Supply Chain Resilience: Reducing Vulnerability to Economic Shocks, Financial Crises, and Natural Disasters [1st ed.] 9789811528699, 9789811528705

Citation preview

Venkatachalam Anbumozhi Fukunari Kimura Shandre Mugan Thangavelu Editors

Supply Chain Resilience Reducing Vulnerability to Economic Shocks, Financial Crises, and Natural Disasters

Supply Chain Resilience

Venkatachalam Anbumozhi Fukunari Kimura Shandre Mugan Thangavelu •

•

Editors

Supply Chain Resilience Reducing Vulnerability to Economic Shocks, Financial Crises, and Natural Disasters

123

Editors Venkatachalam Anbumozhi Economic Research Institute for ASEAN and East Asia ERIA Gelora Bung Karno, Senayan Jakarta, Indonesia Shandre Mugan Thangavelu Jeffrey Cheah Institute for Southeast Asia Sunway University Selangor, Malaysia

Fukunari Kimura Economic Research Institute for ASEAN and East Asia ERIA Gelora Bung Karno, Senayan Jakarta, Indonesia Faculty of Economics Keio University Tokyo, Japan

Institute for International Trade University of Adelaide Adelaide, SA, Australia

ISBN 978-981-15-2869-9 ISBN 978-981-15-2870-5 https://doi.org/10.1007/978-981-15-2870-5

(eBook)

This is a U.S. government work and not under copyright protection in the U.S.; foreign copyright protection may apply 2020 This work is subject to copyright. All rights are reserved 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. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Foreword

Globalisation has integrated societies, countries, and economies through various channels. The emergence of global supply chains as a main channel has improved connectivity and facilitated cross-border trade, foreign direct investment, international knowledge, and technology flows. It is essential to understand how global supply chains work, how they affect economic performance, and what policies help to derive greater benefits from them. Global supply links and geographically concentrated production due to increasing specialisation also result in greater risks. In an interconnected production system characterised by strong economic links, the failure of a single entity or cluster of entities may result in cascading disruptions that can bring down the entire supply chain or a large part of it. The global supply chain is a paradigmatic example of a socio-economic structure in which the impacts of unexpected events propagate rapidly through the system. This was vividly illustrated by the compound effects of Japan’s triple disaster earthquake and Thailand’s severe floods in 2011. Moreover, incidents like the seemingly unlikely volcano eruption in Indonesia in 2015 make companies and policymakers realise of how little control they have over many of the vulnerabilities they are confronted with while managing the supply chain risks. Still, some communities appear to be able to cope with disastrous events more effectively than others and return to normality or to a new state from which they can operate. Natural disasters and man-made financial crises, which affect integrated production processes, have become important factors influencing the efficiency of the present global supply chains. Although advanced and developing countries are exposed to the same types of disaster risks, the consequences tend to be more severe in developing countries, where disasters disproportionally affect the marginalised people and firms. Understanding and addressing these risks along the supply chain, where it has become a growing threat to prosperity, have thus become a critical challenge in research, policy, and business practices. Over the last decade, a series of studies have been conducted by the Economic Research Institute for ASEAN and East Asia (ERIA) to investigate the direct impacts of natural and man-made disasters on specific geographical areas. However, only a few analyses have sought to quantify the cascading economic v

vi

Foreword

effects of extreme weather events coupled with economic shocks on the global supply chain. The great complexity of the global economic system in conjunction with methodological and data gaps makes it difficult to estimate the domino effects of unexpected disaster events and economic shocks along regional value chains. The impacts of disasters on supply chains are highly diverse. They affect different entities and production systems in different ways, and they may extend well beyond the here and now. When the disaster impacts spill across space and time, they may be either restrained or amplified through socio-economic networks. They may be influenced by market mechanisms operated through insurance, government action in the forms of infrastructure investment, early warning systems and recovery assistance, and actions of individuals as they strengthen business continuity plans. Looking at the shifting risk management strategies, the recovery phase is often a window of opportunity to learn from experience, mitigate future vulnerability and exposure, and enhance resilience. It is important, in a world where extreme shocks and weather events are expected to become more frequent and severe, that policymakers and affected communities and business entities resolve to build better, inclusive, and resilient value chains. The vulnerabilities of the global supply chain in terms of geography, time scale, climate change, natural disasters, and economic shocks are diverse and difficult to understand consequences for governments and the business community. This book provides a timely exploration of macroeconomic approximations connected to such vulnerabilities, as well as innovative ideas to measure resilience through numerical indicators. The chapters in the book offer insights into the challenges and opportunities for global supply chains, with a particular focus on entry and upgrading possibilities. Business continuity plans and related challenges are also examined, and the book offers concrete policy recommendations and suggests a number of inventions that would allow developing countries to better harness the benefits of disaster recovery. This volume is a useful tool for anyone interested in global trade, supply chains, business, and disaster resilience issues. May 2020

Hidetoshi Nishimura President Economic Research Institute for ASEAN and East Asia Jakarta, Indonesia

Contents

Part I 1

2

3

4

Understanding the Vulnerability of Asian Value Chains: Assessment Methods and Quantitative Impacts

Global Supply Chain Resilience: Vulnerability and Shifting Risk Management Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Venkatachalam Anbumozhi, Fukunari Kimura and Shandre Mugan Thangavelu Economic Shocks and Uncertainties: How Do Firm Innovativeness Enable Supply Chain and Moderate Interdependence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shandre Mugan Thangavelu and Venkatachalam Anbumozhi Supply Chain Resilience in the Global Financial Crisis: An Empirical Study on Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Willem Thorbecke How Do Production Networks Affect the Resilience of Firms to Economic and Natural Disasters: A Methodological Approach and Assessment in Japan, Taiwan and Thailand . . . . . . . . . . . . . . Michael C. Huang and Atsushi Masuda

3

15

37

77

5

Economic Risk Characteristics of an Indonesian Palm Oil Value Chain and Identifying Sources of Uncertainty in Policy Making . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 A. Faroby Falatehan and Budi Indra Setiawan

6

Robustness of Production Networks Against Economic Disasters: Thailand Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Hiroyuki Nakata, Yasuyuki Sawada and Naoki Wakamori

vii

viii

Contents

Part II

Resilience and Recovery of Global Value Chains and Production Networks: Policy Frameworks and Risk Management Strategies

7

Complexities in Supply Chain Resilience Against Economic Shocks and Forming Public Private Partnerships: Korean Automotive Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Venkatachalam Anbumozhi and Ji Hong Kim

8

A Contingent Resource-Based Perspective of Tourism Value Chain and Robustness: European Experiences . . . . . . . . . . . . . . . . 185 Meinhard Breiling

9

Assessing the Competitive Advantage of Public Policy Support for Supply Chain Resilience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 Eiji Yamaji

10 Achieving the Resilience of Production Networks During Economic Crisis: The Case of Chinese SMEs . . . . . . . . . . . . . . . . . 239 Minquan Liu and Hang Tai 11 Stranded Assets and Protecting Value of Food Value Chain from Disasters and Other External Shocks . . . . . . . . . . . . . . . . . . . 281 Vangimalla R. Reddy, Venkatachalam Anbumozhi and Mura Jyostna Devi 12 Regional Frameworks for Advancing Supply Chain Resilience and Business Continuity Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 Venkatachalam Anbumozhi and Fukunari Kimura Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339

Part I

Understanding the Vulnerability of Asian Value Chains: Assessment Methods and Quantitative Impacts

Chapter 1

Global Supply Chain Resilience: Vulnerability and Shifting Risk Management Strategies Venkatachalam Anbumozhi, Fukunari Kimura and Shandre Mugan Thangavelu Abstract Global supply links and geographically concentrated production networks allow a local events such as natural disasters to become a global disruption or global events such as financial crisis have pervasive impact on local economies. While global supply chains can not be the cause of adverse shocks, they act as effective transmission mechanism. Greater connectivity initially decreases risk through risk dispersing and diversification and increase the overall robustness of the supply chain. However, beyond a certain threshold, it increases the supply chain’s fragility and thus systemic risks. Resilience address the supply chain’s ability to cope with the consequences of an avoidable risk events in order to return to its original operations. The chapters in the book are divided into two parts. Part one deals with the assessment methods and quantitative impacts of vulnerable supply chains. Part two discuss resilience mechanism, risk management strategies of global supply chains, and enabling policy frameworks. The combined effects of disasters and economic shocks underlines the growing importance of informing the market players and the policy makers of risk management and hot spots of supply chain resilience as well as technical assistance programs. Keywords Risk management · Resilience · Transmission channel · Vulnerability JEL Classification F64 · L5 · Q27

V. Anbumozhi (B) · F. Kimura Economic Research Institute for ASEAN and East Asia, Jakarta, Indonesia e-mail: [email protected] F. Kimura Keio University, Tokyo, Japan S. M. Thangavelu Jeffrey Cheah Institute for Southeast Asia, Sunway University, Selangor, Malaysia e-mail: [email protected] Institute for International Trade, University of Adelaide, Adelaide, Australia © Springer Nature Singapore Pte Ltd. 2020 V. Anbumozhi et al. (eds.), Supply Chain Resilience, https://doi.org/10.1007/978-981-15-2870-5_1

3

4

V. Anbumozhi et al.

1.1 Introduction Over the last five decades, Asia has transformed into to a center of world’s economic gravity benefitting from globalization through various channels: international trade of manufactured goods and services, foreign direct investment, the international migration of the highly skilled people, cross border knowledge and technology flows, etc. The emergence of global supply chains has contributed to this economic connectivity and increased the interdependency of the countries. Global supply links and geographically concentrated production networks allow a local event to become a global disruption or global events have pervasive impacts on local economies. The risk of breakdown of an entire system due to shocks such as natural disasters, resource scarcity and melting down of credit systems, have become systemic in nature. The global financial and economic crisis of 2007–2008 is an classical example of such an connected shock which did not originate in Asia, but Asian economies were hit by the down turn in export demand and in some cases, by the turbulence in foreign exchange and capital markets resulting from a sudden stop in the inward flows, and or withdrawal of capital. Moreover, South East Asia and East Asia is a highly disaster-prone region. In the period 1990–2017, the region experienced nearly 40% of the global total of world natural disasters (ADB 2019). Asia’s share of the total death toll from these disasters has been nearly 80% in the same period (Carvalho et al. 2016). It is also interesting to observe that nearly 58% of natural disaster in Asia occurs in the South East and East Asia region, where greater supply chain connectivity and production networks exists (Ando and Kimura 2013). In particular, countries that are prone to natural disasters are China (681 disasters), Japan (291), Hong Kong (103), Indonesia (412), Philippines (529), Thailand (119), Vietnam (177), Bangladesh (312), India (604), Iran (193), Pakistan (166), and Sri Lanka (81 disasters). Global links and geographically concentrated production due to increasing specialization allow a local event to become global disruption, National economies have therefore become more vulnerable to so called systemic risk, i.e., the risk of breakdown of an entire system.

1.2 Global Supply Chains as Both Causality and a Transmission Channel of Systemic Risks In an economic system characterized by strong linkages the failure of a single business entity or cluster of entities may result in cascading disruptions that can bring down the entire system or large parts of it (Schwarcz 2008). Growing global supply chains also increases the risks of shocks growing quickly worldwide. While supply chain itself is not the cause of adverse shocks, it may act as a very effective transmissions mechanism. It is not obvious if and when a shock originating in one part of the supply chain will have system wide effect.

1 Global Supply Chain Resilience: Vulnerability and Shifting …

5

Greater connectivity initially decreases individual risk through risk dispersion and diversification—and increase the overall robustness of the supply chain. However, beyond a certain threshold, it increases the supply chain’s fragility and thus systemic risk (Battiston et al. 2012; Gai et al. 2007; Watts 2002). This threshold differs from system to system and is directly affected by a second system characteristic; the degree of redundancy or back-up in the system (Korhonen and Seager 2008). When the diversity decreases and/or system redundancies are eliminated, substitutability—the extent to which other components of the supply chain can provide the same services in the event of failure also decreases. The greater the redundancy in the supply chain system, the easier it for other elements to take over in the event of a failure in one part of the system (Abe and Thangavelu 2013). Several major shocks, both natural and manmade have increased academic and policy makers attention to global supply chain risks. Global supply chains involve interdependent and interconnected network of firms, industries and economies, and can be considered potential carriers of global contagion. There are many causes, or drivers of supply chain risks and they have become more varied over time, as a result of increased importance, length and complexity of global value chains (Saenz and Revilla 2014). A recent World Economic Forum (2017) survey of company executives ranks external events such as natural disasters, sudden demand shocks and information disruptions as most likely to have significant effects on supply chain (Table 1.1). By distinguishing environmental, geo-political economic and technological factors, the close relationship between global supply chain risk and other categories of systemic risk become clear. Table 1.1 Drivers of global supply chain risks Environmental Geopolitical

Economic

Technical

Natural disasters Extreme weather Pandemic Conflicts and war Export/Import restrictions Terrorism Corruption Organized crime Maritime piracy Nuclear/biological weapons Sudden demand shocks Volatility in commodity prices Border delays Currency fluctuation Global energy shortage investment restrictions Shortage of labor Information and communication disruptions Transport infrastructure failure

59 30 11 46 33 32 17 15 9 6 44 30 26 26 19 17 17 30 6 % influenceable uncontrollable

Source Authors based on World Economic Forum survey (2017)

controllable

6

V. Anbumozhi et al.

1.3 Benefiting from Supply Chain Resilience Through Risk Management Over the last two decades, supply chain vulnerability and its managerial counterpart supply chain risk management have received considerable attention by the practitioners as well as academics (Rao and Goldsby 2009). Supply chain resilience address the supply chain’s ability to cope with consequences of unavoidable risk events in order to return to its original operations or move to a new, more desirable state after being disturbed (Christopher and Peck 2004; Peck 2006). Although the concept is new and theoretical justifications are evolving, a definition grounded in multiple disciplines had recently been proposed by Ponomorov and Holcomb (2009). They design supply chain resilience as the adaptive capacity of the supply chain to prepare for unexpected events, respond to disruptions, and recover from them by maintaining continuity of operations at the desired level of connectedness and control over structure and function (Ponomorov and Holcomb (2009). Supply chain resilience focuses on the economic system’s adaptive capability to deal with temporary disruptive events (Smith 2004; Briano et al. 2009). These disruptions imply certain level of turbulence (Hamel and Valikangas 2003) and uncertainty in the supply chain which together cause threats to the current operations (Sheffi 2005). Depending upon the magnitude of this threatening events, the terms disruptions, crisis or disasters are used (Natarajarathinam et al. 2009). The adaptive capacity has been structured along the three distinct disruption phases into the supply chain readiness, responsiveness and recovery (Sheffi and Rice 2005). Furthermore, all definitions share the view that resilience means to respond and recover at the same or better share if operations and thus include system renewal. Whereas no conceptual differences between the definitions of the supply chain’s adaptive capacity at the system level are apparent in the literature, the perspective on the formative resilience elements are less consistent. The formative elements describe how the supply chain event readiness, response and recovery can be secured and is therefore relevant from a supply chain risk management perspective. Formative resilience capabilities are based on integrating and coordinating resources which often cut across functional areas and thus may become manifest in the supply chain processes.

1.4 The Relationship Between Supply Chain Resilience, Supply Chain Vulnerability and Supply Chain Risk Management Supply chain vulnerability is the susceptibility of the supply chain to the likelihood and consequences of disruptions (Blos et al. 2009; Svenesson 2000). It therefore captures the risk exposure of the supply chain and is often conceptualized together with supply chain risks (Wagner and Bode 2006). Something at risk is vulnerable, and by addressing the vulnerability of the supply chain, supply chain risks are addressed.

1 Global Supply Chain Resilience: Vulnerability and Shifting …

7

Supply chain risks are anything that may disrupt or impede the information, material or product flows from original supplier to the delivery of the final product to the ultimate end user (Peck 2006). Supply chain risks also refer to the possibility and effect of mismatch between supply and demand. Based on the generic risk concept, supply chain risks also combine the two risk dimensions, namely; the likely probability of loss with the impact of that loss for the company (Manuj and Mentzer 2008). A supply chain disruption is unintended, untoward and exceptional situation that leads to the occurrence of risk (Wagner and Bode 2006). Since supply chain risk risks are defined something that exists, whether it is managed or not, a supply will always display a certain degree of vulnerability. Furthermore, since some contemporary supply chain practices, such as global and single sourcing increase the supply chain risks and thus a certain degree of vulnerability may deliberately be taking into account to global value chains. To summarize, the vulnerability as the exposure of risks is characteristics of any supply chain system. Supply chain vulnerability as the exposure of the supply chain to risks is a characteristic of any supply chain system. It is a latent condition which only becomes manifested, if a disruptive event occurs. However, the higher the vulnerability, the more likely a disruptive event and more severe its consequences. As a latent condition, supply chain vulnerability can be analyzed by measuring the exposure of the supply chain to environmental, supply networks and organizational risks (Anbumozhi 2015). As a manifest condition, the impact of supply chain disruption indicates the degree of vulnerability. Since lowering the impact of a supply chain disruption is also the key target of supply chain resilience, it can be assumed that both the concepts are related. More specifically, if supply chain resilience decreases the negative consequences of supply chain risk events by ensuring a fast return to its improved operation, facilitated by business continuity plans, it should, at the same time decrease the supply chain vulnerability in the case of a manifest risk event.

1.5 Supply Chain Resilience and Supply Chain Risk Management The most commonly accepted definition of in the fast-growing body of literature on supply chain resilience and the business requirement was provided by Juttner (2005). They define Supply chain risk management as the identification of potential sources of risk and implementation of appropriate strategies through a coordinated approach among supply chain risk members to reduce vulnerability. One main aim of the supply chain risk management is to reduce the supply chain vulnerability. However, others also emphasize that supply chain risk management aims at increasing the resilience of supply chain. From this perspective, the emphasis is on managing and mitigating the risks and thereby fostering the resilience of the supply chain. Although, both outcome perspective appears reasonable, they are not identical and

8

V. Anbumozhi et al.

have not yet been reconciled. For example, a supply chain risk measure does not necessarily impair the vulnerability of the supply chain and, at the same time increase its resilience. Likewise, a supply chain with high vulnerability could either have high or low resilience. Figure 1.1 summarizes the relationship among the three concepts of supply chain vulnerability, risk management and resilience. Here, supply chain resilience could be linked with three distinguishable risk management actions. Firstly, those addressing primarily the probability of supply chain risks (example avoidance), secondly, risk actions focusing on the effects (e.g., risk sharing and or continuity plans) and thirdly, any actions which increase the knowledge about supply chain risks (e.g., information about supply chain partners), If a supply chain resilience initiative only lowers the probability, the supply chain vulnerability, as a latent exposure to supply chain risks is reduced but it will not have an effect on resilience. In this sense as supply chain risk strategy which avoids certain geographical risk areas has lowered the likelihood of a disruption caused, e.g. by compound disaster in a coastal zone where production facilities are located. It has however, not increased its capability to respond and recover from a disaster., if it still occurred. However, if an initiative successfully addresses the risk effects, the supply chain resilience increases. For example, joint business continuity plan developed together with suppliers are likely to have positive impact on the supply chain’s capacity to respond to and recover from a natural disaster or economic shock event through collaboration. In a similar vein, a globally dispersed portfolio of suppliers is likely to positively effect the supply chain resilience in the case of a risk event (ADBI 2015). The supply chain can fall back onto suppliers from other regions, if only one region is affected by the disruptive events such as disasters, and thereby resilience to the event. Finally, a risk approach, which aims to increasing the knowledge about supply chain risks should also have a positive impact on supply chain resilience. Supply chain risk knowledge can improve the event readiness because it increases the visibility and shorten the time for detection of disaster events. Furthermore, risk knowledge management can decrease the supply chain vulnerability because knowledge is an important prerequisite to take counteractive measures prior to risk a risk event. This in turn, decreases the likelihood of a risk event. In summary it can be assumed that supply chain resilience is related with supply chain vulnerability and risk management. Depending on the primary target of the specific action, either the supply chain resilience or the supply chain vulnerability should be positively affected by supply chain risk management. Aimed at reducing the probability and increasing the knowledge (-)

Supply chain vulnerability

Supply chain resilience

Aimed at reducing the risk effect and increasing the knowledge (+)

Supply chain risk management

Fig. 1.1 The relationship between supply chain resilience, vulnerability and risk management

1 Global Supply Chain Resilience: Vulnerability and Shifting …

9

1.6 Context and Purpose of the Book What are the major factors that amplify the impact of global supply chain disruptions due to economic shocks, external forces and natural disasters? How companies and communities prepare for the emergency situations occurring in the global supply chain? Which risk mitigation measures, insurance tools and emergency plans and mitigation measures can companies and local governments can employ to bring resilience to supply chain disruptions efficiently? How regional disaster relief mechanisms could be made conciliatory to business continuity plans of major supply chains? This book attempts to answer these questions by analysing reactions to the global economic crisis and natural disasters at the firm and industry levels through the lens of global supply chains.

1.6.1 Understanding the Vulnerability of Asian Supply Chains The book is divided into two parts. Part one deepens our understanding on the assessment methods and quantitative impacts of vulnerable supply chains. Following this Introduction, chapter 1 analyse the impacts studies effects of economic shocks on the productive performance of manufacturing firms in Vietnam using the firm level data. It examined productive performance of firms in terms of access to linkages, spill overs, and ownership structures under changing conditions of foreign, private and public ownership during and after major economic shocks. The findings suggest that the local private enterprises did not receive positive foreign horizontal and backward spillovers from the Multi-national enterprises while state owned enterprises horizontal linkages along the supply chain also failed to contribute to the improvement in output productivity. Chapter 2 made a connection of exchange rate risks with that of export performance of East Asian value chains. Intricating networks were found to influence technology transfer, productivity growth, and tumbling prices for final goods. It documented the burgeoning trade surpluses between Asia and the rest of the world in goods produced within regional value chains. Empirical evidence indicates that exchange rates throughout the supply chain affect these surpluses. Given intense pushback against globalization, a harmonized appreciation of Asian currencies would be desirable to reduce supply chain risks. To increase the resilience of supply chains, it suggests Asian countries to improve infrastructure, fight corruption, strengthen protections for intellectual property rights, improve the supply of electricity, and invest in the health and education of its people. Chapter 3 examined the economic impact of the disaster in the supply chain of electronic products, transportation equipment and machinery sectors in Asia. A Multiregion static Computational General Equilibrium (CGE) model based on GTAP v.9.0 is used to capture full-fledged propagation effects stemming from natural disaster

10

V. Anbumozhi et al.

to seven economies in Asia. Three hypothetical disaster cases are simulated based upon the collapse rates estimated: Tokyo earthquake, Taipei earthquake and Thailand floods. The simulation results confirmed that the economic impacts of the natural disaster are propagated to other countries in the region through changes in the production volumes and prices response. Economic impacts stemming from neighbouring economies can be positive or negative depending on the production and trade structure defined by the supply chain network. Increased resilience to the disaster will have an implication for increased macroeconomic stability by reducing propagation of the disaster impacts. Accordingly, Chap. 4 deals with global agriculture value chain originating from Indonesia. Palm oil industry is one of the main contributors to the economic development in Indonesia. It has experienced rapid growth in recent decades and become a significant contributor to the world market for vegetable oils. Oil palm plantations produce crude palm oil used as raw material by downstream industries to provide many types of derivative products such as in food and oleochemicals. In consequence, managing value chains are vital to gain sustainable palm oil development considering the increasing competitiveness and needs to maintain mutual partnerships among stakeholders. This chapter analysed possible external shocks that would give potential impacts on the supply chains of the palm oil industry in Indonesia and recommends what the policymakers should do to anticipate the negative implications. It found that some variables show elastic values represented by the world price of crude palm oil the export price of palm oil and the exchange rate to the export quantity of Indonesian oil. Based on our simulation, the export tax policy of palm oil has caused negative impacts both to the small-scale farmers and the big scale industries in the form of decreasing the price and exported quantity of palm oil, respectively. Drawing on the experiences in Thailand, Chap. 5 illustrated the impacts of market concentration on risk-prevention incentives by closely observing the changes in the price and quantity in the hard disk drive industry before and after the devastating floods in 2011. The combination of high price and low quantity persisting after the floods indicates that the floods triggered the formation of a de facto cartel, and a shift in demand for hard disk drives alone is unable to explain the observed combination.

