced overseas is very high, but the

value-added of the parts (e.g., screw) produced in China, the

assem-bly, and the package is not high As the overseas value-added is much

Overestimated Trade Surplus 17

higher than the domestic value-added, China’s trade surplus caused

by high-tech products is easily overestimated

For example, if α = 0.5 and β = 0.5, China’s trade surplus

calculated by the customs statistical method is overestimated by

0.5(B − 0.5A) As long as B > 0.5A5, China’s trade surplus is likely

to be overestimated

Several factors can explain the difference between the two

statisti-cal methods First, the customs statistics of China’s imports

repeat-edly calculate the value-added return of the country in the import

commodities, which has been calculated in the export of general spare

parts (A) Second, the customs statistics of China’s exports

repeat-edly calculate the overseas value-added of imports, which has been

calculated in imports Therefore, the customs statistics of the total

import and export volume of processing trade is greater than the total

value of the actual production because of these repeated calculations

The value-added approach must be introduced to the international

trade accounting system to more reasonably reflect the value-added

of different regions and different production links in the global

produc-tion chain This approach is the only way through which the real profits

of each side in the multilateral trade can be described accurately

1.5 Theoretical Model of the Value-added Approach

The model has two parts, namely, the Chinese and the overseas parts

Each part has N tradable sectors and produces n products Each

article of trade can be directly used for the final consumption or

the intermediate input of the production of other products In other

words, these two parts of trade are fully liberalized without trade

barriers According to the KPWW method, the global input–output

table, including the import input information in 2002 and 2007, is

sorted out, and the domestic and foreign shares in the value-added

of China’s trade are discomposed.5

5The estimation of the trade appreciation by KPWW uses the two-country

(Kroop-man et al., 2008) and the multinational methods (Kroopman et al., 2010), with the

18 From Trade Surplus to the Dispute over the Exchange Rate

According to the transverse equilibrium relation of the input–

output table,

The total output = the intermediate needs + the final consumption

= (the domestic production demand+ the overseas consumption export),

i.e.,

X C = (A CC X C +A CW X W) + (Y CC+Y CW), (1)where subscript c refers to China; w refers to overseas; X c and X w

are the output vectors of N × 1 representing the total output of

N tradable departments in China and overseas; A represents the

N × N direct consumption coefficient matrix; A ccrefers to the direct

consumption coefficient matrix of the consumption of the Chinese

commodities in China’s total output; and subscriptA cw is the input

of c to w, that is, the direct consumption coefficient matrix of the

consumption of the Chinese commodities in the overseas output

Equation (1) can be expanded to the regional input–output module

composed of two parts (Inter-regional Input–Output Table, IRIO):

former method is improved based on HIY and DRS and is mainly reflected in the

non-competitive input–output table, including the processing trade information of China and

Mexico.

Overestimated Trade Surplus 19

where V represents the increment vector of value of N × 1, and ˆΦ

represents the diagonal matrix of the intermediate input rate vector

As the intermediate input includes the intermediate input of imports

and the domestic intermediate input,

Equation (6) can be rewritten as V = (I − ˆΦ)X = ˆ V B · Y As ˆΦ

represents the intermediate input rate, (I − ˆΦ) is the value-added

ratio ˆV of the total output, and ˆ V B is the value-added ratio of the

V C B CC is China’s value-added ratio in China’s final products,

and ˆV W B W C is the overseas value-added ratio in China’s final

prod-ucts Similarly, ˆV C B CW is China’s value-added ratio in the overseas

final products, and ˆV W B W W is the overseas value-added ratio in the

overseas final products The value-added created by the Chinese or

overseas production is assumed as the unit value Thus,

ˆ

V C B CC+ ˆV W B W C= ˆV C B CW + ˆV W B W W =u,

where u is the unit vector of 1 × N.

Setting E as the final product export, E = E C 0

0 E W

 E c isChina’s final product export to overseas, and E w is the overseas

20 From Trade Surplus to the Dispute over the Exchange Rate

final product export to China (i.e., China’s final product import)

Therefore, the trade volume multiplied by the value-added ratio is

where ˆV C B CC E C and ˆV W B W W E W represent the value-added created

in China in China’s export commodities and the value-added created

overseas in the overseas exports to China, respectively; ˆV C B CW E W

and ˆV W B W C E C represent the value-added with the Chinese

prod-ucts as the intermediate inputs in the overseas exports to China (i.e.,

China’s value-added return in its imports) and the value-added with

the overseas products as the intermediate inputs in China’s exports

to overseas (i.e., the overseas value-added return in overseas imports)

ˆ

V C B CC+ ˆV W B W C= ˆV C B CW + ˆV W B W W =u,

where u is the unit vector.