1.6.2 Policy Frameworks for Improved Resilience and Risk Management Part two of this book discuss particular resilience and recovery mechanisms, risk management strategies of global supply chains, and enabling policy frame works. Chapter 6 addresses the complexities in supply chain resilience against economic shocks and forming public-private partnerships with Korean automotive sector as a case study. It used the cases of natural disasters to understand the interactions and impacts along the global automotive value chain. In addition, several strategies to minimize the risk of supply chain are proposed: First, build strong and flexible

1 Global Supply Chain Resilience: Vulnerability and Shifting …

11

procurement system; Second, risk mitigation with diversification of supply sources; Third, information network across the supply chain; Fourth, build a real-time risk evaluation system with quantifiable index and training; Fifth, minimize the damage with a business continuity plans. A contingent resource-based perspective of tourism value chain resilience is brought in by the Chap. 7. European experiences are studied in detail, with six step approach. (i) the risks with global tourism value chain, (ii) international factors making an impact on the tourism value chain, (iii) impacts on the national tourism value chain in Austria, (iv) case study of food pleasure tourism destinations in Austria, (v) impacts on tourism value chain in Southeast Asia and (vi) altering the tourism value chain in Southeast Asia by employing experience from Austria. Chapter 8 assess the competitive advantage of public policy support for supply chain resilience in Japan. Food flow sometimes suffers disasters, such as heavy rains, flooding, low temperatures, strong winds, volcanic eruptions, earthquakes, and tsunamis. To increase the resilience of the agricultural supply chain, the sixth industrialisation is found to be effective and well established. Japan’s sixth industry contains 60,000 businesses—forming into new supply chains. Most of them are food processing and direct shop businesses. Direct shops are operated by farmers, farmers’ groups, farmers’ companies, municipalities, cooperatives, and producers’ groups. Japan’s agricultural policy also provided direct and indirect support at the national and local levels actions on resilience. Public policy supports the resilience of the agricultural supply chain through measures such as agricultural land use planning, agricultural improvement projects, rapid reconstruction following damage, and improvement of agricultural resources. Achieving the resilience of production networks during economic crisis is the focus of the Chap. 9. It analyzed the effects of rising labor costs, Chinese Yuan appreciation, and the 2008 Global Financial Crisis on the performance of Small and Medium Enterprises (SME)in China. Chinese small business has spearheaded the integration of the Chinese economy with the world economy through their involvement and partnership with the Asian regional and global supply chains. While they have acted as important linkages for the Chinese economy with the world economy, they have also become important conduits of the impact of world events on the national economy. This chapter addressed the impact of the 2008 Global Financial Crisis against the background of rising labor costs and currency appreciation that had already been exerting great strains on global value chains originating or ending in China. Chapter 10 brought new insights on stranded assets and protecting value of food value chain from disasters and other external shocks. Stranded assets are those that have suffered unanticipated or premature write-downs, lost value or turn into liabilities due to external shocks. Environmental risk factors such as natural disasters, climate change, water scarcity etc. that could cause asset stranding of agriculture are poorly understood in the context of food value chain (FVC) . The Value at Risk globally is significant in agriculture due to over-exposure to stranded assets throughout financial and economic systems. This chapter provides a critical evaluation of public policies of the environmental risk disasters and climate change as agriculture asset stranding in FVC. The impact of disasters triggered by natural hazards on

12

V. Anbumozhi et al.

the economic losses of the agricultural value chain and loss of value-added growth with further discussion on the principles of effective disaster risk reduction in FVC. Disaster, when combined with climate change, poses challenges by creating fluctuations in yields, supply shortfalls, subsequent global trading patterns, and substantial effects on FVC. Finally, it presents public policy strategies for building resilient FVCs in partnership with communities. The rationale, cost and benefits of a regional framework for advancing supply chain resilience and business continuity plan is the focus of the last Chap. 11. It analysed the direct, indirect and spill over effects of disasters and climate change on global value chains. Key economic trends and multi-dimensional challenges to disasters risk reduction and supply chain resilience is discussed with various levels of governance system and business continuity plans. It argued that supply chain resilience could be advanced through regional cooperation frameworks. Strengthening legal frameworks for improved coordination along the supply chain is critical for transferring the risks to market forces. Capacity of local business along the supply chain could further be enhanced by investing in innovative financial mechanisms including indexed based insurance instruments and catastrophic risk pooling at regional level in a coordinated way.

1.7 Conclusions The chapters in this volume support the conclusion that the economic shocks, disaster impacts and financial crisis has not reversed the globalization trends but increased the supply chain resilience. Regional supply chains in food, manufacturing and service sectors have emerged as a long-term structural feature of integration into world economy. However, important shifts in local production and demand have taken place and the disruptions due to natural disasters has accelerated pre-existing trends toward business and organizational consolidation. The compound effects of disasters and economic shocks has also underlined the growing importance of informing the upstream and end market players on the importance of risk management and hot spots of supply chain vulnerability. These supply chain disruptions also stimulate innovations towards resilience, along with existential challenges, but the benefits and barriers are unevenly distributed along the supply chains. In addition to the importance of these general patterns, industry dynamics matter for achieving resilience. It is essential to understand conditions in specific industry value chains in the post-crisis recovery stage, and the opportunities and challenges they create for new stakeholders to enter and upgrade within these global supply chains.

1 Global Supply Chain Resilience: Vulnerability and Shifting …

13

References Abe S, Thangavelu SM (2013) Natural disasters and Asia. Asian Econ J 26(3):138–147 ADB (2019) Asian development outlook 2019 strengthening disaster resilience. Asian Development Bank, Manila ADBI (2015) Disaster risk management in Asia and the Pacific. Asian Development Bank Institute, Tokyo Anbumozhi V (2015) Engendering resilient and sustainable ASEAN, in framing the ASEAN sociocultural community post 2015. Economic Research Institute for ASEAN and East Asia, Jakarta Ando M, Kimura F (2013) How did Japanese exports respond to two crises in the international production networks? The global financial crisis and the Great East Asia Japan earthquake. Asian Econ J 26(3) Battiston S, Delli Gatti D, Gallegati M, Greenwald B, Stiglitz JE (2012) Liaisons dangereuses: increasing connectivity, risk sharing, and systemic risk. J Econ Dyn Contr 36(8):1121–1141 Blos M, Quaddus M, Wee H, Watanabe K (2009) Supply chain risk management: a case study of automotive and electronic industries in Barzil. Supply Chain Manag Int J 14(4):247–252 Briano E, Caballini C, Revertria R (2009). Literature review about supply chain vulnerability and resilience. In: Proceedings of the 8th ESEAS international conference on system science and simulation in engineering, Miami, pp 191–197 Carvalho VM, Nirei M, Saito Y, Tahbaz-Salehi A (2016) Supply chain disruptions: evidence from the Great East Japan earthquake. Columbia Business School Research Paper Christopher M, Peck H (2004) Building the resilient supply chain. Int J Logist Manag 15(2):1–13 Gai P, Jenkinson N, Kapadia S (2007) Systemic risk in modern financial systems: analytics and policy design. J Risk Finance 8(2):156–165 Hamel G, Valikangas L (2003) The quest for resilience. Harvard Bus Rev 81:52–63 Juttner U (2005) Supply chain risk management—understanding the business requirements from a practitioner perspective. Int J Logist Manag 16(1):120–141 Korhonen J, Seager T (2008) Beyond eco-efficiency: a resilience perspective. Bus Strat Environ 11:411–449 Manuj I, Mentzer J (2008) Global supply chain risk management. J Bus Logist 29(1):133–155 Natarajarathinam M, Capar I, Narayanan A (2009) Managing supply chains in times of crisis: a review of literature and insights. Int J Phys Distrib Logist Manag 35(4):535–573 Peck H (2006) Reconciling supply chain vulnerability, risk and supply chain vulnerability, risk and supply chain management. Int J Logist Res Appl 9(2):127–142 Ponomorov S, Holcomb M (2009) Supply chain risk: a review and typology. Int J Logist Manag 20(1):124–143 Rao S, Goldsby T (2009) Supply chain risks: a review and typology. Int J Logist Manag 20(1):97–123 Saenz MJ, Revilla E (2014) Creating more resilient supply-chains. MIT Sloan Manag Rev Res Highlight. https://sloanreview.mit.edu/article/creating-more-resilient-supply-chains/. Accessed 13 Sept 2019 Schwarcz R (2008) Systemic risk, duke law school legal studies paper no. 163. Georgetown Law J 97(1). https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1008326. Accessed 13 Sept 2019 Sheffi Y (2005) The resilient enterprises-overcoming vulnerability for competitive advantage. The MIT Press, Cambridge Sheffi Y, Rice JB (2005) A supply chain view of the resilient enterprises. Sloan Manag Rev 47(1):41– 48 Smith R (2004) Operational capabilities for the resilient supply chain. Supply Chain Pract. 6(2):24– 35 Svenesson G (2000) A conceptual framework for the analysis of vulnerability in supply chains. Int. J. Phys. Distrib. Logist. Manag. 32(2):110–134 Wagner S, Bode C (2006) An empirical investigation into supply chain vulnerability. J. Purchas. Supply Manag. 12(6):409–430

14

V. Anbumozhi et al.

Watts DJ (2002) A simple model of global cascades on random networks. Proc Natl Acad Sci 99(9):5766–5771 World Economic Forum (2017) New models for addressing supply chain and transport risk, Geneva. http://www3.weforum.org/docs/WEF_SCT_RRN_NewModelsAddressingSupplyChai nTransportRisk_IndustryAgenda_2012.pdf. Accessed 13 Sept 2019

Chapter 2

Economic Shocks and Uncertainties: How Do Firm Innovativeness Enable Supply Chain and Moderate Interdependence Shandre Mugan Thangavelu and Venkatachalam Anbumozhi Abstract This chapter studies effects of economic shocks on the productive performance of manufacturing firms at Vietnam using the firm level data from 2004 to 2008. The chapter also examines the effects of economic shocks on the productive performance of firms in terms of access to linkages, spillovers, and ownership structures in terms of foreign, private and public ownership. The economic shocks such as global financial crisis tends to have greater impact on domestic firms and SOEs. The findings suggest that the local private enterprises did not receive positive foreign horizontal and backward spillovers from the MNEs, while SOE horizontal linkages also failed to contribute to the improvement in output productivity. Keywords Global financial crisis · Productive performance · Domestic firms · ASEAN value chain JEL Classification F60 · L14

2.1 Introduction The Vietnam economy is showing strong growth for the last decade. In fact, the key to strong growth of the Vietnamese economy is the liberalization policy of the government to enhance competitiveness of the domestic economy by opening it to trade, foreign competition and investment. Since its economic liberalization, the government has put in market friendly policy to attract foreign activities in the domestic economy. In 2007, Vietnam joined the WTO and hence increasing its participation in the global economy. Vietnam tends to have experienced an average real growth

S. M. Thangavelu (B) Jeffrey Cheah Institute for Southeast Asia, Sunway University, Selangor, Malaysia e-mail: [email protected] Institute for International Trade, University of Adelaide, Adelaide, Australia V. Anbumozhi ERIA, Jakarta, Indonesia © Springer Nature Singapore Pte Ltd. 2020 V. Anbumozhi et al. (eds.), Supply Chain Resilience, https://doi.org/10.1007/978-981-15-2870-5_2

15

16

S. M. Thangavelu and V. Anbumozhi

of around 7.1% from 2000 to 2011, which is much higher than the ASEAN 5 countries and it is only surpassed by recently liberalized economies of Cambodia and Myanmar. There are several factors that accounts for the sustained growth of the Vietnamese economy. Firstly, Vietnam has a growing and young labour force. The labour force literacy rate is nearly 97% in 2015. Secondly, Vietnam is rapidly advancing its technology sector and is already among the fastest growing technology markets in the world. Thirdly, the role of the government is also emerging as an important factor for economic stability of the Vietnamese economy. The pro-business approach of the government tends to attract significant foreign direct investment activities in the economy. Current economic policies were triggered by a series of reforms in the 1980s known as doi moi (new thought). The government is now more receptive to the involvement of foreign government in its domestic economy, especially in the key sectors such as the IT sector. The deregulation is carried out by restructuring state-owned enterprises into private enterprises and increasing foreign ownership in domestic industries. In terms of infrastructure, the government has devoted everincreasing resources to the establishment of several industrial parks including the most modern industrial parks in recent years. However, we also observe that as Vietnam is becoming more open to trade and investment from the global economy, it is also increasing expose to global and regional shocks due to its openness to trade and investment. Recent evidence indicates that Vietnam experienced the economic shock on its domestic economy from the Global Financial Crisis (GFC) in 2007–2008. Economic shocks such as Global Financial Crisis (2007–2008) has disruptive effects on the economic activities of firms. In particular, the ability of firms to survive economic shocks will depend on their flexibility and productive performance. This chapter examines the impact of economic shock on the productive performance of Vietnamese firms from the Global Financial Crisis. The study controls for endogeneity in the estimation by adopting the GMM estimation framework. The study also uses firm level data to explore the impact of the economic slowdown of the GFC on the economic and productive activities of the Vietnamese firms by specifically focusing on the crisis period of 2007–2008 and compare to the non-crisis periods. The purpose of this chapter is to investigate the economic shocks on the productivity performance of the Vietnamese manufacturing firms and this is done via a two-stage empirical strategy. We develop several key hypotheses in our study. Firstly, in an open economy, the impact of economic shocks affects the smaller firms more than larger ones. In fact, the larger ones have the investment capacity, economies of scale and scope to hedge the risk of external shocks. In contrast, the smaller ones are more vulnerable to the shocks due to smaller scale and lack of scope to move their operations and investments around. Secondly, linkages with foreign multinationals can enhance local SME development through beneficial linkages between foreign affiliates and domestic SMEs and this could help to smooth out and reduce the persistence of the negative effects of external shocks on the SMEs. This study has several key features. Firstly, with the use of micro-level panel data of 4146 firms from the Annual Statistical Censuses and Surveys during the period of

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

17

2004–2008, we employ the Generalized Method of Moments (GMM) estimation of total productivity factor (TFP) to control for the possible endogeneity of production inputs. In doing so, we also address several gaps in the literature. Most studies on Vietnamese industries tend to focus at industry-level data which would not control for time-specific and firm-specific differences in TFP. In our study, we introduced horizontal and backward foreign linkages in our empirical model, along with firmspecific characteristics such as quality of labor, and industry-level variables such as the presence of SOEs and level of competition. Econometric issues such as heteroskedasticity, unobservable firm-specific characteristics and endogeneity biases of the control variables are also controlled for to ensure robustness of results. The current paper follows closely the study on the Vietnamese firms by Thangavelu (2014) that examined the productive performance of large and small firms in Vietnam manufacturing sector. The results indicate that there are weak backward and horizontal linkages between foreign and domestic firms in the Vietnamese manufacturing sector. This study focuses on the impact of economic shocks on the productive performance of Vietnamese firms. The rest of this paper can be outlined as follows. Section 2.2 provides an overview of the development in Vietnam. Section 2.3 details data construction and measurement. Section 2.4 estimates the productive performance of firms using two stage estimations: (a) estimating the firm level TFP and (b) identifying the sources of productive performance such as linkages and spillovers. Section 2.5 presents the parameter estimates and discusses the main findings. Section 6 concludes with some policy implications.

2.2 The Key Macroeconomic Trends of Vietnam Vietnam transited into a market economy in the early 1990s via the Doi Moi Policies (Economic Reform Policies), which facilitated the development of trade and also increased the inflows of FDI through initiatives such as the promulgation of Law on Foreign Investment as well as membership into ASEAN, APEC and WTO. Since then, Vietnam has experienced rapid GDP growth from openness strategies and FDI inflow. Vietnam’s economy has consistently achieved a high rate of economic growth, in addition to improved standards of living and rapid poverty reduction. During the period 2000–2010, the economy enjoyed an impressive GDP growth rate of 7.2%— the second highest among ASEAN+3 countries following China.1 The accelerated pace of economic growth is fueled largely by growth in the manufacturing and construction sectors which accounted for approximately 40% and realized the value added growth of 10.6%, on average, during the same period.

1 The

figure of the average GDP growth rate is calculated from World Development Indicators, the World Bank.

18

S. M. Thangavelu and V. Anbumozhi 16.0

14.0

Growth Rates (%)

12.0

10.0 GDP Agriculture Industry Services

8.0

6.0

4.0

2.0

0.0

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Fig. 2.1 Real GDP and sectoral growth rates (%) of Vietnam: 1994–2011. Source ADB statistics

The key to strong growth of the Vietnamese economy is the liberalization policy of the government to increase the competitiveness of the domestic economy by opening it to trade, foreign competition and investment. The effects of liberalization of the Vietnamese economy are reflected in terms of real GDP growth at Fig. 2.1. Vietnam tends to have experienced an average real growth of around 7.1% from 2000 to 2011, which is much higher than the ASEAN 5 countries and it is only surpassed by recently liberalized economies of Cambodia and Myanmar. The real growth rate peaked before the Global Financial Crisis at 8.4 in 2006. However, we observed a downward trend in real GDP after the Global Financial crisis in 2008, where the average growth rate was 5.9% from 2008 to 2011. In fact, the global financial crisis in 2007–2008 indicates a structural shift in the economy as there is a clear downward shift in the trend of the growth rate for all sectors in the Vietnamese economy. The growth of the Vietnamese economy also reflects the rising importance of manufacturing for the domestic economy. Figure 2.1 clearly shows the importance of the manufacturing and services sector in the overall growth of the domestic economy. The industrial sector grew at the annual average rate of 10% in 1994–2011 and nearly 8.5% from 2000 to 2011. The services sector grew at the annual average rate of 7.1% from 1994 to 2011, but in it has been showing an upward trend since 2000. In contrast, the agricultural sector is just growing at the average annual rate of 4% since 1994. The declining share of agriculture and the rising share of manufacturing are reflected in Fig. 2.2. It is clear from Fig. 2.2 that the rising of share of manufacturing is with concurrent decline in the share of the agricultural sector. The share of manufacturing to GDP ratio has been rising from 22% in 1990 to over 40% in 2011, and concurrently, we observed that the agricultural sector has been declining to 22% in 2011 from over 39% in 1990. In contrast, the share of services sector to GDP has been remaining steady at 38% from 1990 to 2001. This clearly indicates

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

19

50.0 45.0 40.0

Share to GDP (%)

35.0 30.0 Agriculture Industry Services

25.0 20.0 15.0 10.0 5.0 0.0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Fig. 2.2 Share of key sectors to GDP ratio for Vietnam: 1994–2011. Source ADB statistics

the importance of the industrial and services sectors in sustaining the overall growth of the Vietnamese economy. The rising share of industrial sector is reflected by two key observations. The rising share of industrial sector is highly correlated with the rising share of gross domestic capital formation. The share of gross domestic capital formation to GDP ratio shows a rising trend as it increased to over 40% in 2007 from 12% in 1990 (ADB statistics). It is important to highlight that the economic liberalization of Vietnam is mainly driven by growth in global trade. The share of export to GDP increased to 87% in 2011 from 26% in 1990. The impact of openness is also observed with the rising share of imports to GDP, where it increased from 36% in 1990 to nearly 91% in 2011. The rising trend of the imports suggests that Vietnamese firms (especially the foreign firms) might be increasing their outsourcing activities in the domestic economy. The rising share of imports to GDP ratio clearly indicates that the Vietnamese economic liberalization have reduced the barriers to trade in terms of import tariffs and tax on capital goods. The effects of this liberalization are the rising share of imports to GDP, where domestic firms are likely to outsource some of their key services and other activities to the global production value-chain. The rising share of imports and hence outsourcing is given at Fig. 2.3, where the share of imports increased from 36% in 1990 to nearly 91% in 2011. The impact of global financial crisis is also apparent in FDI inflows in the Vietnamese economy. The trends of GDP growth and FDI inflows as a percentage of GDP in Vietnam since the transition in 1990 are given at Fig. 2.4. The relationship is positive and strong up till the end of 1990s. However, it is also important to note

20

S. M. Thangavelu and V. Anbumozhi 100.0 90.0 80.0

Share to GDP (%)

70.0 60.0 Exports of goods and services

50.0

Imports of goods and services

40.0 30.0 20.0 10.0 0.0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Fig. 2.3 Share of export and import to GDP for Vietnam (%) from 1994 to 2011. Source ADB statistics

Fig. 2.4 Trends of GDP growth (annual %) and FDI inflows (% of GDP) in Vietnam from 1990 to 2011. Source World Development Indicators (WDI), the World bank

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

21

Fig. 2.5 Trends of Annual Growth (%) of Manufacturing Sectors in Vietnam from 1990 to 2011. Source World Development Indicators (WDI), the World bank

that the correlation is less apparent thereafter; for instance, while FDI inflows grew steeply as a percentage of GDP from 2006 to 2008, GDP growth slowed. Thus, this calls to question the assumed positive relationship between FDI inflows and economic growth. In fact, the FDI inflows have steeply declined since the global financial crisis as indicated at Fig. 2.4. In addition, the manufacturing sector is also the key recipients of these FDI inflows across the industries. According to the Foreign Investment Agency (FIA) in Vietnam, the processing and manufacturing industries received the most newly and additionally registered FDI capital in 2012, accounting for 65.5% of the total FDI. Figure 2.5 shows the value added growth of the manufacturing sectors since the Doi Moi policies facilitated the FDI inflows. From the beginning of 1990s, the annual growth largely remained above 8%, except for the decline during the global recession in 2009. This provides preliminary signs of a positive correlation between the entry of foreign investments and the output productivity of enterprises. The positive correlation reflects the need for more semi-skilled and skilled workers as the economy transit to more capital-intensive and skilled intensive activities. Recent study by the World Bank (Vietnam Development Report 2012) reports the importance of the declining labour productivity growth for Vietnam and its impact on sustaining the economic growth momentum in the region. The labour productivity is fairly stable for Vietnam and then we observe a downward trend after the Global Financial Crisis. The average labour productivity has been around 4.9% from 2000 to 2007 and it declined to nearly 3.2% in 2008–2011. Although the decline in labour productivity in the post-crisis period is of a concern, as compared to other selected ASEAN countries, the productivity for Vietnam is quite stable and shows similar trend as to other ASEAN countries.

22

S. M. Thangavelu and V. Anbumozhi

2.2.1 Linkages Between Foreign and Domestic Firms Production linkages are important conduits for the positive impact and spillovers of multinational activities in the domestic economy. MNEs and foreign affiliates typically have more advance technology and better distributional networks than domestic firms in developing countries, which creates a potential for productivity spillovers on domestic firms when different production linkages are formed with their foreign counter-part. As aforementioned, this paper focuses on horizontal and backward production linkages. Horizontal linkages have been widely researched on and positive productivity spillovers through such intra-industry relationship can occur through 4 channels—(a) competition effects, (b) demonstration effects, (c) labour mobility and (d) exports. The first channel refers to the entry of foreign firms into the domestic market as a form of competition with the domestic firms. As a result, domestic firms are incentivized to enhance productivity through better utilization of resources and usage of more advanced technology, thereby creating positive competition effects. However, as Aitken and Harrison (1999) suggested, domestic firms’ market share can also be eroded by the entry of large foreign firms, especially when there is imperfect competition in the product market. Consequently, the competition effects become negative as firms either function with less efficiency due to higher average operating costs or exit the market. On the other hand, demonstration effects occur when domestic firms adopt advanced technology or imitate better practices used by foreign firms, which subsequently improved their productivity. Similarly, domestic firms may also tap on knowledge and expertise of workers previously from MNEs for improving their productivity. Görg and Strobl (2001) did a relevant empirical investigation and found that owners of domestic firms who had worked in an MNE immediately prior to starting their firms in the same industry were more productive than their counterparts without the MNE experience. But as Meyer and Sinani (2009) highlighted, such labor mobility can be limited if foreign firms offer higher wages and attract skilled labor from domestic firms instead. In such cases, the entry of foreign firms may further drain the level of human capital in local companies. Lastly, the presence of MNEs and foreign affiliates can provide distributional networks and relevant knowledge which facilitate export performance. Hence, with horizontal linkages with the foreign firms, domestic firms can boost their export capacity and productivity levels as well. Vertical linkage had been mainly neglected in the earlier part of the empirical research but it is increasingly emphasized, as recent studies find positive and statistically significant vertical spillovers despite non-significant horizontal spillovers from FDI (Smarzynska and Beata 2002). This is especially so for backward linkages. Similar to horizontal linkages, they can facilitate positive productivity spillover through the demonstration effect, competition effect, and labor mobility.

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

23

2.2.2 The Context of Vietnam As a transitional economy with relatively new to FDI, Vietnam has lower levels of technology and smaller skilled workforce as compared to developed countries. Therefore, the local enterprises have a great potential to benefit from productivity spillovers brought about by production linkages with foreign counterparts. This is highlighted in Nguyen et al. (2008), where large FDI inflows had mainly entered the industrial sectors and were restricted in the form of joint ventures with state-owned enterprise before the 1997. In particular, the growth rate of industrial output produced by these FDI enterprises mostly exceeded the growth rate of the entire industrial sector from 1995 to 2003. Therefore, their greater level of productivity would impact positively on local firms. However, many studies on Vietnam have shown that the presence and magnitude of spillovers are dependent on the characteristics of the firms in the sample, as well as the industry and country examined. Hence, it is important to examine the methodologies and diverse findings from these studies so as to address the gaps and draw relevant conclusions for the empirical results derived in this paper. Giroud conducted semi-structured interviews and found that initial linkages formed in Vietnam were weak and productivity spillovers were not as extensive as Malaysia due to lack of collaborative schemes and large technology gap between foreign and domestic firms. For example, foreign firms may have demand for higher quality inputs which domestic suppliers with limited technology capacities cannot produce. Hence, the backward linkages are not formed and productivity spillovers are limited. On the other hand, domestic-market-oriented FDI also enter the Vietnamese market with an advantage over domestic firms in terms of technology and knowledge. Consequently, this negative competition effect led to domestic firms experiencing a negative horizontal spillover. However, Nguyen et al. (2008) observed an upward trend in production efficiency of domestic manufacturing firms in Vietnam, likely due to gradual demonstration effects. This suggests that Vietnamese firms are gaining the ability to meet the demands of foreign firms and compete with MNEs in the same sector. Therefore, their improving absorptive capacity for productivity spillovers may increase the spillover effects from production linkages across the years. In fact, subsequent empirical studies have shown that backward spillover effects are the dominating type of positive spillover in Vietnam, whether the spillover effects are in the form of labor productivity, output productivity or wages. However, the results for horizontal spillover effects remained mixed and inconclusive as it mainly depended on the aspect of spillovers examined and the empirical specification used. Firms’ heterogeneity constitutes an important part of the analysis as many studies included control variables at firm-level or even at industry-level to investigate the possible determinants of spillover effects. The existing technology gap between domestic firms and their foreign counterparts remain an important part of many analyses on Vietnam as it consistently predicted negative spillover effects for domestic firms. Hence, local enterprises with significantly low technological capacity would not be able to reap positive productivity spillovers as they are unable to imitate and utilize the advance technology and processes used in foreign firms. Similarly,

24

S. M. Thangavelu and V. Anbumozhi

the level of technology required in different sectors is also related to the relative technology gap between domestic and foreign firms, thereby resulting in different spillover outcomes. For example, Le Quoc Hoi (2008) found that only domestic firms in low-technology industries experienced negative horizontal spillovers on labor productivity, which suggests that negative competition effect play an important role in these industries as compared to the positive demonstration effects facilitated by technology gap. Moreover, only medium- and high-technology industries experienced positive backward spillovers which implied that these industries have moderate technology gap and are able to benefit from the transfer of technology and knowledge from foreign firms in downstream sectors. The scale of firms as a firm-specific factor was also found to be influential for the spillover effects on domestic firms in Vietnam. In Nguyen et al. (2008), larger hightechnology domestic firms have more opportunities to receive technology transfers from foreign firms than its smaller counterparts. In the study by Le Quoc Hoi (2008), they observe larger domestic firms are able to benefit more in terms of backward productivity spillovers.

2.3 Data Construction and Empirical Methodology We construct our dataset of firms from Annual Statistical Censuses and Surveys: Enterprises from 2004 to 2008, gathered by the General Statistics Office of Vietnam. It provides firm-level information on foreign ownership and production characteristics, like the number of workers, gross revenue, working capital, materials, profits, and export/import status, on top of financial attributes such as liquid asset, fixed asset, liabilities and equity, among many others. Firms are differentiated into three categories, (i) domestic-owned, if there is an absence of state and foreign capital, (ii) state-owned, if the enterprise owns central state or local state capital and (iii) foreign-owned, when there is the presence of foreign capital in the firm. This classification provides nearly 1,446 domestic firms, 890 foreign firms, and nearly 1,810 state-owned enterprises. The data is rebased to 2005 prices using the producer price index. In total, the panel data consisted of 20,730 observations and span across 23 manufacturing sectors based on the Vietnam Standard Industrial Classification 2007 (VSIC 2007)2 . Firms are differentiated into three categories, (i) domestic-owned, if there is an absence of state and foreign capital, (ii) state-owned, if the enterprise owns central state or local state capital and (iii) foreign-owned, when there is the presence of foreign capital in the firm. It is important to highlight that we examine the impact of economic shocks such as GFC (global financial crisis) by analysing pre- and post- effects of GFC. Our empirical strategy includes the following. First, the measurement of TFP rests with an estimation of a Cobb-Douglas production function which requires information 2 VSIC

(2007) is based on International Standard Industrial Classification revision 4 (ISIC Rev.4).

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

25

on a firm’s gross output as well as production inputs. Net output is measured by sales of goods produced net of materials and components purchases. There are three production inputs in the empirical model, labour, intermediate materials, and capital. Labour is the number of labour employed within a firm. Intermediate materials include parts and components that are used in the production processes. Capital is the values of land, building and construction, and machinery and equipment, less the depreciation of assets. All variables are deflated using GDP deflators in 2004 prices.3 Several studies have highlighted the weakness of using the Ordinary Least Square (OLS) estimations for the measurement of TFP, it has been pointed out that the estimators might be biased since the OLS method assumes that the input levels are exogenous. Studies including Griliches and Mairesse (1998), Girma et al. (2001), Lesher and Miroudot (2008) have pointed out that productivity shocks observable by firms may affect both their decisions for inputs level and the respective firm’s TFP, thereby creating a simultaneity problem where the input variables in the OLS estimation are endogenous. Hence, to address this issue, the two-step Blundell-Bond GMM estimation was employed instead. The simplest way to obtain parameter estimates in our base-line econometric specification (3) is to carry out the standard Ordinary Least Squares (OLS) estimations. However, our concern is that OLS estimations tend to convey biased estimates due to firm heterogeneity. The unobservable firm heterogeneity seems plausible given the knowledge that firms operate in a wide range of economic activities like manufacturing, financial intermediation, trade, real estate and consultancy services. To control for unobservable firm heterogeneity, we make use of Fixed Effects (FE) and Random Effects (RE) estimations. The former is undertaken by using OLS with heteroskedasticity-robust estimators to take into account the heteroskedasticity problem that arises from variation in firm size, whereas the latter is obtained by Generalized Least Squares (GLS) with the Swamy-Arora estimators. However, FE and RE estimates may also be biased and inconsistent. The reason is that all of our structural variables, e.g. FDI, financial characteristics, high-tech capital investment, and human capital utilization are very likely to be endogenously determined by other unobserved variables. If the potential endogeneity bias problem exists, FE and RE estimates are not consistent and asymptotically efficient. We address the issue of endogeneity in our study by adopting the Generalized Method of Moment (GMM) of two-step estimation to correct any simultaneity bias in the estimations (Blundell and Bond 2000; Arellano and Bover 1995). In conventional first-differenced GMM estimations, variables are first-differenced to remove the firm-specific time-invariant characteristics. Subsequently, the firstdifferenced equations are instrumented by lagged structural variables to isolate changes in production inputs caused by endogenous factors. However, Blundell and Bond (2000) pointed out that these estimators may yield unsatisfactory results if the lagged structural variables used are weak instruments due to finite-sample bias. Hence, appropriate for our data sample, the authors proposed two-step GMM estimation useful for a large sample across few time periods. Essentially, they introduced 3 GDP

deflators are constructed using information available from World Bank.

26

S. M. Thangavelu and V. Anbumozhi

an additional step by adding lagged first differences of structural variables as IVs in the level equation, which increased the moment conditions in the estimations and the precision of estimates. In addition, the authors allow for first order serial correlation in the error terms in order to obtain any valid lagged internal instruments for equations in first-differences or equations in levels. Therefore, the specification for the firm’s production can be modified as such4 : yit = β0 + βl lit + βk kit + βm mit + αt + ηi + vit + mit ,

(2.1)

where ηi , vit and mit are the additive components of the error term and represent unobserved firm-specific effect, productivity shock (potentially autoregressive) and serially uncorrelated measurement errors, respectively. The two-step Blundell-Bond GMM estimation serves to isolate effects of unobserved firm-specific effect and productivity shock through the use of IVs to resolve the endogeneity issue for the production inputs. The effects of linkages on output productivity of domestic firms are then examined through the regression of the estimated TFP against the production linkages as well as the respective control variables. The key variables of our study are the two types of production linkages foreign firms form with their domestic counterparts. The foreign horizontal linkage (FOR_HORZ) variable5 aims to measures the presence of foreign firms in a particular manufacturing sector and is defined as the share of sales of foreign firms in that sector. Such measurements were also used in Girma et al. (2001), Nguyen et al. (2008) and can be written as follows: FOR_HORZjt =

 ∀j = i

yj,t /Yi,t ,

where yj,t represents the output of foreign firm i, operating in sector j at time t and Yjt is the total output of sector j at time t. Hence, the FOR_HORZ variable increases with rising output share of the foreign firms. The foreign backward linkage (FOR_BACK jt ) variable serves to capture the extent of potential contacts between foreign firms and domestic suppliers, and akin to Smarzynska and Beata (2002), Girma et al. (2001), it is defined as: FOR_BACKjt = Σk αkj FOR_HORZkt for k = j,

4 The

econometric specification is adapted from Blundell and Bond (2000). and Beata (2002), Pattnayak and Thangavelu (2011) used the foreign equity participation averaged over all firms in the same sector (weighted by each firm’s share in sectoral output), which would be more sensitive to the presence of foreign investment in the industry. However, the data limitation in the dataset only allowed us to capture the horizontal linkage as the foreign firm’s share in sectoral output.

5 Smarzynska

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

27

where α kj is the proportion of sector j output supplied to sector k.6 Hence, the backward linkage variable increases with rising foreign presence in sectors supplied by industry j and increasing share of intermediates supplied to sectors with foreign presence. According to the mixed results from the current literature, it is difficult to form any reliable hypotheses for the empirical outcome of the four linkage variables. However, given strong inclination for positive foreign backward spillovers from FDI in many empirical studies on Vietnam seen in Sect. 2.2, this paper proposes that foreign backward linkages are likely to lead to positive productivity improvement. Since horizontal spillovers are often dependent on the trade-offs of positive demonstration effects and negative impacts of competition, the overall influence appears to be ambiguous. In the context that domestic firms are able to improve their technical efficiency progressively across the years and gain more competitiveness, the effects from demonstration and imitation of technology will over-compensate for the negative competition effects. In such instances, we would therefore observe positive foreign and SOE horizontal spillovers. Additionally, firm-specific characteristics are important in accounting for the presence and size of spillover effects on productivity as discussed in Sect. 2.2. In this study, we included a proxy for quality of labor in the empirical framework. While studies have used the ratio of skilled workers as a measurement of labour quality, this information is not available in our dataset. Hence, as suggested in Le Quoc Hoi (2008), the average wage of a firm is used a proxy instead, with the assumption that firms with higher average labour costs per worker employ higher skilled labour. The variable (Labour_Qijt ) is measured as such: Labour_Qijt = Wijt /Lijt where W ijt refers to the total wages paid in firm i, industry j at time t while L ijt refers to the total number of employees in firm i, industry j at time t. This variable aims to capture the quality of human capital in each domestic firm. It is predicted that firms with higher quality of labor is likely to have greater output productivity due to increased efficiency. An industry-level characteristic is examined through the Concentration variable (CONC jt ) and intends to capture the effects of industry concentration and competition. It is proxied by the Herfindahl index7 as:  2 CONCjt = Σi xijt /Xjt 6α

kj is constructed with the use of an input-output table on Vietnam in early 2000s, retrieved from http://stats.oecd.org. 7 The Herfindahl index is a concentration ratio which captures the level of competition in a market or industry by comparing market shares of firms using the relative firm size. A high Herfindahl index indicates the presence of firms with large market shares and hence, a lower level of competition in the industry. Correspondingly, a low Herfindahl index indicates firms each having low market share and thereby, implying a high level of competition.

28

S. M. Thangavelu and V. Anbumozhi

where x ijt is the sales of domestic firm i in industry j; X jt denotes the total sales of industry j. A higher value of the Herfindahl index indicates a high degree of industry concentration and thus, the presence of big firms withholding large market shares. Hence, it is predicted that a higher value of Herfindahl index is likely to have a negative impact on the productivity of domestic firms as they are unable to compete with larger firms. The study also controls for trade by deriving the effective rate of protection (ERPs) for  firms. The ERPs are measured using the formula ERPj = [t j −  Vietnamese aij t i ]/[1 − aij ] where t j is nominal tariff on sector j, aij is the share of intermediate input i in the final value of product j that is obtained from the input-output tables. Thus, ERPs depend on both tariffs for final products as well intermediate products (see Thuyen et al. for Vietnamese ERP). The estimated model can be represented by the econometric specification as follows: TFPijt = α0 + α1 FOR_HORZjt + α2 FOR_BACKjt + α4 LABOR_Qijt + α4 CONCjt + δt + δj + uit

(2.2)

where the subscript i, j and t refer to firms, industries and time respectively. δt and δj are the time and industry dummies, respectively, and uijt denotes the stochastic error term in the regression model. However, before proceeding to the estimation, the empirical strategy must also address three possible econometrics issues which arise if OLS estimation is utilized for specification (3). First, there can be heteroskedasticity in the stochastic error term (uit ) which will subsequently lead to biased estimates. Hence, this study employs heteroskedasticity-robust procedure for all estimations. In addition, there can be considerable unobserved firm-specific heterogeneity given that the firms span across the various segments of the manufacturing industry. Hence, Fixed Effects (FE) and Random Effects (RE) estimations are used to control for such time-invariant firmspecific effects. In particular, FE estimates are obtained via OLS estimation whereas RE estimates are derived using the Generalized Least Squares (GLS) estimations. To further compare the two types of estimates, the Breusch-Pegan test8 is carried out under the null hypothesis that there are no random effects. Thus, the specification can be re-written as such: TFPijt = α0 + α1 FOR_HORZjt + α2 FOR_BACKjt + α3 LABOR_Qijt + α3 CONCjt + δt + ηi + uit

(2.3)

where ηi represented the unobservable time-invariant firm-specific effects. The third concern involves the possible endogeneity of the explanatory variables as they might be determined by unobserved variables. In such cases, the FE and RE 8 It

tests for heteroskedasticity in a linear regression model by checking whether the estimated variance of the residuals from a regression is dependent on the values of the independent variables. If the estimated variance is dependent on the covariates, FE estimates are preferred.

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

29

estimates will be biased. Hence, to address this issue, the two-step Blundell-Bond GMM estimation was employed again. So the final econometric specification can be written as follows: TFPijt = α0 + α1 TFPijt−1 + α2 FOR_HORZjt + α3 FOR_BACKjt + α5 LABOR_Qijt + α6 CONCjt + ηi + vit + mit

(2.4)

where TFPijt−1 is included to account for the dynamic adjustments of the TFP in time period, t. Similar to the TFP estimation, ηi , vit and mit are the additive components of the error term. ηi and mit represent unobserved firm-specific effect and serially uncorrelated measurement errors, respectively. vit refers the unobserved variables which determine the explanatory variables. Two additional robustness checks are undertaken: The Sargan statistics9 test is undertaken to test the null hypothesis that the over-identifying restrictions are valid and the Arellano-Bond (AR) Test examines the null hypothesis of no serial correlation.

2.3.1 Descriptive Statistics and First Stage Regression—TFP Before proceeding to the econometrical tests, it is useful to perform preliminary descriptive analysis on the firms in the sample.

2.3.2 Estimations of Production Technology The results of the TFP measurement is given at Table 2.1, which provide the results of the estimation of the Cobb-Douglas production function. The first panel reports the OLS estimates with heteroskedasticity-robust estimators. However, as aforementioned, OLS estimates are likely to be biased due to potential endogeneity of input levels. For instance, in the context of a positive productivity shock which simultaneously affects both the production input choices and output levels, the input coefficients are likely to be biased upwards in OLS estimation. Therefore, to control for these biases, the second panel reports the GMM estimates, where the lagged dependent variable is used as a regressor and the lagged input variables are chosen as IVs. The input coefficients are lower than the OLS estimates and this suggests that there is likely to be simultaneity biases in the OLS estimation. Therefore, we adopt the GMM-estimated TFP for subsequent empirical analysis.

9 Also

known as Hansen test, it tests for the validity of instrumental variables used by checking for correlation between the residuals and exogenous variables to affirm the exogeneity of the instrumental variables.

30

S. M. Thangavelu and V. Anbumozhi

Table 2.1 Estimations of production technology by OLS and GMM for manufacturing firms in Vietnam: 2004–2008

Dependent variable: yit

OLS

Two-step GMM

Labor, l it

0.6179***

0.3187***

(0.0089)

(0.0307)

0.2580***

0.0209***

(0.0056)

(0.0077)

Capital, k it

0.2305***

0.1281***

(0.0067)

(0.0193)

Total

1.106

0.4758

Number of Obs.

16,172

13,139

Material inputs, mit

Note (1) Heteroskedasticity-robust standard errors in parentheses, (2) ***, ** and * denote statistical significance at 1, 5, and 10%, respectively

Firstly, a comparison is done among the foreign firms, state-owned enterprises and local enterprises with respect to their firm-specific characteristics. The descriptive statistics are presented in Table 2.2. Quality of Labor—Skilled workers require higher wages than low-skilled workers. Therefore, the average wage in a firm is an indicator for the level of human capital in a firm as firms with relatively more skilled workers are likely to also pay higher average wages. Correspondingly, average wage is used as a proxy for the quality of labor in each firm, which in turn signals the firm’s level of productivity and ability to compete with its counterparts in the same industry. In our sample, the labor quality of domestic firms is below average while foreign enterprises comparatively employ higher quality labor. Hence, MNEs may impose a negative competition effect on the domestic firms as they gain a competitive edge and enjoy higher productivity. Labor Mobility—Labor mobility across firms is dependent on the comparative attractiveness of the firms. An indicator of a firm’s appeal is the relative wage level offered to employees of similar qualifications, and it is often observed that MNEs offer higher wages than local private enterprises. In this case, the relative wage Table 2.2 Firm-specific characteristics by type of ownership Type of ownership

Quality of labour* (mil. Dongs)

Wage to sales ratio* (%)

Employment growth (% p.a.)

High-technology investment*

Total factor productivity

Total

16.4

0.18

10.68

0.11

−0.0014

Domestic

11.7

0.19

10.55

0.09

−0.0194

Foreign

24.4

0.24

11.35

0.13

−0.0372

State-owned

16.3

0.16

10.53

0.11

0.0168

*Labor quality is measured as the average wage in each firm. Wage level is proxied by wages as a proportion of total firm sales. High-technology investment is taken as the number of computers per employee

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

31

level is captured by the ratio of wages to firm’s total sales, which proxied the firm’s willingness to pay for each dollar of labor output. Hence, enterprises which offer higher wages to employees of similar caliber will have greater ratio of wages to their total sales. In our sample, the foreign firms have the highest ratio and therefore, they may have a draining effect on domestic and state-owned firms by better attracting more skilled workers. This limits the positive effect of labor mobility for which production linkages can facilitate. This trend is also consistent with the labor growth observed across the firms as foreign enterprises have faster labor growth than its domestic and state-owned counterparts. High–technology Capital Accumulation—Accumulation of High–technology capital contributes to operating performance, research and development, and ultimately, improved productivity. While the dataset lacks information on the expenditure on all high–technology capital in the firms, a proxy can be constructed to examine the trends amongst firms of different ownership. In this case, the number of computers available in the firm per employee is used to compare the incentive for innovation and efficiency. Foreign enterprises display the highest average while domestic firms have the lowest mean. However, greater high-technology capital accumulation does not necessarily translate into higher TFP. In the last panel, we see that SOEs has the highest average TFP despite fewer numbers of computers per employee than foreign firms. In fact, foreign enterprises have the lowest average TFP in our sample while domestic firms fared slightly better. Hence, with larger technology gap from SOEs, domestic firms may be able to receive greater productivity spillovers from the technology and knowledge transfers from SOEs than foreign companies.

2.3.3 Impact of GFC on Productivity of Large and Small Firms We decompose the TFP measure by SMEs, large, state-owned, domestic and foreign owned firms. The decomposition is given at Table 2.3. It is clear that larger and Table 2.3 The productivity (TFP) of Vietnamese firms: size, local, state, and foreign—2004–2008 Small

Large

State

Domestic

Foreign

2004–2006

0.26

−0.28

0.16

−1.80

1.66

2007–2008

−0.12

0.01

−0.01

−1.60

1.67

2004–2006

−5.20

5.30

−3.40

0.90

2.50

2007–2008

−8.00

8.20

−3.50

0.60

3.10

OLS

GMM

Source Author’s calculation Notes Firms size is as small if less than 100 workers

32

S. M. Thangavelu and V. Anbumozhi

foreign firms have higher TFP during the GFC and it might be due to their flexibility to adjust and implement new technologies. It is also likely that foreign firms might have larger scale to be more efficient during economic crisis. In contrast, SOEs have experienced negative TFP pre-crisis and also during the economic crisis. The negative TFP from 2007 to 2008 indicates that there is less creative destruction to effectively adjust to the more competitive equilibrium. This might be due to greater government support that might be given to sustain their activities. We also observe that SMEs are greatly affected by the economic and financial crisis and this might be due to lack of investment and lack of access to financial markets such as weak financial collaterals due to the financial crisis.

2.4 Empirical Results of GFC on Productive Performance of Firms The results of GFC on the productive performance of firms are given at Table 2.4. We provide the OLS and GMM analysis. The first panel provides the OLS estimates with the heteroskedasticity-robust estimators. However, due to unobserved firm-specific differences and endogeneity of control variables, OLS estimates are inclined to be biased. Therefore, the second and third columns report the fixed effects (FE) and GMM to control for effects of firm heterogeneity respectively. Across the specifications, estimations for the linkage variables are not consistent which suggests that there are likely to be issues of firm heterogeneity and endogeneity biases. This is addressed with GMM estimations. We also observe that FE estimates are qualitatively similar to GMM estimates in Table 2.4, which suggests that any endogeneity biases did not qualitatively bias the FE estimates. However, to ensure the robustness of the estimates, the remaining discussions are focused on the GMM estimations of GMM-TFP to ensure firm heterogeneity and endogeneity biases are fully controlled. The results from Table 2.4 indicates that human capital and labour quality are very important for adjustment during an economic crisis. The positive and statistically significant coefficient indicates that stronger human capital allows the firms to adjust to the economic shock much better. the coefficient of quality of labor is found to be positive and statistically significant across all specifications. This highlights the importance of investment in human capital in local firms to improve their output productivity. Consistent with Havranek and Irsova (2011) and Nguyen et al. (2008), this suggests that higher levels of human capital facilitate innovation and imitation of technology and expertise from MNEs and SOEs. On the other hand, the industry-level attribute (level of concentration) is shown to be negative and statistically significant for both the FE and GMM estimations. This provides evidence that high level of concentration in an industry would favor larger firms and disadvantage firms with small market shares, which subsequently impact negatively on the latter’s productivity. This is especially true for industries where majority of market shares

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

33

Table 2.4 Regression results of GFC on productive performance of Vietnamese firms: 2004–2008 Before economic shock (2004–2006)

Economic shock (2007–2008)

OLS

FE

GMM

OLS

FE

GMM

Lagged TFP

0.552*** (20.02)

0.774*** (65.44)

−0.259*** (−6.770)

1.178*** (47.140)

−0.158*** (8.570)

−0.054 (−0.840)

Labour quality

0.010*** (5.210)

0.009*** (18.870)

0.014** (2.740)

0.012*** (22.910)

0.006*** (5.720)

0.002*** (12.540)

Herfindahl index

−0.297** (−2.040)

−0.259* (−1.740)

−0.697* (−1.760)

−0.244* (−1.740)

−0.399** (−2.460)

−0.344** (−2.140)

Horizontal linkage

0.012 (0.270)

0.054 (0.900)

0.0114 (0.760)

0.521 (0.480)

0.103 (0.520)

0.011 (0.030)

Backward linkage

−0.308** (−4.860)

−0.219** (−2.980)

−0.559* (−1.950)

−0.418*** (−4.570)

−0.480** (−4.320)

−0.450** (−3.520)

Effective rate of protection (ERP)

−0.001** (−4.860)

−0.014** (−3.880)

−0.001*** (−3.570)

−0.020*** (−4.260)

−0.644 (3.490)

−0.192** (−2.080)

Domestic

0.001** (2.100)

0.050** (2.140)

0.730** (2.170)

0.001* (1.870)

0.647** (2.520)

0.144** (2.390)

Foreign

−0.020 (−1.090)

0.011 (0.560)

0.693** (2.030)

0.011 (0.560)

0.470** (2.520)

0.144** (2.360)

Firm size

0.148*** (8.060)

0.059** (3.540)

0.118** (2.310)

0.141*** (7.200)

0.483** (2.680)

0.278*** (6.640)

Dummies: time

Yes

Yes

Yes

Yes

Yes

Yes

Dummies: industry

Yes

Yes

Yes

Yes

Yes

Yes

Constant

−0.014 (−0.8340)

−0.010 (−0.270)

−0.963** (−3.290)

−0.040 (−0.160)

−0.283 (−1.430)

0.160** (2.140)

Obs

5135

5135

5135

5135

5135

5135

R-square

0.647

0.550

0.644

0.647

0.648

0.643

*, **, and ***: 10, 5, and 1% level of statistical significance

is dominated by large foreign or state-owned enterprises. In such instances, domestic firms are unable to compete and their productivity is affected by falling profit margins. We also observe that the effective rate of trade protection is negative pre-crisis and also during crisis. However, it has more negative impact during the crisis period, indicating protection during crisis period does not help firms to adjust to economic shocks. We also observe that foreign and larger firms adjust well to economic crisis due to the scale and activities in the GVC due to greater scale of operations and also able to cushion the crisis with it stronger regional and global network. The results indicate that there are weak linkages from domestic firms as the backward linkage is negative. The negative backward linkages at pre-crisis and also during

34

S. M. Thangavelu and V. Anbumozhi

economic crisis indicates that very weak domestic firms and SME activities in the Vietnamese manufacturing sector. This is consistent with previous findings where it was suggested that domestic firms in Vietnam typically have a significant technology gap with the foreign firms which limits their ability to imitate the advance technology used. Additionally, the lack of positive productivity spillovers can be contributed by the rigidity of labor mobility. As seen in Sect. 2.3, there is an evidence of wage differential between domestic and foreign firms, which increases the appeal of employment in a foreign firm. Hence, workers with the relevant expertise may be attracted to foreign enterprises instead of the domestic companies, thereby limiting the potential knowledge spillovers. With regards to backward spillovers, the results are indicative of the lack of transfer of technology and expertise from foreign firms to domestic suppliers. One possible explanation offered by Adams and Tran (2011) is the low quality and lack of variety of intermediate inputs provided by domestic producers, which prevented any production linkages to form between the two parties. This is also supported by Tran et al. (2008) where Japanese multinational such as Fujitsu emphasized that local material suppliers in Vietnam could not meet its expectations for intermediate inputs such that it had substituted for imported materials instead. This paper also proposes that SOEs play an important role in minimizing backward spillovers as they are likely to crowd out partnerships with MNEs. As highlighted in Knutsen and Nguyen (2004), Vinatex is an example of such crowding out; it is a state-owned supplier in the fibers, garments and textiles industry which foreign firms often collaborate with due to its superiority in quality and variety.