According to the customs statistical method, the formula of

China’s total trade volume is as follows:

the total exports + the total imports

=E C+E W = ( ˆV C B CC+ ˆV W B W C)E C

+ ( ˆV W B W W+ ˆV C B CW)E W (11)

Overestimated Trade Surplus 21

According to the value-added statistics, the formula of China’s

total trade volume is

the domestic value-added exports + the overseas value-added imports

imports = ˆV C B CC E C + ˆV W B W W E W (12)The total trade volumes calculated using these two statistical

methods differ mainly in the processing of the returned

value-added China’s exports of products with value-added to overseas

( ˆV C B CC E C) has two parts One part is used for the overseas

final consumption, and the other is used for overseas production

as the intermediate inputs After being used for overseas

produc-tion, the latter can also be returned to China through the overseas

export to China, that is, the return of China’s value-added in its

imports ( ˆV C B CW E W) As this part of the returned value-added has

been included in China’s value-added in China’s exports to

over-seas, ˆV C B CW E W ∈ ˆ V C B CC E C; similarly, ˆV W B W C E C ∈ ˆ V W B W W E W

Thus, the customs statistics repeatedly calculates the returned part

of the value-added The value-added return phenomenon only occurs

in processing trade; therefore, a higher proportion of processing trade

corresponds to a greater degree of repeated calculation Although the

proportion of the imports and exports of processing trade in China’s

total trade volume dropped from the peak of 53.4% in 1998 to 35.8%

in 2011, processing trade still accounts for more than one-third of

China’s total trade volume The traditional customs statistics

exag-gerates China’s total trade volume to a certain extent

When calculating the trade surplus, the formula of the customs

statistical method is as follows:

the total exports − the total imports

=E C +E W = ( ˆV C B CC + ˆV W B W C)E C

− ( ˆ V W B W W + ˆV C B CW)E W (13)The formula of China’s trade surplus by the value-added approach

is as follows:

the domestic value-added exports− the overseas value-added imports

22 From Trade Surplus to the Dispute over the Exchange Rate

In the statistics of the trade surplus, the difference between the

customs statistical method and the value-added statistical method is

ˆ

V W B W C E C − ˆ V C B CW E W

If ˆV W B W C E C  ˆ V C B CW E W, that is, when the value of using

the overseas products as the intermediate inputs in China’s exports

to overseas is greater than that of using Chinese products as the

intermediate inputs in the overseas exports to China, the trade

sur-plus according to the customs statistics will be greater than that

according to the value-added statistics

If ˆV W B W C E C ≺ ˆ V C B CW E W, the trade surplus according to the

customs statistics will be smaller than that according to the

value-added statistics

If ˆV W B W C E C = ˆV C B CW E W, the trade surplus according to the

customs statistics will be equal to that according to the value-added

statistics

1.6 Difference in Calculating China’s Trade between

the Customs Frontier Method and the Value-Added Approach

According to the domestic input–output tables and the import-input

tables of China and the world in 2002 and 2007, China’s total trade

volume and trade surplus are analyzed comparatively with the

cus-toms method and the value-added method globally The domestic

input–output tables of China and the world are used for calculating

A CC , A W W , E C, andE W; the import-input tables of China and the

world are used for calculating A W C and A CW

The direct consumption coefficient matrix of China’s domestic

production and imports in 2007 is from China’s Non-competitive

input–output table in 2007 compiled by the Department of National

Economic Accounting of the National Bureau of Statistics (Xikang

et al., 2011) Since China began to compile the non-competitive

input–output table, including the processing information in 2007,

no official non-competitive table was developed in 2002 The present

study assumes that the import consumption proportion of the

domes-tic product remains unchanged in 2000 and 2002 and approximately

Overestimated Trade Surplus 23

replaces the table in 2002, with the direct consumption coefficient of

China’s imports in 2000 released by the OECD

The direct consumption matrix of the global production and

import is summarized according to the input–output database of

the OECD and the EU The samples include 35 major countries

except China, that is, 27 EU countries, Brazil, Canada, Indonesia,

Japan, South Korea, Norway, South Africa, and the U.S As only the

input–output tables of the developing countries in 2005, except 27

EU countries and the U.S., are available, this study uses the direct

consumption coefficients of these countries in 2005, assumes that the

production technology is fixed, and obtains the input–output tables

and the import-input tables of these countries in 2007 according

to their total input, value appreciation, total output, imports, and

exports of different sectors in 2007 in the STAN Structural Analysis

Database with the RAS method These tables are then combined

after the adjustment with the exchange rate in the current period to

obtain the approximate global input–output table in 2007 and the

corresponding import-input table

As the classifications of 48 departments in the OECD, 65

depart-ments in the EU, and 42 departdepart-ments in China are different, this

study connects these three groups of data one by one using the

international standard industry classification code (ISIC rev 3.1) To

avoid the separation processing, this study finally reviews 40

ments including 20 manufacturing industries The specific

depart-ment classification is shown in Table 1.7

The processed data are included in Formulas (7)–(10); the specific

results are listed in Tables 1.8 and 1.9

Table 1.8 analyzes the changes in the domestic value-added ratio

and the overseas value-added ratio of different trade departments in

China in import and export trades in 2002 and 2007 The proportion

of China’s overseas value-added ratio in its exports continued to rise

with the deepening of the internationalization of production The

average value increased from 11% in 2002 to 15% in 2007, which is

an increase of 4% in five years Accordingly, the average proportion

of China’s domestic value-added ratio return of import also increased

from 12% in 2002 to 16% in 2007 The global economic integration