2.5 Policy Discussions This chapter examines the impact of economic shock on the productive performance of Vietnamese firms from the Global Financial Crisis from 2007 to 2008. It is observed that economic shocks such as Global Financial Crisis (2007–2008) has disruptive effects on the economic activities of firms. In particular, the ability of firms to survive economic shocks will depend on their flexibility and productive performance. The study controls for endogeneity in the estimation by adopting the GMM estimation framework. The study also uses firm level data to explore the impact of the economic slowdown of the GFC on the economic and productive activities of the Vietnamese firms. The results indicate that larger firms, foreign firms and labour quality have positive impact on TFP growth of firms and also able to manage the economic shocks better than domestic and state-owned-enterprises. Larger and foreign firms have higher TFP during the GFC, which indicates that they have economies of scale and size to manage adverse economic shocks. In addition, the larger and foreign firms tend to have greater flexibility to manage the creative destruction due to their greater flexibility and scale to be more efficient during and after economic crisis.

2 Economic Shocks and Uncertainties: How Do Firm Innovativeness …

35

We also observed that the SOEs experienced negative TFP at the pre-crisis and also during the economic crisis. The negative TFP from 2007 to2008 indicates that there is less creative destruction with the SOEs which might be due to the “soft budget” allocated to them from government financial assistance. This indicates that the inefficiencies of the SOEs could be extended from the pre-crisis to the post-crisis, thereby perpetuating and prolonging the inefficiencies in the economy. Given that SMEs employs large number of workers in the economy, improving the productivity of SMEs is very critical to increase the competitiveness of the economy. One possible avenue to deal with such inefficiencies is to create more competitive environment for SOEs to operate in the domestic and regional markets to create more creative destructions among the domestic firms. We also noted that human capital and labour quality are very important for adjustment during an economic crisis. This is very crucial for sustain and inclusive growth of the Vietnamese economy. The key for future growth of the Vietnamese economy is in developing skilled and critical human capital that will be flexible and able to undertake different tasks in the global production value-chain in the regional and global markets. In particular, the government should focus on upper secondary and technical education that will create the necessary foundation for workers to acquire new skills and also understand new innovations in manufacturing and services, thereby undertaking different tasks in the global production value-chain. We also observe the concentration ratio has negative impact on the productive activities. We also noticed that effective rate of trade protection is negative during the pre- and post-crisis of the global financial crisis. This indicates that protectionism tend to hamper the adjustments of firms to improve their efficiency during crisis. We also observed negative backward linkages at pre-crisis and also during economic crisis, thereby indicating very weak domestic firms and SME activities. There are several policy implications for the development of small and medium = sized enterprises for emerging economies such as Vietnam. In fact, the development of SMEs will be very crucial for Vietnam to attain sustainable development for its economy. The ability to create crucial linkages between local firms and foreign firms will be important for Vietnam to link to the global production network. This study will highlight the productive performance of domestic firms and the key determinants of the productivity growth. Promoting the growth of domestic small and medium-sized enterprises (SMEs) represents an important national development objective in most countries for both economic and socio-political reasons. Although this observation applies generally, the goal has particular consequence in developing countries with limited local enterprises that may lack the resource base or sufficient market size to foster further internal expansion. Domestic SME development can increase employment, generate economic growth, create local value added, and improve national innovation and entrepreneurial capabilities during economic crisis.

36

S. M. Thangavelu and V. Anbumozhi

References Adams F, Le Tran (2011) The competitiveness of Vietnam on the East Asian production ladder. Int J Bus Soc Sci 2(14):82–95 Arellano M, Bover O (1995) Another look at the instrumental-variable estimation of errorcomponents models. J Econ 68:29–52 Aitken BJ, Harrison AE (1999) Do domestic firms benefit from direct foreign investment? Evidence from Venezuela. Am Econ Rev 89(3):605–618 Blundell R, Bond S (2000) GMM estimation with persistent panel data: an application to production functions. Econ Rev 19:321–340 Girma S, Greenaway D, Wakelin K (2001) Who benefits from foreign direct investments in UK? Scottish J Polit Econ 48:119–133 Görg H, Strobl E (2001) Multinational companies and productivity spillovers: a meta-analysis. Econ J 111:723–739 Griliches Z, Mairesse J (1998) Production functions: the search for identification. In: Griliches Z (ed) Practicing econometrics: essays in methods and application. Edward Elgar, UK Havranek T, Irsova Z (2011) Estimating vertical spillovers from FDI: Why results vary and what the true effect is. J Int Econ 85(2):234–244 Knutsen H, Nguyen CM (2004) Preferential treatment in a transition economy: the case of stateowned enterprises in the textile and garment industry in Vietnam. Norwegian J Geogr 58:125–135 Le Quoc Hoi (2008) Technology spillovers from foreign direct investment in Vietnam: horizontal or vertical spillovers. Vietnam Development Forum, Hanoi, Vietnam. http://www.vdf.org.vn/ workingpapers/vdfwp085.pdf Lesher M, Miroudot S (2008) FDI spillovers and their interrelationship with trade, OECD Trade Directorate, OECD Trade Policy working paper 2008/01/01, Paris Meyer KE, Sinani E (2009) When and where does foreign direct investment generate positive spillovers? A meta-analysis. J Int Bus Stud 40(7):1075–1094 Nguyen NA, Nguyen T, Le DT, Nguyen DC, Nguyen DN (2008) FDI in Vietnam: is there any evidence of technological spillover effects? DEPOCEN working paper series 2008/18, Development and Policies Research Center, Hanoi, Vietnam Pattnayak S, Thangavelu SM (2011) Backward linkages and technology spillovers in the presence of foreign firms: Evidence from the Indian Pharmaceutical Industry, (with Sanja S. Pattnayak). J Econ Stud 38:275–238 Smarzynska, Beata K (2002) Does foreign direct investment increase the productivity of domestic firms? In search of spillovers through backward linkages, World Bank Working Paper, no.2923, The World Bank, Washington Thangavelu SM (2014) Globalization and performance of small and large firms: case of Vietnamese firms, in Chin Hee Hahn and Dionisus Narjoko edited, Globalization and Performance of Small and Large firms, ERIA Research Report 2013, No. 03, Jakarta Tran TC, Le XS, Nguyen KA (2008) Vietnam’s small and medium sized enterprises development: Characteristics, constraints and policy recommendations. In Lim H, (ed) SME in Asia and globalization, ERIA Research Project Report 2007-5, pp 323–364

Chapter 3

Supply Chain Resilience in the Global Financial Crisis: An Empirical Study on Japan Willem Thorbecke

Abstract Intricate trading networks have emerged in East Asia. They are associated with technology transfer, mushrooming productivity growth, and tumbling prices for final goods. This chapter recounts the emergence and evolution of manufacturing production networks in the region. It then documents the burgeoning surpluses between Asia and the rest of the world in goods produced within these regional value chains. Empirical evidence indicates that exchange rates throughout the supply chain affect these surpluses. Given intense pushback against globalization, as evidenced by the British vote to exit the European Union, a harmonized appreciation of Asian currencies would be desirable to reduce the risks of protectionism. To increase the resilience of supply chains, Asian countries should also improve infrastructure, fight corruption, strengthen protections for intellectual property rights, improve the supply of electricity, and invest in the health and education of its people. Keywords Global value chains · Exchange rates · Asia JEL Classification F15 · F23

3.1 Introduction Intricate trading networks have emerged in East Asia. They were developed by multinational corporations (MNCs) seeking to maintain price competitiveness. MNCs in advanced Asia tried to lower production costs by relocating factories to lower cost locations. These foreign direct investment (FDI) flows not only reduced costs but also transferred technological and managerial know-how, increased local procurement, multiplied trade in intermediate goods, and strengthened distribution networks (Gaulier et al. 2005). These value chains have multiplied efficiency gains and caused prices of consumption goods to tumble. W. Thorbecke (B) Research Institute of Economy, Trade and Industry, Center for International Development at Harvard, 1-3-1 Kasumigaseki, Chiyoda-ku, Tokyo 100-8901, Japan e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 V. Anbumozhi et al. (eds.), Supply Chain Resilience, https://doi.org/10.1007/978-981-15-2870-5_3

37

38

W. Thorbecke

Production activities within these networks can be fragmented into individual modules, and the modules can be allocated to different locations based on differences in comparative advantage. For instance, research and development and technology-intensive activities can be performed in advanced countries and laborintensive assembly can be performed in lower wage countries. This type of trade is vertical intra-industry trade. The three industries in the region where FDI flows and intermediate goods trade have been especially important are clothing, automotives, and electronics. The Appendix presents data on the flow of intermediate goods in these three sectors within East Asia and shows that the electronics sector is far and away the most important. Indeed, when measured at the International Standard Industrial Classification (ISIC) four-digit level, the most traded category within the region every year for the last 25 years has been electronic parts and components. In 2014, intraregional trade in electronic parts and components equaled US$ 300 billion and was more than twice as much as trade in the next leading category. The most exported category from East Asia to the rest of the world in every year since 1994 has been computers and office equipment. In 2014, East Asia’s exports of final electronic goods (FEG), including computers, phones, and consumer electronics exceeded US$ 760 billion. Given the importance of the electronics industry in the region, this paper investigates this sector. Sato et al. (2013) have used industry-specific exchange rates to investigate changes in the competitiveness of the electronics industry in East Asia. They found that, while prices in the electronics sector keep falling, nominal exchange rate changes also exert large effects on individual countries’ competitiveness. Electronics firms have thus not been able to offset the effects of exchange rate appreciations by cutting costs. Sato et al. also noted that Asian producers within the electronics value chain typically invoice their exports in dollars and try to keep dollar prices fairly constant in response to exchange rate changes. This pricing-to-market (PTM) strategy implies that appreciations in the short run squeeze exporters’ profit margins. This paper examines the effects of exchange rates on final electronics exports. As Chinn (2005) discussed, this approach combines into a single relationship (1) factors such as exchange rate pass-through that affect import prices relative to domestic prices and (2) the effect of relative price changes on exports. Obtsfeld and Rogoff (2000) noted that exchange rates often have a life of their own. Building on this insight, Chinn argued that exchange rates can be regarded as more exogenous than relative prices. This implies that examining how exchange rates affect exports can lead to valid inferences. As documented in the next section, upstream economies in the electronics industry such as Japan, South Korea, and Taiwan specialize in producing semiconductors, sensors, and other parts and components and shipping these to downstream countries such as China and ASEAN for assembly into smartphones and other final electronics goods and re-export. To take account of value-added coming from countries providing parts and components, this paper includes exchange rates in upstream countries as well as exchange rates in downstream countries exporting final goods. Ahmed (2009), employing a

3 Supply Chain Resilience in the Global Financial Crisis …

39

theoretical model that derived the demand and supply of exports in downstream countries as a function of inputs from upstream countries and other factors, showed that exchange rate appreciations in both upstream and downstream countries should reduce the quantity of exports from downstream countries to the rest of the world. China has become far and away the leading exporter of final electronics goods in Asia, exporting more than ten times more than the next leading exporter. Results from dynamic ordinary least squares (DOLS) estimation over the 1993–2016 period indicate that appreciations in both upstream countries and in China would significantly reduce exports. As China’s supply chains have become deeper and more of the value added is produced in China, one might expect that exchange rates in supply chain countries would matter less. However, estimating trade elasticities over the 1993–2014 period and forecasting China’s exports over the 2015–2016 period using actual values of the independent variables indicates that exchange rates in supply chain countries remain a crucial explanatory variable. The other leading downstream countries within the electronics value chain are members of the Association of Southeast Asian Nations (ASEAN). For the leading ASEAN exporter of computers and other final goods, Malaysia, panel DOLS results using data over the 1993–2014 period indicate that the Malaysian exchange rate matters and exchange rates in upstream countries do not. However, for the other leading exporters, the Philippines and Thailand, the results suggest that exchange rates in upstream countries also affect exports.1 How do these findings relate to risks facing the value chain? In the HeckscherOhlin model with two factors and complete factor mobility, the Stolper-Samuelson theorem implies that a decrease in the relative price of the import competing good (due, for instance, to trade liberalization) would benefit the abundant factor and harm the scarce factor. For Western countries such as the US and the UK, this implies that trade liberalization would increase the return to capital and decrease the return to labor. Since China’s WTO Accession in 2001, China has run enormous surpluses with the West. China’s surplus in processing trade (largely electronics goods) has remained a little less than half a trillion dollars for the last five years. Combined with a growing surplus in ordinary trade (largely labor intensive goods), China’s surplus in its two major customs regimes equaled US$ 750 billion in 2015. While this was partly offset by a deficit in tourism trade, the enormous surpluses that China runs in manufactured goods year after year is causing massive job losses abroad. This is increasing the risk of protectionism in key Western markets. Autor et al. (2013) found that job losses in the U.S. following China’s WTO Accession in 2001 occurred in sectors most exposed to competition from China. They also reported that these job losses were not offset by job gains in other sectors. Acemoglu et al. (2014) found that the loss of U.S. manufacturing jobs over this period was “stunning,” amounting to 33% of total U.S. manufacturing jobs. They reported that competition from China’s imports caused more than 2.4 million job losses between 1999 and 2011. Pierce and Schott (2016) reported that the largest drops in U.S. manufacturing employment and the largest increases in imports were 1 Results

for Singapore were inconclusive.

40

W. Thorbecke

in goods for which China secured the greatest tariff reductions when it joined the WTO. Che et al. (2016) reported that counties in the U.S. that were most exposed to competition from China were much more likely to support politicians who supported protectionism. This anger among those harmed by foreign competition was seen by Britain’s vote to exit the European Union (Brexit).2 According to a British government study, Brexit may cause British GDP to be 9.5% less each year for the next 15 years (Onyanga-Omara 2016). Colantone and Stanig (2016) found that British regions that are exposed to Chinese import competition tended to vote for Brexit. Those losing from trade liberalization thus seem willing to countenance large aggregate losses in order to change the status quo. Widespread displeasure with globalization is also seen by the rise of populist politicians in the EU and the US. In many European countries politicians espousing extreme nationalism and even neo-Nazism are gaining popularity. In the US every major candidate for president has espoused protectionism, especially against China. Thus there is a risk of extremist policies and protectionism in the West. Section 3.3 documents that regional supply chains depend heavily of final goods exports to the US and Europe. Protectionism in the West would thus be a devastating shock for regional value chains. Since East Asian countries within global value chains are running large surpluses in a value-added sense against the US and Europe, a concerted appreciation against the dollar and the euro would help to rebalance trade with the West and deflect protectionist pressures. This would also increase the purchasing power of Asian consumers and allow more final goods to flow to the region. While protectionism poses an extreme risk to regional value chains, mundane factors such as poor infrastructure, low quality education, and insufficient protection of intellectual property also jeopardize the slicing up of the value chain. Kimura and Ando (2005) have analyzed how trade-FDI-technology linkages can be promoted. They showed that firms fragment production when the cost saving arising from fragmenting production exceeds the costs of linking geographically separated production blocks. This latter cost is called the service link cost. The service link cost can be lowered by, inter alia, improving infrastructure, educating workers, fighting corruption, strengthening the rule of law and the enforcement of contracts, protecting intellectual property, and ameliorating information asymmetry problems. This paper takes stock of how countries in the region are doing at lowering the service link cost and points to areas for improvement.3 2 Britain’s

exit from the EU may or may not lead to more protectionism. It is a symptom, though, that those harmed by import competition are willing to take great risks to their own economies to change the political equilibrium. In this type of environment, protectionist policies are a greater risk. 3 An example of how service link costs are important for the electronics industry comes from the Philippines. Intel was the first semiconductor firm to open a factory in the Philippines. However, the investment climate is marred by poor infrastructure and corruption. An unfavorable investment climate, high electricity costs, and other infrastructure problems caused Intel to move production to China (see, e.g., Calimag 2008).

3 Supply Chain Resilience in the Global Financial Crisis …

41

Protectionism in the West is largely a risk to the demand side of Asian value chains, since Western economies are far and away the largest consumers of final goods produced within East Asian supply chains. High service link costs in Asian countries represent a risk to the supply side. Ando and Kimura (2012) investigated the impact of a demand shock (the Global Financial Crisis) and a supply shock (the Great East Japan Earthquake) on Japanese exports and production networks in the machinery industries. They found that the demand shock caused massive dislocation and permanent changes. After the supply shock, on the other hand, coordination activities recovered quickly. This suggests that MNCs in Asia are better able to cope with supply side shocks than with demand side shocks. If this is true, then protectionism in the West could cause more disruption to supply chains than high service link costs. The next section examines the emergence and development of electronics production networks in Asia. Section 3.3 considers China, which has become more and more central in global value chains. Section 3.4 presents empirical evidence on the effects of exchange rates on China’s exports. Section 3.5 presents empirical evidence on the effects of exchange rates on ASEAN’s exports. Section 3.6 discusses risks to the supply chain. Section 3.7 concludes.

3.2 The Development of Electronics Production Networks in East Asia Japanese electronics companies such as Sony, Matsushita (Panasonic), and Toshiba produced a string of high quality products during the post-War period that rapidly penetrated global markets. These products included transistor radios, television sets, the Sony Walkman, video cassette recorders, and computers. Japanese semiconductor production then exploded in the 1980s. In 1980 no Japanese manufacturers were in the top ten but within six years Japan had become the world’s leading semiconductor supplier.4 The rapid import penetration by Japan sparked consternation and protectionist pressure in the U.S. To deflect the protectionist pressure, the U.S., Japan, France, Germany, and the U.K. on 22 September 1985 agreed to push down the value of the dollar to reduce large U.S. trade deficits. The Japanese yen appreciated by 60% and Japanese exporters lost price competitiveness. Japanese electronics firms then cut costs by moving labor-intensive operations to lower wage nations. They manufactured technology-intensive parts and components in Japan, and exported these abroad for assembly into final goods and re-export. This pattern is evident in Fig. 3.1. The figure shows that, as the yen began appreciating, Japanese outward foreign direct investment increased by 50%. Japanese multinational corporations (MNCs) thus began aggressively transferring factories abroad. 4 These

data come from the Japanese Society of Semiconductor Industry Specialists.

42

W. Thorbecke

5.4

13 Foreign direct investment

Yen/dollar exchange rate

12

5.2

11

5.0

10

4.8

9

4.6

8

4.4

7

4.2 1980

Log of foreign direct investment

Log of yen/dollar exchange rate

5.6

6 1985

1990

1995

2000

2005

2010

Fig. 3.1 Yen/Dollar exchange rate and Japanese outward foreign direct investment. Source CEPIICHELEM database

Figure 3.2 shows that, as the yen appreciated, exports of electronic parts and components to East Asian neighbors soared. What countries did Japanese MNCs ship these parts and components to? Figures 3.3 and 3.4 indicate that electronic parts and components and FDI flowed largely to the newly industrialized economies (NIEs) of South Korea and Taiwan. These economies had many locational advantages including good infrastructure, high quality human capital, and stable governments. However, between the Plaza Accord in September 1985 and the middle of 1989 the Korean won appreciated by 30% and the New Taiwan dollar by 45%. In addition, as Yoshitomi (2003) noted, labor

Log of yen/dollar exchange rate

5.4

Electronic p&c exports

11

5.2

10

5.0

9

4.8

8 7

4.6 Yen/dollar exchange rate 4.4 4.2 1980

6 5 1985

1990

1995

2000

2005

2010

Log of electrronic parts and components exports

12

5.6

Fig. 3.2 Yen/Dollar exchange rate and Japanese electronic parts and components exports to East Asia. Note East Asia includes China, Malaysia, Indonesia, the Philippines, Singapore, South Korea, Taiwan, and Thailand. Source CEPII-CHELEM database

Log of electronic parts and components exports

3 Supply Chain Resilience in the Global Financial Crisis … 10

43

ASEAN

9 NIEs 8 7 6

China

5 4 3 1980

1985

1990

1995

2000

2005

2010

Fig. 3.3 Japanese electronic parts and components exports to parts of East Asia. Note NIEs refers to South Korea and Taiwan. ASEAN refers to Malaysia, Indonesia, the Philippines, Singapore, and Thailand. Source CEPII-CHELEM database 8 ASEAN

Billions of U.S. dollars

7 6 5

China 4 NIEs

3 2 1 0 82

84

86

88

90

92

94

96

98

00

02

04

Fig. 3.4 Japanese outward FDI to parts of East Asia. Note FDI refers to foreign direct investment. NIEs refers to South Korea and Taiwan. ASEAN refers to Malaysia, Indonesia, the Philippines, Singapore, and Thailand. Source Japanese Ministry of Finance

migration from the agricultural sector to the manufacturing sector stopped at the end of the 1980s and this caused wages to soar. It thus became uneconomical for Japanese MNCs to perform labor intensive assembly operations in the NIEs. Figures 3.3 and 3.4 indicate that they redirected production to the Association of Southeast Asian Nations (ASEAN).

W. Thorbecke Log of electonic parts and components exports

44 9 8

Singapore

7 6

Indonesia Malaysia

5 4 3

Phil Thailand

2 1 1980

1985

1990

1995

2000

2005

2010

Fig. 3.5 Japanese electronic parts and components exports to individual ASEAN countries. Source CEPII-CHELEM database

Figure 3.5 shows that electronic parts and components went to Malaysia, the Philippines, Singapore and Thailand but not to Indonesia. The figure also shows that in recent years there has been a fall in the flow of goods to the Philippines. One problem with Indonesia and the Philippines is that their roads, ports, electricity supply, and other infrastructure are often poor. ASEAN countries experienced currency and banking crises in 1997–98. Schnabl and Spantig (2016) noted that stability and growth in Asia favor trade, and the instability during the crisis curbed the appetite of Japanese firms to invest in ASEAN. Figures 3.3, 3.4 and 3.5 indicate that there were declines in FDI and intermediate goods flowing from Japan to ASEAN after the crisis. Japanese companies left existing factories in ASEAN but began looking for other locations for new production. As Figs. 3.3 and 3.4 indicate, China in the 2000s soon became the leading destination for Japanese Asian supply chains. As Chen (2008) argued, China’s accession to the World Trade Organization in 2001 gave foreign investors confidence that China would sustain an FDI-friendly environment through fair and coherent enforcement of the relevant laws and regulations. In addition, foreign investors were attracted by the high quality roads, ports, and airports and other infrastructure in the Pearl River Delta and the Yangtze River Delta. While South Korea and Taiwan were used for labor-intensive assembly in the 1980s, their industrial structures have advanced significantly since then. Many electronics firms in the NIEs have developed strong brands and independent technologies. We can understand the progress that South Korea and Taiwan have made by examining a specific sector, the liquid crystal display (LCD) industry. LCDs are flat panel displays that are used for watches, calculators, cell phones, digital cameras, laptop and desktop monitors, personal digital assistants, and televisions. As Han

Log of electronic parts and components exports

3 Supply Chain Resilience in the Global Financial Crisis …

45

12 China ASEAN

10

Japan

8

6

4 NIEs 2 1980

1985

1990

1995

2000

2005

2010

Fig. 3.6 The NIEs electronic parts and components exports to parts of East Asia. Note NIEs refers to South Korea and Taiwan. ASEAN refers to Malaysia, Indonesia, the Philippines, Singapore, and Thailand. Source CEPII-CHELEM database

et al. (2012) documented, Japanese firms were unrivaled in this industry in the 1980s. Then, according to Han et al., as the Korean won depreciated following the 1997–98 Asia Crisis, Korean firms gained market share. Korean firms also developed their own technologies.5 Japanese firms, partly to weaken the competitiveness of Korean firms, transferred technology to Taiwanese firms. By 2002, both Korea and Taiwan had surpassed Japan in the production of large-sized LCD panels. Japan, however, retained the advantage in the most technologically advanced parts of the industry such as the production of lithography equipment and advanced materials and components. As Korea and Taiwan have advanced technologically in the LCD and other electronics industries, they have outsourced labor-intensive operations to lower wage regions of Asia. Figure 3.6 shows that before China’s WTO accession ASEAN was an important location but after 2000 more and more production was relocated to China. However, as wages have increased in China, MNCs in the NIEs have relocated production to Vietnam. Korea’s intermediate goods exports to Vietnam increased almost four times after 2008 to reach $8 billion. In addition, Korea and Taiwan are now competing more and more with Japan. Over the last ten years, Korean electronics firms such as Samsung and LG have earned high profits and gained market share while Japanese firms such as Sharp and Panasonic have languished. Sato et al. (2013), investigating this phenomenon, found that Korean firms gained markedly in terms of price competitiveness relative to their Japanese peers. They measured competitiveness using real effective exchange rates calculated specifically for the electronics sector. As the Korean won depreciated relative to the 5 Technology

development in Korea and other East Asian economies is related to declines in transportation costs and increases in liberalization throughout the region. I am indebted to the reviewers for this observation.

W. Thorbecke Log of electronic parts and components exports

46 12

China ASEAN

10

NIEs 8

Japan

6 4 2 0 1980

1985

1990

1995

2000

2005

2010

Fig. 3.7 ASEAN’s electronic parts and components exports to parts of East Asia. Note NIEs refers to South Korea and Taiwan. ASEAN refers to Malaysia, Indonesia, the Philippines, Singapore, and Thailand. Source CEPII-CHELEM database

Japanese yen during the 2008 Global Financial Crisis, the real effective exchange rate for the electronics industry in Korea depreciated both absolutely and relative to the Japanese yen. Sato et al. also reported results from structural vector autoregression estimation indicating that the won depreciation and the yen appreciation caused a significant increase of Korean electronics exports relative to Japanese electronics exports. ASEAN countries have also been increasingly active in sending electronic parts and components to their Asian neighbors. Figure 3.7 shows that the leading recipient of these goods is China, but many parts and components flow to other ASEAN countries. An example helps to clarify how production links in ASEAN work. Hiratsuka (2011) discussed in detail the operations of Hitachi Global Storage Technologies (HGST), until recently a leading producer of hard disk drives (HDD) in Thailand. He documented how HGST procured most parts and components from Indonesia, Malaysia, the Philippines, and Singapore. The close location meant that supplier firms could send parts and components by overnight express and could also send engineers to improve communication with HGST engineers. These engineers would also come quickly when there was a problem with defective parts. HGST procured media from Malaysia and Singapore; printed circuit boards from Indonesia, the Philippines, and Thailand; pivots from Malaysia, Singapore, and Thailand; voice coils from Indonesia, Malaysia, and Thailand; and bases from Malaysia and Thailand. These parts and components were also procured from other countries. Hiratsuka (2011) noted that employing multiple suppliers increased competition between the suppliers and reduced the risk of parts and components being unavailable due to natural disasters, political problems, and other factors. While the parts and components listed above were obtained through arms-length transactions, core

3 Supply Chain Resilience in the Global Financial Crisis …

47

components such as heads and suspension are procured through intra-firm trade with HGST’s head office in the U.S. and its affiliate in Mexico. The evidence of Hiratsuka (2011) makes clear electronic parts and components come into a country and are used to produce electronic parts and components and final goods that often flow out of the country. It is possible to learn about a country’s place in the electronics value chain by examining variables such as the ratio of electronic parts and components exports to electronic parts and components imports or the ratio of FEG exports to electronic parts and components exports. Countries whose value of electronic parts and components exports far exceed its value of electronic parts and components imports tend to add a lot of value to these intermediate goods. Countries whose FEG exports far exceed their electronic parts and components exports tend to be further downstream in the value chain and to be used as export platforms. Table 3.1 presents these and other data for East Asian countries. Column (1) indicates that South Korea and Taiwan are the leading exporters of electronic parts and components to the region. Column (2), which presents data on electronic parts and components exports divided by electronic parts and components imports, indicates that the ratio is especially high for Taiwan and South Korea. This suggests that these economies add a lot of value to their electronic parts and components exports. Column (3) indicates that China is the unrivaled exporter of FEG exports, exporting more than 10 times what the next leading exporter does. Although not shown in the Table 3.1, Table 3.1 East Asian exports of electronic parts and components, final electronics goods (FEG), and related statistics, 2014 Country

Electronic parts and components exports to East Asia (billions of USD)

Electronic parts and components exports divided by electronic parts and components imports

FEG exports to the world (billions of USD)

FEG exports divided by electronic parts and components exports

China

33.5

0.25

538.6

8.95

Japan

27.2

1.31

31

0.9

Malaysia

37.9

2.27

42.1

0.73

Philippines

12.7

1.43

13.1

0.71

Singapore

28.2

1.12

21.3

0.42

South Korea

50.7

1.92

50.1

0.86

Taiwan

67.8

2.69

22.7

0.3

7.3

1.01

36.7

3.66

Thailand

Note Final electronics goods include computers, telephone equipment, and consumer electronics goods. Electronic Parts and Components Exports Divided by Electronic Parts and Components Imports represents electronic parts and components exports to East Asia divided by electronic parts and components imports from East Asia. FEG Exports Divided by Electronic Parts and Components Exports represents final electronics goods exports to the world divided by electronic parts and components exports to the world Source CEPII-CHELEM Database

48

W. Thorbecke 140 China

Billions of U.S. dollars

120 100 80 60

ASEAN

40

NIEs

20 Japan 0 1980

1985

1990

1995

2000

2005

2010

Fig. 3.8 East Asia’s exports of electronic parts and components to locations in East Asia. Note NIEs refers to South Korea and Taiwan. ASEAN refers to Malaysia, Indonesia, the Philippines, Singapore, and Thailand. East Asia includes the NIEs, ASEAN, Japan, and China. Electronic components comes from the HS classification numbers 8540, 8541, and 8542. Source CEPII-CHELEM database

China exports more FEG exports to the world than the next 16 leading exporters together. Column (4) presents data for the ratio of FEG exports to electronic parts and components exports. This value equals 8.95 for China and 3.66 for Thailand, suggesting that these countries are downstream. The ratio is only 0.42 for Singapore and 0.3 for Taiwan, indicating that these countries are upstream in the value chain. Figure 3.8 shows total electronics goods exports from all of East Asia to the individual countries and regions. It makes clear that Japan since the 1980s is the most upstream location. ASEAN was the most downstream until 2003. Since then China has become more and more of the final link in regional value chains. Figure 3.9 shows the exports to the world of the final electronics goods produced using electronic parts and components. These final goods include computers, telephones, and consumer electronics goods. The figure shows that explosion of final electronics goods from China. Thus China is not only the final link but has also become the central country within regional value chains. Because China has become so crucial within East Asian production networks, the next section investigates China’s imports and exports in detail.

3.3 China’s Ordinary and Processing Trade One way to understand China’s role in regional value chains is to consider China’s two main customs regimes, processing trade and ordinary trade. Processed exports are goods produced using parts and components that are imported duty free for the

3 Supply Chain Resilience in the Global Financial Crisis …

49

600 China

Billions of U.S. dollars

500 400 300 200

ASEAN 100

NIEs Japan

0 1980

1985

1990

1995

2000

2005

2010

Fig. 3.9 East Asia’s exports of final electronic goods to the world. Note NIEs refers to South Korea and Taiwan. ASEAN refers to Malaysia, Indonesia, the Philippines, Singapore, and Thailand. East Asia includes the NIEs, ASEAN, Japan, and China. Final electronic goods include computer, telephones, and consumer electronic goods. Final electronic goods correspond to the HS classification numbers 8469-73, 8517-22, 8525-39, 8543-48. Source CEPII-CHELEM database

purpose of assembling goods for re-export. Ordinary exports are goods produced using local inputs and also using imported inputs that are not imported duty free. The three key industries within regional value chains are the clothing, electronics, and automotive sectors. China’s intermediate imports are disproportionately in the electronics sector. This is clear in Figure 3.14 in the Appendix that plots intermediate imports in the three key industries. Given the overarching role that electronics plays in China’s value chains, processing trade reflects largely trade in electronics goods. Figure 3.10 plots China’s processing imports and exports. Figure 3.10 indicates that the gap between processed exports and imports for processing keeps growing. This implies that more and more of the value-added of processed exports such as smartphones, tablet computers, and consumer electronics goods comes from China. The gap is even wider than the figure indicates, since each year more than USD 70 billion of “imports” for processing are produced in China and round-tripped out of China and back into take account of favorable tax provisions (see Xing 2012). This increase in value added reflects several factors. First, the Chinese government has steered Chinese firms towards higher value-added activities (Republic of China 2012). Second, China’s capital deepening has allowed more parts and components to be produced domestically (Knight and Wang 2011). Third, China’s industrial clusters and processing supply chains have become deeper (Kuijs 2011). Fourth, a growing knowledge base in China has facilitated the transfer of technology from FDI firms and the progress towards higher value added tasks within the value chain (see Kiyota et al. 2006; Yusuf et al. 2003).

50

W. Thorbecke 900 Processed Exports

Billions of U.S. dollars

800 700 600 500 400 300

Imports for Processing

200 100 0 94

96

98

00

02

04

06

08

10

12

14

Fig. 3.10 China’s processing trade, 1993–2015. Source China customs statistics

To shed further light on China’s processing trade, Table 3.2a breaks down processed exports and imports for processing by country and region. The top panel of Table 3.2a indicates that US$ 310 billion in parts and components for processing come from East Asia (including China’s “exports” to China). Korea has become the leading source of imports for processing, shipping almost US$ 86.5 billion to China in 2015. This surge reflects the growth of the Korean electronics industry after 2009. Taiwan is the second leading provider of imports for processing, exporting US$ 70.2 billion to China in 2015. Japan is the third leading supplier, sending China US$ 48 billion of these goods. Imports for processing from Europe and the U.S. were both under US$ 25 billion in 2015. The middle panel of Table 3.2a indicates that the largest recipients include the U.S. (US$ 186.2 billion), and Europe (US$ 107.8 billion). South Korea and Taiwan together only received US$ 63.2 billion of processed exports. Because of these trading patterns, South Korea and Taiwan together ran surpluses with China of US$ 108.8 billion in processing trade. The U.S and Europe then ran deficits that are recorded at US$ 170 billion and US$ 95.8 billion, respectively. However, their actual deficits in processing trade are much larger because the Chinese Customs data records more than US$ 200 billion as being exported to Hong Kong whereas in reality the lion’s share of these exports are transshipped through Hong Kong to advanced economies. One reason why so many of China’s processed exports have gone to the U.S, and the EU is because their markets are open and the levels of protectionism in these countries are low. Figure 3.11 plots China’s ordinary imports and exports. Figure 3.11 indicates that ordinary exports have grown very rapidly, and only started to decelerate in 2015. The figure also shows that ordinary imports have fallen in 2015. The fall in imports reflects the drop in the prices of primary products such as crude oil, as well as a decrease in import demand arising from China’s slowdown and from President Xi

Taiwan (2)

Japan (3)

ASEAN-4 (4)

79.6

83.9

87.1

98.7

86.5

2012

2013

2014

2015

70.2

76.2

72.5

68.5

71.9

70

61.6

48.0

55.6

56.4

62.8

64.9

42.1

34.3

35.2

36.2

38.2

44.7

34.9

39.6

46.1

47.2

47.8

47.0

2010

2011

2012

2013

2014

2015

16.2

18.3

16.6

16.2

17.7

15

65.1

73.5

75.9

78.6

75.2

65.5

27.5

31.8

31.7

32

28.6

25.2

Processed exports (billions of U.S. dollars)

71.1

2011

Imports for processing (billions of U.S. dollars)

2010

(a)

S. Korea (1)

186.2

194.8

186.7

184.6

175.6

162.6

23.4

24.8

21.6

19.8

21.9

21.7

US (5)

25.6

27.6

26.0

29.7

34.2

33.2

6.8

8.2

7.4

6.4

7.3

6.7

Germany (6)

Table 3.2 a China’s processing trade, 2010–2015. b China’s ordinary trade, 2010–2015

82.2

91.6

88.2

93.8

101.4

94.1

14.2

15.2

14.9

13.5

15.6

13.2

Rest of Europe (7)

19.6

22.0

22.8

22.8

21.7

18.1

2.4

5.3

4.8

4.6

3.9

3.1

Australia and Brazil (8)

329.5

376.9

365.4

359

341.5

291.7

161.7

205.3

196.1

183.5

160

127.9

Rest of the World (9)

(continued)

798.9

884.3

860.5

862.8

835.5

740.3

447.5

524.5

497

481.2

469.8

417.4

Total (10)

3 Supply Chain Resilience in the Global Financial Crisis … 51

Taiwan (2)

ASEAN-4 (4)

−54.2

52.3

−55.9

−57.9

−54

−40.0

−37.8

−39.9

−50.9

−39.5

2011

2012

2013

2014

2015

17.1

17.9

19.5

15.8

10.3

3.9

−6.8

−3.4

−4.5

−6.2

−16.1

−16.9

45.3

55.9

56.8

62.6

63.6

59.2

2010

2011

2012

2013

2014

2015

47.6

48.8

42.2

36.1

34.2

29.2

72.2

83.8

82.9

88.7

100.5

89.4

61.6

68

71.8

68

67

49

Ordinary imports (billions of U.S. dollars)

−55.0

−36.2

2010

(b)

Japan (3)

Balance in processing trade (billions of U.S. dollars)

S. Korea (1)

Table 3.2 (continued)

89.9

96.4

94.9

90

81.4

63.8

162.8

170.0

165.1

164.8

153.7

140.9

US (5)

68.4

81.9

73.5

72.3

73.5

57.9

18.8

19.4

18.6

23.3

26.9

26.5

Germany (6)

71.4

81.3

75.5

70.4

70.8

56.7

68

76.4

73.3

80.3

85.8

80.9

Rest of Europe (7)

102.3

134.6

139.5

122.6

122.4

91.0

17.2

16.7

18.0

18.2

17.8

15.0

Australia and Brazil (8)

351.4

451.5

467.0

416.9

401.9

286.0

167.8

171.6

169.3

175.5

181.5

163.8

Rest of the World (9)

(continued)

924.0

1,111.1

1,109.9

1,021.8

1,007.6

768.3

351.4

359.8

363.5

381.6

365.7

322.9

Total (10)

52 W. Thorbecke

Taiwan (2)

Japan (3)

38.6

36

38.4

46.1

46.9

2011

2012

2013

2014

2015

19.9

20.1

17.2

15.4

15.2

12.4

63.5

69

67

66.4

66.5

51.1

100.4

97.4

92.7

78.8

61.3

46.5

ASEAN-4 (4)

−19

−20.7

−25

−28.7

−27.7

−17.3

−20.8

−24.2

−17.5

−12.3

2011

2012

2013

2014

2015

−8.7

−14.8

−15.9

−22.3

−34

−38.3

38.8

29.4

20.9

10.8

−5.7

−2.5

111.8

82.6

67.1

61.3

54.2

43.7

201.7

179

162

151.3

135.6

107.5

US (5)

−28.9

−41.2

−36.1

−36.1

−34.1

−26.1

39.5

40.7

37.4

36.2

39.4

31.8

Germany (6)

73.6

63.7

52.7

50.7

56.6

49

145

145

128.2

121.1

127.4

105.7

Rest of Europe (7)

−69.6

−88

−93.8

−79.3

−82.8

−61.1

32.7

46.6

45.7

43.3

39.6

29.9

Australia and Brazil (8)

219.6

108.4

31.9

22.6

−8.5

19.7

571.0

559.9

498.9

439.5

393.4

305.7

Rest of the World (9)

296.6

93.9

−22.4

−33.8

−90.6

−47.6

1,220.6

1,203.8

1,087.5

988.0

917.0

720.7

Total (10)

Notes ASEAN 4 includes Indonesia, Malaysia, the Philippines, and Thailand. Rest of Europe includes Austria, Belgium, Denmark, Finland, France, Greece, Ireland, Luxembourg, Netherlands, Italy, Portugal, Spain, Sweden and United Kingdom Source China Customs Statistics

−16.8

−15.2

2010

Balance in ordinary trade (billions of U.S. dollars)

30.1

2010

Ordinary exports (billions of U.S. dollars)

S. Korea (1)

Table 3.2 (continued)

3 Supply Chain Resilience in the Global Financial Crisis … 53

54

W. Thorbecke 1,400 Ordinary Exports

Billions of U.S. dollars

1,200 1,000

Ordinary Imports

800 600 400 200 0 94

96

98

00

02

04

06

08

10

12

14

Fig. 3.11 China’s ordinary trade, 1993–2015. Source China customs statistics

Jinping’s crackdown on government officials receiving luxury imported goods (Qian and Wen 2015). To shed further light on China’s ordinary trade, Table 3.2b breaks down ordinary exports and imports by country and region. The top panel of Table 3.2b shows that ordinary imports from South Korea and Taiwan increased up until 2014 and slowed down in 2015. The middle panel shows that ordinary exports to ASEAN have also increased rapidly. Thus China runs trade deficits with advanced East Asia and small surpluses with ASEAN in ordinary trade. The surplus with ASEAN reflects the fact, noted by Gaulier et al. (2011), that Chinese ordinary exports are lower priced goods that are especially likely to penetrate markets in the South. Gaulier et al. and others have also noted that automobiles and sophisticated capital goods are key exports from Germany. After increasing rapidly until 2014, these have slowed in 2015. The bottom panel shows that China’s deficit with Australia and Brazil has fallen. This reflects the fall in the prices of iron ore, coal, and other primary inputs that these countries export to China. The data above indicate that China is running huge surpluses in the manufactured products that make up the lion’s share of processing and ordinary trade. The huge surplus highlights China’s rising value-added in manufacturing. The increase in China’s value added in processing trade indicates that more of the technology-intensive parts and components are being produced in China. There is now ferocious competition between China and traditional suppliers such as South Korea and Taiwan in producing cutting edge chips and other inputs into final goods. South Korea and Taiwan still have a technological advantage, but China is closing the gap quickly. Competition within the supply chain and mushrooming productivity growth have reduced prices for consumers. Figure 3.12 shows that the prices of computer, cell phones, and other final electronic products. For comparison, it also shows the import

3 Supply Chain Resilience in the Global Financial Crisis …

55

6.0 All goods

Log of price index

5.6

5.2

4.8 Final electronics goods 4.4

4.0 92

94

96

98

00

02

04

06

08

10

12

14

Fig. 3.12 Price index for East Asia’s exports of final electronic goods and for all goods imported by the U.S. Note Final electronics goods include computers, telephone equipment, and consumer electronics goods. Source CEPII-CHELEM database, Bureau of Labor Statistics, and author’s calculations

price index for all goods imported by the U.S. The figure makes clear that FEG prices have tumbled not only absolutely but also relative to the prices of other goods that the U.S. imports.

3.4 Exchange Rates and China’s Exports What is driving China’s burgeoning exports? Since processed exports represent final goods produced using imported parts and components and since the electronics industry is at the heart of processed exports, investigating China’s processed exports will shed light on the factors affecting the electronics value chain in East Asia. Using dynamic ordinary least squares estimation and quarterly data over the 1994– 2010 sample period, Cheung et al. (2012) reported that a 10% RMB appreciation would reduce processed exports by between 9 and 12% and ordinary exports by between 13 and 19%. They also reported that income elasticities exceed unity for processed exports but are ambiguous for ordinary exports. Using autoregressive distributed lag (ARDL) estimation and quarterly data over the 1996–2009 sample period, Ahmed (2009) reported that a 10% RMB appreciation relative to non-Asian currencies would reduce processed exports by 17 and a 10% RMB appreciation relative to Asian supply chain countries would increase China’s processed exports by 15%. Ahmed’s findings imply that a 10% appreciation of both the RMB and supply chain country’s currencies against the rest of the world would reduce processed exports by 32%. Ahmed also found that a 10% RMB appreciation against all currencies would reduce ordinary exports by 19%. His results for

56

W. Thorbecke

ordinary exports were not as clear cut when he divided the exchange rate into the portion relative to non-Asian currencies and the portion relative to Asian supply chain countries. Using panel DOLS techniques and annual data over the 1992–2013 period, Thorbecke (2015) found that a 10% appreciation across the entire supply chain (including China and the supply chain countries) would reduce processed exports by between 13 and 20%. He also reported that GDP elasticities vary between 1.8 and 2.8 depending on whether time trends are included. To investigate the factors affecting China’s processed exports, this section performs estimation with the sample period extending to 2016Q1. The dependent variables are China’s processed and ordinary exports. These data come from the China Customs Statistics. Following Cheung et al. (2012), exports are deflated using the Hong Kong to U.S. re-export unit value indices. These are obtained from the Hong Kong Census and Statistics Department. Following previous authors, the independent variables include the real effective exchange rate and real GDP in importing countries. The real effective exchange rate is CPI-deflated and obtained from the IMF International Financial Statistics. Real GDP in measured as real GDP in OECD countries and is obtained from the OECD. Since parts and components in supply chains countries contribute to the valueadded of China’s exports, exchange rates in supply chain countries are also included in the regression. The main countries providing imports for processing to China in most years are Germany, Japan, Malaysia, the Philippines, Singapore, South Korea, Taiwan, Thailand, and the U.S. For these nine countries, weights wi, t are calculated by dividing the value of imports for processing coming from supply chain economy i to China by the value of imports for processing coming from all nine economies at time t. These weights can then be used to calculate weighted exchange rates in supply chain countries using the formula: ssreert =

9 

(reeri,t /reeri,t−1 )wi,t ,

(3.1)

i=1

where ssreert is the weighted exchange rate in the supply chain countries at time t and reert is the real effective exchange rate in country i at time t. Data on real effective exchange rates come from the CEIC database and data on imports for processing that are used to calculate the weights come from China Customs Statistics. Augmented Dickey-Fuller tests are used to test the maintained hypothesis that each of the variables have unit roots in their autoregressive lag operator polynomials. The tests indicate that the variables are integrated of order one (I(1)). Johansen cointegration tests are then used to test the null hypothesis of no cointegrating relations among the equations against the alternative hypothesis of more than one relation. The number of lags for the Johansen tests are determined by applying the Schwarz Information Criterion to the unrestricted vector autoregression. For processed exports, both cointegration rank tests (Maximum Eigenvalue and Trace) indicate that there is one cointegrating equation. For processed exports, the Maximum Eigenvalue Test

3 Supply Chain Resilience in the Global Financial Crisis …

57

indicates that there is one cointegrating equation and the Trace Test the p-value for the null hypothesis of no cointegrating equations equals 0.054. Dynamic ordinary least squares (DOLS), a technique for estimating cointegrating, is thus used to estimate the equations. Stock and Watson (1993) showed that DOLS yields consistent and efficient estimates. Montalvo (1995) showed that the DOLS estimator has smaller bias and root mean squared error than other cointegrating regression estimators in cases where the sample is not large enough to justify applying asymptotic theory. DOLS estimates of the long run trade elasticities can be obtained from the following regression: ext =α1 + α2 r mbt + α3 ssreert ∗ + α4 yt ∗ +

K 

β1,k reert+k

k=−K

+

K  k=−K

β2,k ssreert+k +

K 

β3,k yt+k ∗ + εt ,

(3.2)

k=−K

where rmb represents the renminbi real effective exchange rate, ssreer represents the weighted real exchange rate in supply chain countries, y represents real GDP in OECD countries, and K represents the number of leads and lags of the first differenced variables. K is set equal to one. The data extend from 1993Q1 to 2016Q2. Quarterly dummies are also included in the estimation. The results are presented in Table 3.3. They indicate that, for processed exports, a 10% appreciation of the renminbi would reduce exports by 22% and a 10% appreciation in East Asian supply chain countries would reduce exports by 26%.6 A 10% increase in real GDP in the rest of the world would increase processed exports by 19%. The results also indicate that, for ordinary exports, only the renminbi exchange rate matters. The finding that exchange rates in supply chains do not matter for ordinary exports may be because more of the value-added of ordinary exports comes from China rather than from imported parts and components. Cheung et al. (2012) suggested that there may be a structural break associated with China’s WTO Accession in 2001.7 They did not find strong evidence for this though. Breakpoint tests were conducted for each quarter of 2001, but there was no evidence of structural breaks for either processed exports or ordinary exports. As China’s supply chains have become deeper recently, do exchange rates in supply chain countries still matter for China’s processed exports. To investigate this, Eq. (3.2) is estimated up until 2014Q4 and then out of sample forecasts are obtained using actual values of the independent variables over the next five quarters. The model is estimated both including and excluding exchange rates in supply chain countries. For processed exports, the root mean squared error (RMSE) measure of 6 Although

not shown, the residuals are generally well behaved. and Koivu (2007) also argued that there may be a structural break associated with China’s WTO Accession.

7 García-Herrero

58 Table 3.3 Dynamic OLS estimates for China’s processed and ordinary exports

W. Thorbecke (1) RMB REER

(2)

Processed exports

Ordinary exports

−2.17***

−2.42***

(0.22)

(0.30)

REER in supply chain countries

−2.56***

0.19

(0.56)

(0.79)

GDP

1.88**

−0.72

(0.78)

(1.08)

Time

0.03***

0.05***

(0.00)

(0.01)

No. of leads and lags

1, 1

1, 1

Adjusted R-squared

0.99

0.99

No. of observations

89

89

Sample period

1993:1–2016:2

1993:1–2016:2

Notes DOLS(1, 1) estimates, Heteroskedasticity-consistent standard errors in parentheses, For column (1) the independent variable is China’s processed exports to the world and for column (2) the independent variable is China’s ordinary exports to the world, RMB REER represents the renminbi real effective exchange rate, REER in Supply Chain countries represents a weighted average of exchange rates in supply chain countries, GDP represents real GDP in OECD countries, and Time is a time Trend. Seasonal dummies are included *** (**) denotes significance at the 1% (5%) level Source Author’s calculations

the model’s forecast error for the 2015Q1–2016Q1 period equals 0.19 when ssreer is included and 0.29 when ssreer is excluded. This indicates that, even in 2015 and 2016 as China’s supply chains have become deeper, the specification with exchange rates in supply chain countries included forecasts far better than the specification with only the renminbi included as an exchange rate measure. For ordinary exports, on the other hand, the RMSE is identical whether ssreer is included or not.

3.5 Exchange Rates and ASEAN’s Exports While there is a plethora of studies investigating elasticities for China’s exports within global value chains, there are not many investigating these for ASEAN. One reason, as Fig. 3.9 makes clear, is that ASEAN exports are much smaller than China’s exports. Another reason is that, unlike in China where there is a clear link between imported inputs (imports for processing) and subsequent exports (processed exports),

3 Supply Chain Resilience in the Global Financial Crisis …

59

finding data for ASEAN that links imported inputs with subsequent exports is more difficult. As discussed above, such a link should exist between imports of electronic parts and components and exports of final electronics goods. This section thus investigates exports of final electronics goods from ASEAN and seeks to explain this using not only ASEAN exchange rates but also exchange rates in countries providing electronic parts and components. Final electronics goods include goods in the following categories: computer equipment, consumer electronics goods, and telecommunications equipment. Bilateral exports of final electronics goods between ASEAN countries and major importing countries are employed. These data are measured in U.S. dollars and are obtained from the CEPII-CHELEM database. As before, the sample period begins in 1993. It extends to 2014, the last year for which data are available. Since ASEAN’s exports of FEG represent imports by countries such as the U.S., the data are deflated using U.S. import price deflators for computers and for telecommunications equipment. The deflator data come from the U.S. Bureau of Labor Statistics. Countries that imported small quantities of FEG in some years are excluded from the sample because these can have very large percentage changes from year to year due to idiosyncratic factors such as a new importer opening up business rather than due to macroeconomic variables such as exchange rate changes. These large changes can cloud inference. The major ASEAN exporters of final goods are Malaysia, the Philippines, Singapore, and Thailand. Malaysia has 12 countries that consistently imported large quantities, the Philippines has three, Singapore has twelve, and Thailand has eight.8 For Malaysia and Singapore, 12 countries provide enough cross sectional variation to identify in an econometric sense how exchange rate changes will impact exports. For the other countries, the cross sectional sample size is too small. It will be possible, though, to estimates trade elasticities for exports from Malaysia, the Philippines, and Thailand together.9 Bilateral real exchange rates between Malaysia and the countries importing the final goods are obtained from the CEPII-CHELEM database.10 To take account of exchange rates in supply chain countries, data on the value of electronic parts and components coming from the leading supply chain countries are used. As in Sect. 3.4, these countries are China, Germany, Japan, Malaysia, the Philippines, Singapore, South Korea, Taiwan, Thailand, and the U.S. Weights (wi,t ) are calculated for each of 8 For Malaysia, these countries are Australia, Canada, France, Germany, Japan, Mexico, the Nether-

lands, Singapore, Taiwan, Thailand, the U.S., and the U.K. For the Philippines, these countries are Germany, Japan, and the U.S. For Singapore, they are Australia, China, Germany, Hong Kong, Japan, Malaysia, the Netherlands, South Korea, Taiwan, Thailand, the U.K., and the U.S. For Thailand, they are Canada, France, Germany, Hong Kong, Japan, Malaysia, Singapore, and the U.S. 9 Singapore is not included with the other three because it is at a higher level of development and its role as an entrepôt makes it different from the other three. 10 In this paragraph the discussion focuses on Malaysia as the exporter of final goods, but the same procedures are used to calculate data for the Philippines, Singapore, and Thailand as final exporters.

60

W. Thorbecke

these countries by determining the value of imports of electronic parts and components each country provided to Malaysia divided by the value of electronic parts and components coming from all of the supply chain economies together. The weights (wi,t ) can be combined with bilateral exchange rates between the upstream countries and the countries importing the final electronics goods using the formula: wr er j, t =



wi, t ∗ r eri, j, t .

(3.3)

i

where wr er j, t is the weighted exchange rate between the upstream countries providing electronic parts and components to Malaysia and country j purchasing final electronic goods from Malaysia at time t, wi, t is the value of electronic components exported from supply chain economy i to Malaysia divided by the value of electronic components exported from all the supply chain economies to Malaysia at time t, and r eri, j, t is the bilateral real exchange rate between supply chain economy i and country j purchasing the final good from Malaysia. An increase in r er represents an appreciation of the Malaysian ringgit and an increase in wr er represents an appreciation in supply chain countries relative to the country purchasing the final goods. To specify the econometric model a battery of panel unit root tests and Kao residual cointegration tests are performed. They indicate that there is a positive trend in Malaysia’s final electronics goods exports. They also point to cointegrating relationships between the variables. Panel dynamic ordinary least squares (DOLS), a technique for estimating cointegrating relations, is thus employed. The estimated model takes the form: F E G j, t = β0 + β1 wr er j, t + β2 r er j,t + β3 y ∗j,t +

p 

α1,k wr er j, t−k

k=− p

+

p 

α2, k r er j, t−k +

k=− p

+ u i, j, t , t = 1, . . . , T ;

p 

α3,k y ∗j, t−k

k=− p

j = 1, . . . , N .

(3.4)

Here F E G j,t represents final electronics goods exports from Malaysia to country j, wr er j,t represents weighted exchange rates between the countries providing electronic parts and components to Malaysia and country j importing final electronics goods from Malaysia, r er j,t represents the bilateral real exchange rate between Malaysia and importing country j, and yj,t * represents real GDP in country j. Country j fixed effects and linear trends are included in the estimation. The Mark and Sul (1999) approach is used to allow for heterogeneity in the long run variances. The number of lags and leads are determined for each cross section by the Schwarz Information Criterion.

3 Supply Chain Resilience in the Global Financial Crisis …

61

Table 3.4 Panel DOLS estimates of Malaysia’s and ASEAN’s final electronics goods exports (1)

(2)

(3)

Malaysia

Malaysia, Philippines, Thailand

Singapore

Bilateral real

−1.86***

−0.78***

−0.36

Exchange rate

(0.34)

(0.16)

(0.24)

Weighted real exchange rate in upstream countries

0.21

−0.64***

1.29***

(0.37)

(0.20)

(0.28)

Real GDP

1.93***

1.94***

0.80

(ULC-deflated)

(0.69)

(0.51)

(0.54)

Adjusted R-squared

0.970

0.952

0.973

S.E. of regression

0.203

0.259

0.143

Country fixed effects

Yes

Yes

Yes

Heterogeneous linear trend

Yes

Yes

Yes

Number of cross sections

12

23

12

Sample period

1993–2014

1993–2014

1993–2014

No. of observations

235

456

232

Notes The number of leads and lags of the first differences of the independent variables are selected by the Schwarz Information Criterion. An increase of the bilateral real exchange rate implies an appreciation of the exporting country’s exchange rate. The weighted exchange rate in upstream countries represents exchange rates in upstream countries providing electronic parts and components to ASEAN countries exporting final electronics goods. The weights are determined by the share of parts and components coming from each of the upstream countries. Source Author’s calculations *** (*) denotes significance at the 1% (10%) level

The results are presented in Table 3.4. Column (1) presents results for Malaysia, column (2) for Malaysia, the Philippines, and Thailand, and column (3) for Singapore. The results in column (1) indicate that a 10% appreciation of the Malaysian ringgit would reduce FEG exports by 19%. The results also indicate that exchange rates in supply chain countries do not affect Malaysia’s exports. A 10% increase in real GDP in the rest of the world would increase FEG exports by 19%. The results in column (2) indicate that a 10% appreciation in the ASEAN exporting country (Malaysia, the Philippines, or Thailand) would reduce FEG exports by 8% and a 10% appreciation in supply chain countries would reduce FEG by 6%. A 10% increase in real GDP in the rest of the world would increase FEG exports by between 19%. Comparing exchange rate elasticities in Table 3.4 indicates that the Malaysian ringgit seems to matter more for Malaysia’s FEG exports than the exchange rates in other ASEAN countries matter for their FEG exports and the exchange rates in supply chain countries seem to matter less for Malaysia’s FEG exports than for FEG exports from the other ASEAN countries. GDP in the rest of the world matters a lot, as it did also in the case of China’s processed exports.

62

W. Thorbecke

For Malaysia, the large decline of FEG exports in response to an appreciation of the ringgit could reflect the fact that producers can relocate assembly from Malaysia to China in response to appreciations in Malaysia. The results for Singapore in column (3) are not compelling. Singapore’s role as an entrepôt may cloud inference.

3.6 Abating Risks to the Supply Chain 3.6.1 Protectionism Pressures Arising from China’s Surpluses If we correct processing and ordinary trade for round-tripping imports, China ran a surplus in 2015 of $422 billion in processing trade and $330 billion in ordinary trade. China’s combined surplus in processing and ordinary trade thus equaled $752 billion or 6.7% of Chinese GDP in 2015. This implies that China ran a huge surplus in the manufactured products that make up the lion’s share of processing and ordinary trade. While its surplus in ordinary trade appeared in 2014, its surplus in processing trade has remained a little less than half a trillion dollars for the last five years. Its total goods surplus, including other trade regimes such as customs warehousing trade, was over 5% of GDP in 2015 and its current account surplus just under 3%. According to Chinese balance of payments statistics, its current account surplus was below its merchandise trade surplus largely because of a deficit in tourism trade. Politically, the large surpluses in manufacturing goods are likely to lead to dislocation, extremist policies and protectionism in trading partners (Schwartz and Bui 2016). The discussion above indicates that free trade played an important role in the development of regional value chains. What are the prospects for the surplus declining? Large capital outflows from China heading for housing, agriculture, companies, and other foreign investments will keep the renminbi from appreciating and keep China’s current account surplus large. The capital outflows, however, represent a misallocation of resources, as the returns would be much higher from investing in the people of China. For instance, a study at Peking University found that air pollution in China reduces people’s life expectancy by 5.5 years (Chen et al. 2013). Another study found that polluting industries contaminated between 8 and 20% of China’s arable land and produced hundreds of “cancer villages” where residents die young because of exposure to toxins (see Chin and Spegele 2013). As another example, Fang et al. (2012) reported that the return to an additional year of education in the PRC equalled 20% per year. While schools are good for residents of Beijing and Shanghai, they are substandard for rural children. Rozelle

3 Supply Chain Resilience in the Global Financial Crisis …

63

(2010) noted that most rural children cannot afford pre-school, and that elementary school attendance is hampered by poor accessibility and long, dangerous commutes. Bad health, sanitation, nutrition, and psychology management also restrict students’ ability to learn. Problems such as anemia, vitamin deficiencies, and visual difficulties are prevalent and can be remedied inexpensively. For instance, one multivitamin with iron will address both anemia and vitamin deficiencies and costs about 3 U.S. cents per student per day (Rozelle 2010). Smith et al. (2014) found that 34% of 3–5 year olds and 40% of 8–10 year olds in Guizhou Province were infected with intestinal worms, and that being infected significantly worsened academic performance, working memory, processing speed, height for age, and body mass index for age. In addition, deworming medicines are cheap and safe. The State Council initially allocated 60 million renminbi for deworming in Guizhou, but then did not follow through because it was “not a priority” (Smith et al. 2014). Rural students will be the urban workers of tomorrow, and investing in them now will pay high dividends later. Currently China’s exports and imports remain imbalanced. Its surpluses are intensifying protectionist pressures abroad. Large capital outflows from China will cause foreign trade to continue to play a disproportionate role. If investments could be rechanneled into the health, education, and welfare of its people, China’s long run growth trajectory would rise, the process within China would be more peaceful and harmonious, and protectionist pressures abroad would abate.

3.6.2 Protectionism Pressures Arising from Exchange Rates in Supply Chain Countries Equation (3.2) indicated that exchange rates in supply chain countries play an important role along with the renminbi in China’s processed exports. What has happened to exchange rates throughout the supply chain? Figure 3.13 plots the CPI-deflated real effective exchange rate for the renminbi and for supply chain countries. It shows that China’s exchange rate has appreciated 47% between 2005Q2, when China abandoned its fixed peg to the U.S. dollar, and the end of 2015. By contrast, there has been a 4% depreciation in supply chain countries since the second quarter of 2005. This depreciation across supply chain countries has happened in spite of the fact that key supply chain countries have run large current account surpluses over this period. For instance, Taiwan’s surplus has averaged more than 10% since 2009. Korea’s surplus has been at least 6% of GDP since 2013. Singapore’s surplus has been at least 17% of GDP since 2009. Japan’s surplus was low in 2013 and 2014 but jumped to 2.9% of GDP in 2015. Since countries throughout the supply chain run surpluses because of their role in regional production networks, market forces militate for joint exchange rates appreciations. However, because regional economies compete in third markets, each

64

W. Thorbecke

Log of Real Effective Exchange Rtae

4.9 RMB

4.8 4.7 4.6 4.5 4.4

Weighted Exchange Rate in Supply Chain Countries

4.3 4.2 4.1 94

96

98

00

02

04

06

08

10

12

14

Fig. 3.13 The RMB real effective exchange rate (REER) and the weighted REER in supply chain countries. Source The Bank for International Settlements, China Customs Statistics, the International Monetary Fund International Financial Statistics, and calculations by the author

country resists appreciation. The two largest supplier economies, South Korea and Taiwan, are reeling from competition with China, implying that there is little prospect for either of their currencies to appreciate. Exchange rates between Japan and Korea have been particularly volatile. Before the onset of the Global Financial Crisis (GFC) in 2007, the Korean won appreciated against the yen. Then, as the GFC began, the yen’s appeal as a safe haven currency caused a major appreciation of the yen against the won and other currencies. In late 2012, however, Prime Minister Abe’s policies contributed to a major yen depreciation against the won. These wide swings in the yen/won exchange rate have caused dislocation and disruption in sectors such as the electronics industry where these two countries compete extensively (see Sato et al. 2013). A better strategy would be if countries maintained more intra-regional stability in their exchange rates, and if they appreciated together against the U.S. dollar in response to market forces. A joint appreciation would have an attenuated effect on real effective exchange rates in the region, since so much trade is with regional trading partners. It would also reduce outsized surpluses in regional value chains relative to Western countries and forestall protectionist pressures. One major obstacle is that East Asian countries still assign strong weights to the U.S. dollar in their exchange rate policies. One reason they do this is that another suitable anchor currency has not emerged in East Asia. Schnabl and Spantig (2016) noted that underdeveloped capital markets and limits on capital flows have restricted China’s ability to function as an anchor currency. The Chinese government has been

3 Supply Chain Resilience in the Global Financial Crisis …

65

pursuing a policy of internationalizing the renminbi. As China over time liberalizes its capital account and develops deep and liquid capital markets, its attractiveness as an anchor currency should increase. This would provide one path to reduce the role of the U.S. dollar in exchange rate policy in the region and to facilitate a joint appreciation against the U.S. dollar. However, renminbi internationalization and the renminbi’s rise as an anchor currency may take a long time. If East Asian countries adopt regime’s characterized by multiple-currency, basket-based reference rates with reasonably wide bands, the large surpluses that they are running in their current accounts and in regional value chains would cause their currencies to appreciate together against the U.S. dollar. To achieve this, countries in the region would have to give greater play to market forces in exchange rate determination and dialogue extensively about exchange rate policy.

3.6.3 Factors Affecting the Service Link Cost Kimura and Ando (2005) have shown that factors that increase the cost of linking geographically separated production blocks can hinder the slicing up of the value chain. The service link cost is higher when infrastructure, education, and health in downstream countries are poor, when the rule of law and private contracts are not enforced, and when intellectual property is not protected. How are downstream supply chain countries in Asia doing in areas such as these? One way to shed light on this is to examine data collected by the World Economic Forum (WEF 2016). The WEF surveys more than 13,000 executives in 144 economies concerning the competitiveness and the ease of doing business in different economies. The evidence presented below is also summarized in Table 3.5. Indonesia has fallen to 62nd place in the world in infrastructure and 53rd in terms of the quality of institutions in the 2015–2016 survey. It is doing better at fighting corruption, with the WEF (2016) reporting that all measures related to bribery and ethics are improving. In public health it ranks badly (96th) and its weakest ranking is in labor market efficiency (115th). Malaysia ranks as one of the 20 most competitive economies in the world. The WEF (2015) reports that it ranks well across the board, and that it excels in goods market efficiency (6th) and financial market development (9th). Malaysia also improved 13 places in technological readiness. The Philippines is weak in terms of infrastructure (90th) and health and primary education (86th). In the area of infrastructure, roads (97th), ports (103rd), air transport (98th), and electricity and telephone infrastructure (95th) are concerns. For health and primary education, tuberculosis (126th) and primary education enrollment rate (100th) need to be addressed. Ethics and corruption (91st) are also problem areas.

66

W. Thorbecke

Table 3.5 Taking stock of how Asian countries can improve their business climates Country

Salient considerations (rankings compared to 144 countries are in parentheses)

China

China’s institutions (51st), infrastructure (39th), and health and primary education (44th) are acceptable. Its transport infrastructure (21st) is excellent

India

India’s electricity and telephone supply ranks 115th. The quality of India’s institutions has jumped 10 places to 60th, although in protection of intellectual property India ranks 103rd

Indonesia

Indonesia ranks 62nd in infrastructure and 53rd in the quality of institutions. Measures related to bribery and ethics are improving. It ranks badly in terms of health (96th) and labor market efficiency (115th)

Malaysia

Malaysia ranks well across the board and is one of the 20 most competitive economies in the world

Philippines

The Philippines is weak in terms of roads (97th), ports (103rd), air transport (98th), and electricity and telephone infrastructure (95th) are concerns. Tuberculosis (126th) and primary education enrollment (100th) need to be addressed. Ethics and corruption (91st) are also problem areas

Thailand

Thailand needs improvement in intellectual property protection (113rd), public trust in politicians (118th), bribes related to imports and exports (102nd), and the quality of primary education (89th)

Vietnam

Vietnam should improve property rights (96th), irregular payments and bribes (106th), the accountability of private institutions (126th), the quality of roads (93rd), the incidence of tuberculosis (103rd) and HIV/AIDS (109th), and the state of higher education and training (99th)

Source World Economic Forum (2016) Note Ranking data come from surveys of 13,000 executives by the World Economic Forum (2015). The executives rate 144 economies in several areas related to competitiveness and ease of doing business

Thailand needs improvement in intellectual property protection (113rd), public trust in politicians (118th), bribes related to imports and exports (102nd), and the quality of primary education (89th). Vietnam should improve in property rights (96th), irregular payments and bribes (106th), the accountability of private institutions (126th), the quality of roads (93rd), the incidence of tuberculosis (103rd) and HIV/AIDS (109th), and the state of higher education and training (99th). China ranks 28th overall in competitiveness. Its institutions (51st), infrastructure (39th), and health and primary education (44th) are all acceptable. Its transport infrastructure (21st) is excellent. India’s electricity and telephone supply ranks 115th, and the WEF (2015) reports that less than one in five Indians access the Internet regularly and that less than two in five own cell phones. The quality of India’s institutions has jumped 10 places to 60th, although in protection of intellectual property India ranks 103rd. Thus, while the business climate in India is improving, there is room for further improvement in electricity supply, protecting intellectual property, and fighting corruption.

3 Supply Chain Resilience in the Global Financial Crisis …

67

Most downstream countries in Asian value chains can thus ameliorate their business climates. As they do so, they will attract more FDI and strengthen not only their own economies but also the resilience of East Asian value chains. Attracting FDI offers the promise of promoting the transfer of technology and managerial knowhow to small and medium sized enterprises (SMEs) in the host country. Kimura and Lim (2010) argued that governments in emerging Asia should maintain an FDI friendly environment and be patient as industrial agglomeration takes root. Initially MNEs may not procure much from SMEs in the host countries. However, competitive pressures will cause MNEs to begin looking to local SMEs for parts and components. Kimura and Lim noted that MNEs will transfer technologies to local firms and entrepreneurs so that the MNEs can obtain parts and components that meet their requirements in terms of price, quality, and timing. Education and human capital development play important roles in facilitating the transfer of technology from MNEs to local firms (Kiyota et al. 2006). Investing in human capital needs to progress in stages. Heckman (2005) observed that it is beneficial for mothers to spend time with their young children. Thorbecke et al. (2013) emphasized that children in emerging Asia need to receive adequate nutrition, healthcare, and primary education. Yusuf et al. (2003) noted that high school students need high quality education in science and math and university students need scientific and engineering training. The technical training will help workers in host countries to become involved in more remunerative activities such as the engineering and design aspects of production. The educational system should focus on providing students with marketable skills that businesses need.11 Researchers at the World Bank have emphasized that governments should not undertake educational reform by fiat, but instead after an open discussion between schools, businesses, universities, parents, students, and the government (see, e.g., Bodewig 2012). Bodewig (2012) reported that Vietnamese employers seek engineers and white collar workers who possess, not just technical skills but also the ability to think critically, solve problems, and communicate. Bodewig emphasized that employers in other countries also desire these qualities. In order to nurture these skills, Sawa (2013) argued that a well-rounded education is important. He recommended that engineers and other specialized workers both before and during college receive training in literature, foreign languages, world history, and other liberal arts subjects in addition to their technical training. He noted that these seemingly unrelated subjects would help future workers to acquire the skills of creativity and logical thinking.

11 These

paragraph and the next one draws on Thorbecke (2013).

68

W. Thorbecke

3.7 Conclusion Production networks in the electronics industry have emerged and now crisscross East Asia. China is becoming more and more central within these networks. The networks have multiplied efficiency gains, led to technology transfer to developing and emerging countries, and caused prices of final goods to plummet. This chapter has traced the evolution of these production networks over time. Several factors contributed to lowering the cost of linking geographically separated production blocks and the slicing up of the value chain in Asia. China’s WTO accession gave investors confidence that China would follow the rule of law. In addition, China’s superb infrastructure in the Pearl and Yangtze River Deltas made producing there more attractive. The growing human capital of residents in urban China has also led to technology transfer and more of the higher value added activities being relocated to China. Low tariffs in the electronics sector have also facilitated the flow of electronic parts and components throughout the region. The paper also discusses other factors that have contributed to fragmenting production in the region. Partly because of the efficiency of production networks, East Asia has run large surpluses in manufacturing trade. These surpluses are likely to lead to dislocation, extremist policies and protectionism in trading partners (Schwartz and Bui 2016). One factor driving these surpluses is capital outflows from China. Previously the surpluses were driven by reserve accumulation by the People’s Bank of China, but now they are driven by private capital flows to real estate, companies, and other foreign investments. These capital outflows not only multiply the risks of protectionism and extremist policies among trading partners, they also lead to a misallocation of resources. Returns from investing in China would be much higher. Fighting air and water pollution, de-toxifying the soil, cleaning up cancer villages, and investing in the health and education of rural children, to name only a few, would yield economic and social returns that are much higher than the returns available from capital outflows. The government of China should thus prioritize investing in the people of China. These investments would also set the stage for innovation-led growth that China needs to sustain growth. Asian surpluses are also supported by undervalued exchange rates in upstream Asian countries. In spite of massive current account surpluses year after year in economies such as Taiwan, South Korea, and Singapore, their currencies have depreciated or stayed the same over the last several years. The results in this paper indicate that these exchange rates exert first order effects on exports within regional value chains. It would be desirable for regional currencies to appreciate together against the U.S. dollar in response to market forces and to maintain intra-regional exchange rate stability. To achieve this goal, countries in the region should focus more on real effective exchange rates, give greater play to market forces in exchange rate determination, and dialogue extensively about exchange rate policy.

3 Supply Chain Resilience in the Global Financial Crisis …

69

Asian countries should also continue to lower service link costs. Poor infrastructure, corruption, inadequate protections for intellectual property rights, insufficient electricity supply, health problems, and low quality education plague several downstream countries. The discussion in the previous section highlighted specific areas where each country could improve. Doing so would promote the slicing up of the value chain and help to maintain the healthy functioning of production networks in Asia. For firms, one of the great risks of exchange rate fluctuations is that they are uneven across East Asian countries. For instance, as discussed in Sect. 3.2, when the Korean won depreciated relative to the Japanese yen following the Global Financial Crisis, Korean electronics firms gained markedly in terms of price competitiveness relative to their Japanese peers and Japanese firms were decimated. On the other hand, when the yen depreciates rapidly against the won Korean firms suffer dislocation. In addition, exchange rate volatility between East Asian nations increases uncertainty and thus the service link cost that firms face when fragmenting production across national borders (Hayakawa and Kimura 2009). To mitigate these risks, policymakers in East Asia should seek to maintain relative stability in intra-regional exchange rates. Supply chains in Asia are remarkably efficient, as they have allowed high quality computers, cell phones, and other sophisticated products to be produced at ever lower prices. Partly because of this, East Asia is running enormous surpluses with the rest of the world in manufacturing products year after year and protectionist pressures abroad have reached the boiling point. If Asian exchange rates could appreciate together, the risks of protectionism would abate. A concerted appreciation would also maintain exchange rate stability in the region and facilitate the slicing up of the value chain. If regional economies could continue to improve their business climates, they would become more prosperous. Appreciating exchange rates and rising prosperity would allow Asian consumers to enjoy more of the fruits their labors instead of exporting these abroad. Acknowledgements This research was conducted as part of the project of the Economic Research Institute for ASEAN and East Asia (ERIA) ‘Reducing the Vulnerability of Supply Chains and Production Networks’. The author thanks Venkatachalam Anbumozhi, Michael C. Huang, Fukunari Kimura, Shandre M. Thangavelu, and other project participants for valuable comments and suggestions. The opinions expressed in this paper are the sole responsibility of the authors and do not reflect the views of ERIA.

Appendix This appendix presents data on intermediate goods inflows into individual East Asian countries from the region as a whole. It examines the three key industries within regional value chains, automobiles, clothing, and electronics (Figs. 3.14, 3.15, 3.16, 3.17, 3.18, 3.19, and 3.20).

70

W. Thorbecke 160 Electronics

Billions of U.S. dollars

140 120 100 80 60 40

Automotive

20 0 1980

Clothing 1985

1990

1995

2000

2005

2010

Fig. 3.14 Flow of intermediate goods in three leading industries to China. Note East Asia includes China, Japan, Indonesia, Malaysia, the Philippines, Singapore, South Korea, Taiwan, and Thailand. Intermediate goods in the electronics sector are defined as International Standard Industrial Classification (ISIC) product category 321 (electronic valves and tubes); intermediate goods in the clothing sector are defined as ISIC product categories 171, 172, and 173 (spinning, weaving and finishing of textiles, manufacture of other textiles, and knitted fabrics and articles); and intermediate goods in the automotive sector as ISIC product categories 342 and 343 (bodies for motor vehicles and parts for motor vehicles). Source CEPII-CHELEM database 32 Electronics

Billions of U.S. dollars

28 24 20 16 12 8

Clothing

4 0 1980

Automotive 1985

1990

1995

2000

2005

2010

Fig. 3.15 Flow of intermediate goods in three leading industries to South Korea. Note East Asia includes China, Japan, Indonesia, Malaysia, the Philippines, Singapore, South Korea, Taiwan, and Thailand. Intermediate goods in the electronics sector are defined as International Standard Industrial Classification (ISIC) product category 321 (electronic valves and tubes); intermediate goods in the clothing sector are defined as ISIC product categories 171, 172, and 173 (spinning, weaving and finishing of textiles, manufacture of other textiles, and knitted fabrics and articles); and intermediate goods in the automotive sector as ISIC product categories 342 and 343 (bodies for motor vehicles and parts for motor vehicles). Source CEPII-CHELEM database

3 Supply Chain Resilience in the Global Financial Crisis …

71

24 Electronics

Billions of U.S. dollars

20 16 12

Clothing

8 4 Automotive 0 1980

1985

1990

1995

2000

2005

2010

Fig. 3.16 Flow of intermediate goods in three leading industries to Japan. Note East Asia includes China, Japan, Indonesia, Malaysia, the Philippines, Singapore, South Korea, Taiwan, and Thailand. Intermediate goods in the electronics sector are defined as International Standard Industrial Classification (ISIC) product category 321 (electronic valves and tubes); intermediate goods in the clothing sector are defined as ISIC product categories 171, 172, and 173 (spinning, weaving and finishing of textiles, manufacture of other textiles, and knitted fabrics and articles); and intermediate goods in the automotive sector as ISIC product categories 342 and 343 (bodies for motor vehicles and parts for motor vehicles). Source CEPII-CHELEM database 30 Electronics

Billions of U.S. dollars

25 20 15 10 5 Automotive 0 1980

Clothing

1985

1990

1995

2000

2005

2010

Fig. 3.17 Flow of intermediate goods in three leading industries to Taiwan. Note East Asia includes China, Japan, Indonesia, Malaysia, the Philippines, Singapore, South Korea, Taiwan, and Thailand. Intermediate goods in the electronics sector are defined as International Standard Industrial Classification (ISIC) product category 321 (electronic valves and tubes); intermediate goods in the clothing sector are defined as ISIC product categories 171, 172, and 173 (spinning, weaving and finishing of textiles, manufacture of other textiles, and knitted fabrics and articles); and intermediate goods in the automotive sector as ISIC product categories 342 and 343 (bodies for motor vehicles and parts for motor vehicles). Source CEPII-CHELEM database

72

W. Thorbecke 20 Electronics

Billions of U.S. dollars

16

12

8

4 Automotive 0 1980

Clothing 1985

1990

1995

2000

2005

2010

Fig. 3.18 Flow of intermediate goods in three leading industries to Malaysia. Note East Asia includes China, Japan, Indonesia, Malaysia, the Philippines, Singapore, South Korea, Taiwan, and Thailand. Intermediate goods in the electronics sector are defined as International Standard Industrial Classification (ISIC) product category 321 (electronic valves and tubes); intermediate goods in the clothing sector are defined as ISIC product categories 171, 172, and 173 (spinning, weaving and finishing of textiles, manufacture of other textiles, and knitted fabrics and articles); and intermediate goods in the automotive sector as ISIC product categories 342 and 343 (bodies for motor vehicles and parts for motor vehicles). Source CEPII-CHELEM database 10 Electronics

Billions of U.S. dollars

8

6

4 Clothing 2

0 1980

Automotive 1985

1990

1995

2000

2005

2010

Fig. 3.19 Flow of intermediate goods in three leading industries to Philippines. Note East Asia includes China, Japan, Indonesia, Malaysia, the Philippines, Singapore, South Korea, Taiwan, and Thailand. Intermediate goods in the electronics sector are defined as International Standard Industrial Classification (ISIC) product category 321 (electronic valves and tubes); intermediate goods in the clothing sector are defined as ISIC product categories 171, 172, and 173 (spinning, weaving and finishing of textiles, manufacture of other textiles, and knitted fabrics and articles); and intermediate goods in the automotive sector as ISIC product categories 342 and 343 (bodies for motor vehicles and parts for motor vehicles). Source CEPII-CHELEM database

3 Supply Chain Resilience in the Global Financial Crisis …

73

10 Electronics

Billions of U.S. dollars

8

6

Automotive

4

2 Clothing 0 1980

1985

1990

1995

2000

2005

2010

Fig. 3.20 Flow of intermediate goods in three leading industries to Thailand. Note East Asia includes China, Japan, Indonesia, Malaysia, the Philippines, Singapore, South Korea, Taiwan, and Thailand. Intermediate goods in the electronics sector are defined as International Standard Industrial Classification (ISIC) product category 321 (electronic valves and tubes); intermediate goods in the clothing sector are defined as ISIC product categories 171, 172, and 173 (spinning, weaving and finishing of textiles, manufacture of other textiles, and knitted fabrics and articles); and intermediate goods in the automotive sector as ISIC product categories 342 and 343 (bodies for motor vehicles and parts for motor vehicles). Source CEPII-CHELEM database

References Acemoglu D, Autor D, Dorn D, Hanson G, Price B (2014) Import competition and the great US employment sag of the 2000. In: NBER working paper no. 20395 Ahmed S (2009) Are Chinese exports sensitive to changes in the exchange rate? In: International finance discussion papers no. 987 Ando M, Kimura F (2012) How did the Japanese exports respond to two crises in the international production networks? ‘The global financial crisis and the great East Japan earthquake. Asian Econ J 26:261–287 Autor D, Dorn D, Hanson G (2013) The China syndrome: local labor market effects of import competition in the United States. Am Econ Rev 103:2121–2168 Bodewig C (2012) What skills are employers looking for in Vietnam’s workforce? In: East Asia and Pacific on the rise. www.worldbank.org Calimag MG (2008) Intel prepares to close Philippine plant. In: Bloomberg Business Week CheY, Lu Y, Schott P, Tao Z (2016) Does trade liberalization with China influence U.S. elections, Yale University Manuscript Chen Y (2008) The legal climate for foreign investment in China after its WTO accession. Bond Law Rev 20, Article 2 Chen Y, Ebenstein A, Greenstone M, Li H (2013) Evidence on the impact of sustained exposure to air pollution on life expectancy from China’s Huai river policy. Proc Nat Acad Sci 110:12936–41. https://scholars.huji.ac.il/avrahamebenstein/publications/evidence Cheung Y, Chinn M, Qian X (2012) Are Chinese trade flows different? J Int Money Fin 31:2127– 2146

74

W. Thorbecke

Chin J, Spegele B (2013) China’s bad Earth. Wall Street J. http://www.wsj.com/articles/ SB10001424127887323829104578624010648228142 Chinn M (2005) Doomed to deficits? Aggregate U.S. trade flows re-examined. Rev World Econ 141:460–485 Colantone I, Stanig P (2016) Brexit: data shows that globalization malaise, and not immigration, determined the vote. Bocconi Knowl Fang H, Eggleston K, Rizzo J, Rozelle S, Zeckhauser R (2012) The returns to education in China: evidence from the 1986 compulsory education law. In: NBER working paper no. 18189 García-Herrero A, Koivu T (2007) Can the Chinese trade surplus be reduced through exchange rate policy? In: BOFIT discussion paper, no. 6/2007 Gaulier G, Lemoine F, Unal D (2011) China’s foreign trade in the perspective of a more balanced economic growth. In: CEPII working paper no. 2011-03 Gaulier G, Lemoine F, Unal-Kesenci D (2005) China’s integration in East Asia: production sharing, FDI, and high-tech trade. In: CEPII working paper no. 2005-09 Han I, Oh K, Yoo J (2012) Changes in competitiveness of LCD industry of East Asia: from bamboo capitalism to water lily. Int Telecommun Pol Rev 19:15–41 Hayakawa K, Kimura F (2009) The effect of exchange rate volatility on international trade in East Asia. J Japan Int Econ 23:395–406 Heckman J (2005) Emerging economic arguments for investing in the health of our children’s learning. Interview on 24 Oct 2005. http://www.childrenofthecode.org/interviews/heckman.htm Hiratsuka D (2011) Production networks in Asia: a case study from the hard disk drive industry. In: ADBI working paper 301 Kimura F, Ando M (2005) Two-dimensional fragmentation in East Asia: conceptual framework and empirics. Int Rev Econ Fin 14:317–348 Kimura F, Lim H (2010) The internationalization of small and medium enterprises in regional and global value chains. In: ADBI working paper 231 Kiyota K, Matsuura T, Urata S, Wei Y (2006) Reconsidering the backward vertical linkages of foreign affiliates: evidence from Japanese multinationals. In: Faculty of business administration working paper no. 243 Knight J, Wang W (2011) China’s macroeconomic imbalances: causes and consequences. World Econ 34:1476–1506 Kuijs L (2011) How will China’s external current account surplus evolve in coming years. In: East Asia and Pacific on the rise. www.worldbank.org) Mark N, Sul D (1999) A computationally simple cointegration vector estimator for panel data. Ohio State University Manuscript Montalvo J (1995) Comparing cointegrating regression estimators: some additional Monte Carlo evidence. Econ Lett 48:229–234 Obtsfeld M, Rogoff K (2000) Perspectives on OECD economic integration: implications for U.S. current account adjustment. In: Global economic integration: opportunities and challenges. Federal Reserve Bank of Kansas City, Kansas City Onyanga-Omara J (2016) Hard Brexit could cost U.K. $81 billion per year. In: USA Today, 3A Pierce J, Schott P (2016) The surprisingly Swift decline of U.S. manufacturing employment. Am Econ Rev 106:1632–1662 Qian N, Wen J (2015) The impact of Xi Jinping’s anti-corruption campaign on luxury imports in China. In: Working paper, Yale University. http://aida.wss.yale.edu/~nq3/NANCYS_Yale_ Website/resources/papers/QianWen_20150403.pdf Republic of China (2012) External trade development overview. Bureau of Foreign Trade, Taipei (in Chinese) Rozelle S (2010) China’s 12th 5 year plan challenge: building a foundation for long-term, innovationbased growth and equity. In: Paper presented at NDRC-ADB international seminar on China’s 12th five year plan. Beijing Sato K, Shimizu J, Shrestha N, Zhang S (2013) Industry-specific real effective exchange rates and export price competitiveness. Asia Econ Pol Rev 8:298–325

3 Supply Chain Resilience in the Global Financial Crisis …

75

Sawa T (2013) Lack of liberal arts education is sapping Japan’s creativity. In: The Japan Times Schnabl G, Spantig K (2016) (De)Stabilizing exchange rate strategies in East Asian monetary and economic integration. Singapore Econ Rev (forthcoming) Schwartz N, Bui Q (2016) Where jobs are squeezed by Chinese trade, voters seek extremes. In: The New York Times. http://www.nytimes.com/2016/04/26/business/economy/where-jobs-aresqueezed-by-chinese-trade-voters-seek-xtremes.html?mabReward=A3&action=click&pgtype= Homepage®ion=CColumn&module=Recommendation&src=rechp&WT.nav=RecEngine& _r=0 Smith S, Rozelle S, Medina A (2014) Executive function: a better way to evaluate the impact of deworming on children’s educational outcomes? In: Paper presented at the Freeman-Spogli Institute for International Studies, Stanford University, Palo Alto. http://fsi.stanford.edu/multimedia/ executive-function-better-way-evaluate-impact-deworming-childrens-educational-outcomes Stock J, Watson M (1993) A simple estimator of cointegrated vectors in higher order integrated systems. Econometrica 61:783–820 Thorbecke W (2015) Measuring the competitiveness of China’s processed exports. China World Econ 23:78–100 Thorbecke W (2013) FDI, production networks, and trade in Asia, ADB Manuscript Thorbecke W, Lamberte M, Komoto G (2013) Promoting learning and industrial upgrading in ASEAN countries. In: Thorbecke, Hsieh (eds) Industrial restructuring in Asia: implications of the global economic crisis. Sage, New Delhi World Economic Forum (2016) The global competitiveness report 2015–2016. World Economic Forum, Geneva Xing Y (2012) Processing trade, exchange rates, and China’s bilateral trade balances. J Asian Econ 25:540–547 Yoshitomi M (2003) Post crisis development paradigms in Asia. ADB-I, Tokyo Yusuf S, Altaf A, Eichengreen B, Gooptu S, Nabeshima K, Kenny C, Perkins D, Shotten M (2003) Redrawing the international boundary of the firm in East Asia: the evolution of international production networks. In: Yusuf et al. (eds) Innovative East Asia: the future of growth. World Bank, Washington DC

Chapter 4

How Do Production Networks Affect the Resilience of Firms to Economic and Natural Disasters: A Methodological Approach and Assessment in Japan, Taiwan and Thailand Michael C. Huang and Atsushi Masuda Abstract This chapter examines the economic impact of the disaster in the supply chain of electronic products, transportation equipment and machinery sectors in Asia. A Multi-region static CGE model based on GTAP v.9.0 is used to capture fullfledged propagation effects stemming from natural disaster to seven economies in Asia. Three hypothetical disaster cases are simulated based upon the collapse rates estimated: Tokyo earthquake, Taipei earthquake and Thailand floods. The simulation results confirmed that the economic impacts of the natural disaster are propagated to other countries in the region through changes in the production volumes and prices response. Economic impacts stemming from neighbouring economies can be positive or negative depending on the production and trade structure defined by the supply chain network. Increased resilience to the disaster will have an implication for increased macroeconomic stability by reducing propagation of the disaster impacts. Keywords Natural disaster · GTAP model · Disaster risk management · Supply chain JEL Classification Q54 · R13 · C68

This research was conducted as a part of the project of Economic Research Institute for ASEAN and East Asia (ERIA) ‘Reducing the Vulnerability of Supply Chains and Production Networks.’ The authors are deeply indebted to the members of this project for their invaluable suggestions. The opinions expressed in this paper are the sole responsibility of the authors and do not reflect the views of ERIA. M. C. Huang (B) Visiting Scholar, National Graduate Institute for Policy Studies (GRIPS), Tokyo, Japan e-mail: [email protected] A. Masuda Japan Credit Rating Agency, Tokyo, Japan © Springer Nature Singapore Pte Ltd. 2020 V. Anbumozhi et al. (eds.), Supply Chain Resilience, https://doi.org/10.1007/978-981-15-2870-5_4

77

78

M. C. Huang and A. Masuda

4.1 Introduction The Asia and pacific is the most disaster-prone region in the world. The region has borne brunt of the physical and economic damages of natural disasters due to the geological formation. According to data collected by the Centre for Research on the Epidemiology of Disasters (CRED), human toll from natural disasters in the last 100 years has been staggering. UNESCAP (2014), during the period between 2004 and 2013, natural disasters in Asia and the Pacific caused economic damage of over US$ 560 billion, of which 85.5% was accounted for by 29 upper-middle-income and high-income economies.1 The region lacked comprehensive natural disasters assessment capacities of planning and preparedness, the reports notes, adding that while it generated one quarter of the world’s GDP, it accounted for 85% of deaths and 42% of global economic losses due to natural disasters. The natural disaster such as earthquakes or floods causes both direct impact and indirect impact on the economic activities of the affected area.2 The direct impact includes destruction of the production facilities or infrastructure such as roads, transportation, and lifeline of water, electricity and natural gas supply with loss of lives. The indirect impact is propagation of the direct impact to area outside of the directly affected area of physical damage on infrastructure; furthermore, it affects economic activity and results in unemployment, exports decline and supply-chain disruptions. Recently lean and complex supply chain serves as devise to amplify the propagation of the disaster impacts. According to the CRED disaster database,3 the disaster impact shows that the casualty rates from the natural disaster are falling, whereas their economic cost is rising restlessly and massively. The amplification of the disaster impact propagation causes more economic losses may be due to the development of the lean and complex production networks. Resilience of the physical production facility or pertinent infrastructure can be enhanced through the adoption of the higher standards of the construction. However, resilience of the supply chain cannot be reinforced just by the physical investments for the safety; the effective disaster preparedness would require the foreseen impact assessments. The economic development has been rapid in many countries of the region of Asia and Pacific. As economies grow, so does the value of the infrastructure and assets that could potentially increase its vulnerability and being easily disrupted by a natural disasters. According to UNISDR, during 1970 and 2013, the world reported over US$ 2.8 trillion in economic losses from natural disasters. Asia and the Pacific alone reported US$ 1.15 trillion of economic losses, amounting to 40.7% of the global 1 The

29 countries are Australia, Bangladesh, Brazil, Canada, Chile, China, Colombia, France, Haiti, Indonesia, India, Italy, Japan, Malaysia, Mexico, Myanmar, New Zealand, Pakistan, Portugal, Poland, Sri Lanka, Sweden, Taiwan, Thailand, Turkey, Vietnam, Ukraine, United Kingdom and United States. 2 Haraguchi and Lall (2015) include “undermined trust in public authorities” as intangible and unpriced floods damages. In this context, damages due to disrupted financial service should be counted as “indirect damage” which is not included in the usual loss estimates. 3 EM-DAT, Disaster Trend (http://www.emdat.be/disaster_trends/index.html).

4 How Do Production Networks Affect the Resilience of Firms …

79

total economic losses. Moreover, the share of Asia and the Pacific in the global total has shown an increasing trend, reaching half of global economic losses in recent years, as the experiences we learnt from the case of the Great East Japan Earthquake (GEJE) and Thai Flood in 2011. The impact of large-scale natural disasters is significant on national economies while production facilities, assets and infrastructure were destroyed, as well as human lives and their capabilities directly related to employment. Most of all, under the increasing interconnection and dependence on regional production network, the disaster impact will lead to supply chain disruption in the regional economy. Especially while the production assets mainly located on the lower land or dense area with less resilience against disaster due to a lack of available space and rapid development in the cities. These incapacities have put the key industries exposed to natural disasters and the regional supply chain disruption as potential economic exposures. Haraguchi and Lall (2015) observe that massive excess demands had arisen after the initial impact of large natural disaster with limited response in prices. As a result, quantitative rationing occurs since production constraint cannot avoid unmet excess demand. The firms behave like not rendering short term profit taking by maintaining their product price even under the situation of market shortage after the initial disaster impact. By reviewing the potential change on quantity and prices affected by a largescale natural disaster, firms may also be able to find a way to increase production or make some hedge from the expected losses. It may be rational economic behaviour to avoid possible loss of the reputation. Any calculation on the economic losses stemming from the natural disaster should take account of these volume effect and price effect. The research aims to unfold the impact of global supply chain disruptions by visualizing the disaster impact through displaying some key economic indicators. By interpreting the simulation results from an economic model, we hope to clarify the trade and production interdependence among Asia and ASEAN supply chain, as well as to provide tangible policy implications for disaster risk management. After the section of literature review, the research will proceed as following: (i) to simulate the possible impacts of earthquake and floods damages based upon the potential largescale hazard cases—a. Tokyo Earthquake (Japan), b. Taipei Earthquake (Taiwan) and c. Thai floods (Thailand); then, (ii) to examine the economic impact of the disaster in the supply chain network in the Asia. For the methodology, multi-region CGE model based on GTAP v.9.0 is used to capture full-fledged propagation effects of economic indicators stemming from natural disaster to seven economies in Asia; and finally, (iii) to make policy recommendations for Asia and ASEAN states on their supply chain resilience.

80

M. C. Huang and A. Masuda

4.2 Literature Review 4.2.1 Economic Losses of the Natural Disasters

Share in GDP (%)

Damages (billion Japanese Yen)

Economic damages and impact stemming from natural disasters are among the important source of the macroeconomic fluctuations. Large-scale natural disasters cause massive destruction of the infrastructure as well as production facilities. The production and commercial activities may be disrupted as a result. UNISDR (2015) refers direct disaster losses as directly quantifiable losses such as the number of people killed and the damage to buildings, infrastructure and natural resources, of which caused by the destruction of infrastructure caused by natural disasters. The loss estimated by the authorities normally include the physical damages of the construction structures such as houses, buildings, bridges, roads, ports, stations, and other lifelines. Figure 4.1 shows the estimated loss from the natural disaster and its ratio to the GDP in Japan. According to the bar chat, the direct economic losses of infrastructure facilities from the natural disaster vary significantly year by year. During 1980– 2000, the normal year saw the economic loss around 0.5% of GDP. After 2000, these levels went down less than 0.3% of GDP. The Japanese government invested in flood controls and building structure resilience. The notable exceptions to this trend is 1995 and 2011. In both year, the annual facilities losses from the natural disaster exceeded 1% of GDP. In 1995, Kobe earthquake caused unprecedented losses to the production chain and infrastructure network in the Kobe region of Japan, especially the freight handling capacity in the harbour area

Fig. 4.1 Economic losses due to natural disasters in Japan (1981–2017). Source White Paper on Disaster Management 2018, Cabinet Office of Japan. (http://www.bousai.go.jp/kaigirep/hakusho/ h30/honbun/3b_6s_16_00.html)

4 How Do Production Networks Affect the Resilience of Firms …

81

was severely affected, which resulting a substantial decline of the Kobe Port’s global ranking. The total mortality reached to 6434. According to the statistics made by AON (Catastrophe Insight), Kobe was believed to be relatively less exposed to the earthquake risk, thus preparedness was not too high in comparison to the other high risk areas. Total economic loss was estimated as 2.4% of GDP. In 2011, the GEJE hit the coastal region of northeast Japan with the total causality mounted more than 18,000. Total economic loss is estimated as 3.8% of GDP. These are the two typical large-scale disasters which hit Japan recently. These two earthquakes had quite a different pattern of cause of casualty. In Kobe, the main cause of death is crush of the building. 80% of human losses are caused by crush of building or flying furniture. In contrast, the main cause of death in GEJE is drawing in the Tsunami following the earthquake. The affected area was known possible Tsunami damage from the historical records of the Tsunami. On the other hand, indirect disaster losses include declines in production or revenue, or those subsequent or secondary results of the initial destruction (UNISDR 2015), such as business interruption losses due to the disruption of supply chain. In 2011, GEJE and the flooding in Thailand offered an example of the global indirect impacts from local events of northeast part of Japan and greater Bangkok area in Thailand. Table 4.1 shows top 10 costliest economic loss disaster event recorded during 1980 and 2013. This table shows the estimate of the total economic loss that included direct and indirect losses incurred as a result to the natural disaster. Thus the figures are different from one shown in Fig. 4.1. Kobe earthquake and GEJE were recorded as top two costliest events, 3.5% and 2.3% of GDP respectively. Kobe is one of the transportation hub in Japan. In Tohoku area, there located several important production factories of the leading Japanese firms. Table 4.1 Top 10 costliest global economic loss event during 1980–2013 Dates

Country

Disasters

Economic loss (US$ bil.)

March 2011

Japan

Earthquake/Tsunami

210

August 2005

USA

Hurricane

125

1.0

January 1995

Japan

Earthquake

103

2.1

May 2008

China

Earthquake

85

2.4

October 2012

USA, etc.

Hurricane

72

0.5

Jul-Dec. 2011

Thailand

Flooding

45

14.1

January 1994

USA

Earthquake

44

0.6

November 1980

Italy

Earthquake

19

4.1

February 2011

New Zealand

Earthquake

15

10.2

September 1999

Taiwan

Earthquake

14

5.0

Current price

% of GDP 3.8

Source AON Benfield Annual Catastrophe Report 2013, EM-DAT, arranged by authors

82

M. C. Huang and A. Masuda

Thus in both cases, the initial economic losses have been amplified through their cascading effects through lean and complex supply chain in Japan due to the severely damaged transportation networks, causing the delay of goods interchange. The GEJE had also different dimension of the nuclear power station accidents. Fears of the nuclear contamination caused reputation issues of the agricultural and fishery products in the affected region. This factor did work in amplifying the economic impacts of the disaster. Macroeconomic variables are subject to the influence of various factors. Thus it is difficult to identify the pure result of the disaster impact propagation by pure empirical studies. The full-fledged economic impacts of the natural disaster need to be estimated by using an economic model of the affected economy. The model should take account of both the initial impacts and its propagation effects. Since the Asian economies are structured by a complex network of the supply chain, the multiregional model is appropriate to simulation of the disaster impact on the domestic economic activities as well as the foreign trades.

4.2.2 Impact Assessment of the Economic Losses of the Natural Disasters The estimation of the economic loss of the natural disaster has attracted an extensive research attention. Todo et al. (2015) conduct a seminal study based upon high frequency micro data on the firm transactions after the disaster in the case of GEJE. The study result reveals that the supply chain network tends to have positive impacts to facilitate the recovery process from the production interruption after the disaster. Firms with complex supply chain network tend to see prolonged period of the production interruption since the recovery process takes more time. However the business locations in the affected area could receive supports from other business points and were able to accelerate the recovery process. It concludes that the deepening of supply chain network will increase resilience of the economy to the disaster. Moriguchi et al. (2015) investigate the impact of GEJE on the prices and consumption behaviour. It found that enormous excess demands emerged in particular for the daily required goods. Nevertheless, the prices did not increase drastically. As a result, quantitative rationing occurred for the distribution of the daily goods. Households were divided into two categories; households that purchased more and household that could not purchase the required goods. It was also confirmed that the price increases were generally restrained, but many households were forced to purchase more expensive substitutes in order to fulfil the unmet demands. Generally speaking, the price response was restrained, but demand-supply gap expanded substantially and resulted in inefficient allocation of the resource, in particular for the search activities of the demanded goods. The above two studies drew precious lessons based upon the detailed observation on the behaviours of the household and firms during the disaster and recovery process.

4 How Do Production Networks Affect the Resilience of Firms …

83

Another class of the studies have tried to figure out the macroeconomic impacts of the natural disasters. Warren et al. (2010) present the outline of the Input-Output (IO) procedure for calculating the economy-wide economic impacts in supply chain as a tool of disaster consequence analysis. The procedure first estimated the final demand impacts and then calculate the upstream impacts by using the multipliers obtained from the input-output tables. Strictly speaking, there is some doubt as to whether the input-output table remains the same even after the severe destruction of the production facilities. However, the methodology to make adjustment of the IO table based upon the disaster damages have not developed yet. Thus the standard practice is to apply the same IO table for the sake of estimation of the disaster impacts. The implicit assumption is no single disaster is large enough to change the IO structure of the entire economy. An economy consists of many sectors, of which has interdependent correlation with one another. The change of intermediate good, value-added (labour and capital) input will make difference on the output and the final demands of household and government activities. The economic wide response to the disaster impact can be better captured by using Computable General Equilibrium (CGE) model. CGE model takes account of the price effects with which the regular IO model could not deal lineally. Under CGE model scope, firms will modify its production as the reaction to the shock on factor change, determining a new level of quantity price. A CGE model may serve as a description of an economy by using a system of simultaneous equations while the “general equilibrium” model implies that all markets, sectors that are modelled together with corresponding inter-linkage. A multi-sector CGE model could describe the whole economy and its transactions between diverse economic agents, such as production sectors, households, and governments (Sue Wing 2011). In disaster impact studies, CGE models are often being used for it could provide comprehensive and simultaneous changes on several indicators. For example, Rose and Guha (2004) apply the CGE model to analyse the lifeline losses from earthquakes and its resilience against disasters. Such impact assessment could be made through static and dynamic approaches. Rose and Liao (2005) use a CGE model to analyse the impact of water service disruptions by earthquakes and estimated resilience in various situations. Furthermore, these CGE models are used widely to analyse the aggregate welfare and distributional impacts of policies through multiple markets for various types of tax and subsidy (Adam 2013). The CGE model thus serves as a more adequate tool in order to assess volume and impacts caused by natural disasters. As noted in the above, the empirical studies found the initial quantitative rationing under the excess demands will be followed by price response. CGE model is rational choice in order to assess both volume and price impacts. For running CGE model, it only requires small data requirements (e.g., 1-year IO table) and the calibration method (Hosoe et al. 2010). For the disaster scale estimate, Huang and Hosoe (2016) uses geographic business census to calculate the sectoral aggregated impact, which could be calibrated into the shocks result from disasters. The model can manipulate backward or forward impacts on losses in sectoral capital stock as disaster “shocks.” Its economic impacts could thereby be derived from a shock, or the implementation of a specific policy after a disaster could thereby be

84

M. C. Huang and A. Masuda

captured. This approach is adapted to interpret disaster shocks on markets and prices in the short run. In this research, for better understand the supply chain disruption caused by largescale natural disasters, the impact assessment are essential for vulnerability identification. With the support of such evidence-based research, the foreseen potential losses and sectoral vulnerability could raise the risk awareness, and it will be much convincing for firms and government to take plausible countermeasure to strengthen the capacity for risk reduction. For this reason, we will use a multi-regional CGE model to investigate the change on various economic indicators and the interdependence of supply chains with policy implications.

4.3 Methodology 4.3.1 GTAP World Trade CGE Model In order to capture the impact worldwide simultaneously, we will use a multi-region, multi-sector CGE model, based on GTAP database (version 9.0). GTAP (Global Trade Analysis Project) is a large CGE model developed at Purdue University and used widely around the world. The GTAP v.9.0 database collates data for 57 production sectors in each of the 140 regions of the world, adjusted to make a consistent data set and liner intermediate coefficients for 2011. This has then been aggregated to 10 sectors in 10 regions. The model is closed by fixing the quantity of land, labour, capital, and natural resources; these factors are set immobile within a region (but mobile between sectors) for maintaining their geographical feature and for better viewing the vulnerability of global supply chain and the overall disaster impact in a nation-wide scope.4 The GTAP model has regional “households” (one per region) that aggregate taxes then distribute income within the region for savings and spending. These are different to private households, which receive money from the regional household for spending on taxes and purchases. It is a multi-regional model that includes households, producers, and government for each region, interactions between the regions, and global savings. The income comes from producers in the value-added they receive from capital, labour, and agricultural land. It also comes from taxes paid by private households, governments, producers, and other regions. Subsidies are treated as negative taxes. The income is then distributed to government and private household for expenditure, and to global savings. The private households and government both use their funds to purchase goods and services both from the producers in their own region and from other regions. Producers also sell their products to the savings sector as investment 4 If

these factors were set immobile, the disaster impact would just reveal proportionally as the shocks given in the scenarios due to the assumption of optimization for firms and labor.

4 How Do Production Networks Affect the Resilience of Firms …

85

goods, to other producers as intermediate goods, and import from and export to other regions. Different from the linear regression in the Input-Output model, CGE models use a non-linear function to come out with the optimal level of production from firms while minimizing the costs of all intermediate goods used. For production function of gross output, A Leontief-type function is employed. The gross output is made up of value added, an intermediate input, and an energy composite. Gross output is transformed into domestic goods and exports with a constant elasticity of transformation (CET) function, while constant elasticity substitute (CES) function is assumed for production of composite goods made with domestic goods and imports, following the assumption and illustration made by Armington (1969). It is implied that a share of a goods to be imported and exported. The commodity market and Armington composite goods are used by a representative household and the government as well as for investment and the intermediate input. The model is calibrated to the Armington elasticity of substitution provided by the GTAP Database v. 9.0 (Hertel 1996). In the commodity market. The Armington composite goods are used by a representative household and the government, as well as for investment and the intermediate input. Household utility depends on the consumption of various goods. The value-add input part, the factor variables are labour and capital, operated under Cobb-Douglas function. The reduction on factors are different from the usual economic shock like financial crisis or the recession because the assumed shocks are placed on the production function. Whereas, the impact may cause the change on economic activities and final demand as consequences. While shocks occur on factors, the sectoral production will also change accordingly, and furthermore, affecting the original equilibrium on output price, external trades, and finally the welfare. In the new equilibrium, the results of economic indicators could be observed and compared with the base run, which representing the scenario without natural disasters. Therefore, CGE models are capable tool for analysing the details of a macro economy driven by a price mechanism, all is similar to the reality. A diagram of the GTAP model structure is given in Fig. 4.2. In the CGE model scope, the interdependence of supply chain between regions could be illustrated in details, including production level, labour and price change, external trades, and welfare change. The functions show the interrelated economic activities reflecting on supply chain. For instance, the reduced production will change a region’s output, affecting other country’s import for it is production. And finally, the different consumption levels also determine a new welfare level of a region. The reaction of a region or sector in responding the disaster impact on their capital stock could be foreseen in the simulation results. The important functions are stated as followings: ’ ’ Household  s demand for goods  Government  s demand for goods   f  f p g ph, j F Fh, j − S p − T d X i = αpqi ph, j F Fh, j − S p − T d X i = αpqi h, j

h, j

86

M. C. Huang and A. Masuda

Expenditure

Expenditure

Fig. 4.2 The GTAP world trade model structure. Source Rearranged from Hertel (1996), King (2012)

with the savings, budget consts. are modified 

f

ph F Fh − S p − T d

h

where pq = price of Armigton’s goods, pf = factor price, S p = private saving, T d = tax Armington composite good production function  η η 1/ηi ∀i Q i = γi δm i Mi i + δdi Di i where Qi = Armington’s composite good; γ i = scale par. func.; δ= share par.; ηi = substitution elasticity par. Import demand function Domestic good demand function  ηi  ηi q 1/(1−ηi ) q 1/(1−ηi ) γi δm i pi γ δd p Q i ∀i Di = i pdi i Q i ∀i Mi = 1+τ m pm ( i) i i where M i = sectoral import; τ i = import tariff rate; pm = import price; pd = domestic price

4 How Do Production Networks Affect the Resilience of Firms …

87

Gross domestic output transformation function  1/φi φ φ ∀i Z i = θi ξ ei E i i + ξ di Di i where Z i = sectoral output; ξ = share par.; E i = export good; Di = domestic good; ei = exchange rate; φ i = trans. elasticity par. Export supply function Domestic good supply function

φ

φ 1/(1−φi ) 1/(1−φi ) z z i θi ξ ei (1+τi ) pi θi i ξ di (1+τiz ) piz Ei = Z i ∀i Di = Z i ∀i pe pd i

i

where τ z = production tax; pz = output price; pe = export price The welfare function: Expenditure func. (Minimum expenditure attaining a utility level)

 ep(pq , UU) = minXP pq · UU(Xp ) = UU before (UU0) & after (UU1) a shock (Welfare = Equivalent variations (EV))    WF = ep pq0 , UU1 − ep(pq0 , UU0 where ep(·): expenditure func.; XP: Consum.; UU: utility level; UU(·): utility func.

4.3.2 Region and Sector Classification We use the data from the GTAP v.9.0 database which contains IO table of 57 sectors from 140 regions. The research focuses on the countries that play important roles in the global supply chain and the regional production networks (Table 4.2). We aggregate 10 major sectors from the database (Table 4.3) and mainly concentrate on the industries that have high interdependence: Electronic products (EEQ), Transportation equipment (TEQ) and Machinery (MCH). We will observe the impact of these regions and selected sectors under the disaster scenarios. The focal points for supply chain disruptions are the sectoral output, change of trade volume, trading partner and welfare.

4.3.3 Disaster Scenarios for Simulation For the disaster impact assessments of the supply chain disruption, we created three most potential disaster scenarios in cities in Asia—Tokyo (Japan), Taipei (Taiwan) and Thailand. The large-scale natural disaster scenarios are based on the assumed

88 Table 4.2 Region for analysis

Table 4.3 Industry category

M. C. Huang and A. Masuda Abbreviation

Region

TWN

Taiwan

CHN

China

KOR

South Korea

JPN

Japan

THL

Thailand

MAY

Malaysia

IDO

Indonesia

RASEAN

Rest of ASEAN

ROA

Rest of Asia

ROW

Rest of World

Abbreviation

Sector

AGR

Agriculture

ENG

Coal, Oil, Gas, Petroleum

FOD

Food manufacture

MAN

Textile and printing

STL

Steel and metals

EEQ

Electronic products

TEQ

Transportation equipment

MCH

Machinery and manufacture

TRS

Transportation

SRV

Service

LAB

Labour endowment

damages due to the destruction in buildings and other infrastructure such as transportation and communication system. The building damage will be recalculated as the reduction on sectoral capital stock based on the scientific data provided by the engineering research centre and existing literature. The details of the sectoral estimate and regional aggregation are specified in the annex. By implementing the shocks on capital stock, based on the cobb-Douglas function of labour and capital, the CGE model could generate the new equilibrium of the new level of expenditure, production and other indicators such as prices, external trades, and welfare. By viewing the condition before and after shocks, we may understand how the impact beforehand and identify reactions that a country and sectors may take in responding the disruption of global supply chain.

4 How Do Production Networks Affect the Resilience of Firms …

4.3.3.1

89

Tokyo Earthquake (Japan)

Despite Tokyo’s importance as the capital city of Japan and the role in Asia-pacific for the commercial activity, it is extremely vulnerable to earthquakes for its location situating between of four plates. The Great Kanto Earthquake in 1923 smashed the city with the impact of 50% of Japan’s GDP. Although the disaster preparedness has been reinforced with advance architecture technology in the 21st century, great losses of a potential large-scale earthquake remain inevitable. According to the Japan’s cabinet office report of different scenarios of Tokyo Inland Earthquake in 2015, a scenario with the most severe disaster impact was chosen—the Tokyo Bay Earthquake of M7.3. In the experience of Great Kanto Earthquake, the massive impact was not only caused by the earthquake, but the fire that came afterward, being regarded as a compound disaster. The Tokyo Earthquake scenario also includes the potential fire disasters with number of houses burned estimated. For the assumed damage, we thus use the estimated collapsed and burned house rate in the affected area of Tokyo Earthquake. The house damage rates multiplied by the output share from affected regions, based on the prefectural level Input-output table of the affected area.

4.3.3.2

Taipei Earthquake (Taiwan)

In 1999, the impact of an M7.6 earthquake struck the central Taiwan brought massive shock to the semiconductor production networks. It greatly affected that the global supply chain of computer manufacturing.5 Taiwan’s capital city—Taipei is also located in the area that is prone to earthquakes. Nearby the city, an active fault of 30 km leaned to the city where most of commercial, political and production activities are concentrated. According to the Taiwan Seismic Simulation Database (TSSD), an M7.5 earthquake triggered by the Shan-jiao fault is believed to bring devastating loss to the buildings and communication system. The assumed damage of an M7.5 earthquake in Taipei was based on a prior study on the compound disaster by Huang and Hosoe (2016). Different from the compound disaster assumption of an earthquake with a power crisis,6 the study will only use the assumption of earthquake impact and apply their assumed sectoral damage with some essential arrangement to fit the selected sectors.

5 Chichi,

Taiwan Earthquake of September 21, 1999 (M7.6) (http://www.absconsulting.com/ resources/Catastrophe_Reports/Chichi-Taiwan-1999.pdf). 6 The power crisis was triggered due to the shutdown of all the nuclear power plants.

90

4.3.3.3

M. C. Huang and A. Masuda

Thai Flood (Thailand)

The Manson season in the Indo-china has brought abundant rainfall and sometimes make flood. The impact of the Thai flood in 2011 lasted for months and caused disruption on automobile parts and hard disk drive (HDD), as well as other electronic parts. Isono and Kumagai (2014) and Kirdruang (2013) make an impact assessment of the 2011 Thai Flood, analysing the sectoral impact and economic losses. Since the disaster type of flood and earthquake are different. Instead of using building collapse rate for damage estimation. Based on their assumed sectoral damage of the existing classified sector in the study. For some sectors such as energy, transportation sectors and the decrease on labour endowment, we use the national account statistics and affected labour force to be multiplied by the labour ratio. Although the flood type of natural disaster may be different from earthquake, the consequences of the factory incapability for operation are no different, especially in static model. If it were the dynamic model, the recovery speed may need adjusting under circumstances.

4.3.3.4

Disaster Sectoral Impact for Simulation

The Table 4.4 shows the sectoral impact on the capital stock and labour endowment. The reduced capital stock represents the damage of production capacity, namely, the collapsed building including factories. For the scenario of Tokyo Earthquake and Taipei Earthquake, the reduction rate is estimate based on the geographical industrial agglomeration. This implies that if the earthquake occurs in other places or city, the impact and reduction rates may not be the same. For the scenario of Thai Flood, the sectoral impact is calibrated from Isono and Kumagai (2014). Table 4.4 Sectoral impact on capital stock and labour endowment (%)

Sector

Tokyo earthquake

AGR

−1.2

−5.7

−17.6

ENG

−2.7

−6.1

−23.4

FOD

−1.8

−3.9

−15.6

MAN

−1.7

−9.6

−11.1

STL

−1.4

−6.4

−18.0

EEQ

−1.4

−11.6

−15.0

TEQ

−2.1

−4.1

−19.8

MCH

−2.1

−6.1

−15.6

TRS

−3.1

−13.5

−11.1

SRV

−3.4

−8.2

−2.8

Labour endowment

−2.7

−7.4

−3.8

Source Recalculated by authors

Taipei earthquake

Thai flood

4 How Do Production Networks Affect the Resilience of Firms …

91

The reduced labour endowment does not necessarily mean the mortality of labour endowment, it also indicate the temporary unemployment due to the damage on the collapsed buildings or the incapability of factories. The estimate sectoral impacts are used as the reduction for capital and labour endowment. Table 4.4 show the comparison of the sectoral impact in the three assumed large-scale disaster scenarios.

4.3.3.5

Disaster Impact Indicators for Simulations

For analysing the supply chain disruption, we will interpret the impact from the three scenarios separately from their changes of (a) Sectoral output and price, (b) External trades, (c) GDP and (d) Welfare. These are the key indicators to observe the reaction related to the supply chain disruption that could be viewed immediately after natural disasters. However, based on the static and macro model scope, the linkage with foreign direct investments (FDI) or affiliates could not be captured in the study.

4.4 Simulation Results The study uses a multi-region and multi-sectoral CGE model to simulate the largescale natural disasters by reducing the capital stock and labour endowment from the disaster-hit country. Based on Cobb-Douglas production function, the reduced factors will force firms to adjust their production and export. Meanwhile, the trading partner will also need to adjust their production and external trades in responding the change. The consumption change will also affect the demand while making the price change, creating a new equilibrium of the optimal of collective selections by firms. The simulation results show the change after the assumed disaster shock. The potential disaster impacts provide information of sectoral and regional vulnerability, which could give implication for risk identification to make better preparedness beforehand. We are able to observe the loss and gain of how a country and industries react and switch their trading partners in the global supply chain disruption.

4.4.1 Tokyo Earthquake In the changes on output and prices (Fig. 4.3), Japan’s production of three sectors show a decline of 4–5% while TEQ and MCH showed a slight increase. While Taiwan and Thailand were affected slightly, Korea’s production on three sectors increased by 1–3%. All these three target sectors in Malaysia Indonesia show slight increase while R-ASEAN significant increase in EEQ and MCH by 3–4%, whereas TEQ showed only an increase by 1%. Despite Korea and R-ASEAN’s increase output, the price did not change significantly, where 0.5% price increase could be found in Taiwan,

92

M. C. Huang and A. Masuda

Fig. 4.3 Changes on output (left) and price (right) (%)

Fig. 4.4 Changes on export (left) and import (right) (%)

Korea, Thailand and R-ASEAN, implying their linkage with Japan’s decrease of aggregate demand. For changes on external trades (Fig. 4.4), Japan’s EEQ showed an increase by 5% in all three sectors. On the other hand, Korea’s export in the three sectors had increased by 2–3%; The R-ASEAN export has increased drastically in EEQ and MCH by 3%, while only 2% increase revealed in TEQ. Japan’s import increase slightly by less than 1%; all other countries show slight fall, expect for the EEQ in R-ASEAN. As for the change of trading partner (Fig. 4.5), in EEQ, Japan has decreased its import from all the regions from 6–7%, only R-ASEAN by less than 6%. The import of Taiwan and Korea showed a slightly decrease, while all other regions have increases, especially R-ASEAN, increase by 2–5%. For TEQ, similar to EEQ, the import of Japan has decrease of 4–6% from all regions, especially Korea and RASEAN. Most regions show increase, especially Korea by 1–4% while other regions have only 1–2%. Taiwan, however, show a decrease of 1.5% from R-ASEAN. As for MCH, Japan shows decline of 6–7% from most of regions. While most regions show an increase, Korea and R-ASEAN have most significant increase by 3–5%.

4 How Do Production Networks Affect the Resilience of Firms …

93

Fig. 4.5 Changes on trade volume with partner country (%)

In GDP and welfare change (Fig. 4.6), Japan’s GDP would be affected by 3.5%, while Korea and the R-ASEAN showed a strong growth by 1.3% and 1.9% respectively. Other regions show very insignificant impact by less than 0.1%. The welfare is determined by the gross production that domestic household could consume. The welfare of Japan show great reduction by nearly US$ 89,442 million while only little decline on other regions, such as US$ 6.5 for Rest of the world; US$ 1508 million for Korea; US$ 1208 million for R-ASEAN.

94

M. C. Huang and A. Masuda

Fig. 4.6 Changes on welfare and GDP

4.4.1.1

Sectional Remarks

While the assumed M7.3 inland earthquake hit Tokyo, significant decrease in Japan’s production the three target sectors despite the small impact of its sectoral impact accounts 1–2%, implying that the production could be affected by the loss of logistic system because the Tokyo Port is now the biggest port in Japan. Meanwhile, the gap of production capacity will quickly fulfilled by Korea. Giving the increasing of original equipment manufacturer (OEM) trend for components in Asia production networks, the effect of production substitute will be greater than of which in Kobe Earthquake and GEJE. On the other hand, regarding to the import flows, there showed rigidity import demand of Japanese product for Korea and R-ASEAN. The R-ASEAN will play a vibrant role in the supply chains, especially in EEQ and MCH sectors. Moreover, the stimulus production will increase Korea and R-ASEAN’s GDP growth.

4.4.2 Taipei Earthquake In changes on output and price (Fig. 4.7), Taiwan’s production of three selected sectors shows a decrease of 5–8%%. Slight decrease also showed in EEQ and MCH of Indonesia respectively. The production in Korea and Japan shows an increase of 2–4%, this change indicated that Korea and Japan will quickly increase production to cover the gap in Taiwan, similar to the period of Taiwan’s Chichi Great Earthquake in 1999. As consequences, Taiwan’s output price has fallen by 1%, nevertheless, the change in Japan, Korea and ROA is insignificant. As well as for other regions, indicating that ASEAN countries may gradually increase their production to respond the demand in Korea and Japan. Under the Taipei earthquake, Taiwan’s export of all three sectors show decrease by about 5% (Fig. 4.8). It could also be found that Indonesia also shows a 2–5% decrease in export especially in EEQ and MCH. Meanwhile, Korea and Japan show a strong growth in export by 3–5% in all target sectors, ROA also increase export

4 How Do Production Networks Affect the Resilience of Firms …

95

Fig. 4.7 Changes on output (left) and price (right) (%)

Fig. 4.8 Changes on export (left) and import (right) (%)

by 1–3%. The import in Taiwan decreased significantly by 5–7%, especially TEQ and MCH. Japan shows a decline in three target sectors by more than 2%, for Korea it is less than 1%. The import increase of the three sectors also shows increase in Indonesia, implying that these countries have switched trading partners in the production network. As for changes on trade volume of trading partner (Fig. 4.9), the import flow of target three sectors shows decrease from most regions showed a decrease, and the import from Taiwan decreased significantly by 4–8%, while Korea increases and Japan even shows a substantial increase from China, Korea and ASEAN countries. This trend basically applies to all three sectors. On the other hand, Indonesia reduces import from most of countries, Taiwan at the most. It is also notable to see that for China, Thailand, Malaysia and R-ASEAN reduces import from Taiwan by 7– 10%, followed by Japan with 2–3% decrease, indicating the strong interdependence among these countries. The change of trade volume indicates that the disruption of supply chain in Taiwan could be quickly covered by Korea and Japan, while necessary intermediate goods of EEQ, TEQ and MCH could be sourced from China, Thailand, Malaysia and R-ASEAN. Indonesia, however, may face a situation that such merchandise from ASEAN neighbours may prioritize their destination to Japan and Korea, resulting the substantial decrease in import.

96

M. C. Huang and A. Masuda

Fig. 4.9 Changes on trade volume with partner country (%)

The GDP of Taiwan shows a decrease of 7%, and China, Indonesia the ROW also show slight decline by 0.1%, 1.2% 0.2% respectively (Fig. 4.10). Meanwhile, GDP growths may be seen in Korea (1.3%), Japan (0.9%), Rest of Asia (0.7%), Thailand (0.2%) and R-ASEAN (0.1%), implying that the production has much to be substituted to these three regions. The welfare decrease shows in in Taiwan (US$ 17,043 million), Japan (US$ 5919 million), Korea (US$ 1411 million), and the Rest of Asia (US$ 1797 million). However, welfare increase shows in Thailand (US$ 13 million), Indonesia (US$ 121 million) and the ROW (US$ 1354 million).

4 How Do Production Networks Affect the Resilience of Firms …

97

Fig. 4.10 Changes on welfare and GDP

Although the supply chain disruption occurs in Taiwan may be a stimulus for GDP growth, the unstable prices and supply may cause negative impact on the welfare inferior in the production network.

4.4.2.1

Sectional Remarks

Similar to the disaster case in Japan, the assumed M7.5 earthquake in Taipei may cause a production and trade substitution from Taiwan to Korea, Japan, ROA and RASEAN. The simulations results on GDP growth and welfare also show that although the production capacity could be covered by Japan, Korea or other Asian countries, there is negative impact on welfare due to Taiwan’s competitiveness in producing some key parts so that the production efficiency could be distorted. In EEQ and TEQ, the impact on Indonesia was smaller than Taiwan but showed a notable consistency, implying that the production network has been more closely connected. Moreover, the disaster impact on Taiwan may also significantly affect the import from China to Taiwan, which takes considerable parity in Taiwan’s intermediate production in EEQ and MCH.

4.4.3 Thai Flood Thailand’s production of the three targeted sectors has decreased by approximately 6–8% (Fig. 4.11), surprisingly, Korea also shows decrease from 2–5% especially in EEQ and MCH. On the other hand, the R-ASEAN showed a substantial increase by 12% in EEQ and MCH, but only 2% in TEQ, implying their substitute features on EEQ and MCH sectors and strong dependence on Thailand’s TEQ sector. Indonesia also showed slight increase in EEQ and TEQ by 0.2–1%, but 0.4% decrease in MCH. Except for the increase in Japan by 1%, Taiwan, China, Malaysia, Indonesia and ROA’s changes are insignificant. Unlike the cases of Taiwan indicated previously, the Thailand’s output price only showed an increased only by 0.5%, but R-ASEAN increase by more than 1%. Output prices also show decrease in Korea and Malaysia

98

M. C. Huang and A. Masuda

Fig. 4.11 Changes on output (left) and price (right) (%)

by 0.3–0.6% while these changes are insignificant in Taiwan, China, Japan, Indonesia and ROA. The changes in export after the Thai flood is much diversified (Fig. 4.12). Thailand’s export in the targeted sectors showed a decrease by 3–4%, while Indonesia also shows a similar trend by smaller scale of 1%. On the other hand, export increase could be seen in Korea (5–8%), Japan (4–5%), Taiwan, China and Malaysia by 1–3%. Interestingly, R-ASEAN show and increase in export by 9–10% in EEQ and MCH, 5% in TEQ respectively. Thailand’s import of the target sectors showed a decrease by 3–4%, R-ASEAN also showed decrease by 5% in TEQ. Import changes of Taiwan, China and Japan are in insignificant as Malaysia and Indonesia show an increase by 1% in EEQ and MCH while R-ASEAN shows a substantial import increase in EEQ. The relatively smaller changes comparing with export could imply that these countries are strengthening the domestic production capacity to respond the export demand. In the changes on import volume for the trading partners (Fig. 4.13), substantial increase R-ASEAN shows in EEQ and MCH; Thailand shows import decrease but Korea changes even more. Looking into the component, imports from Japan show the biggest proportion. For EEQ and MCH, R-ASEAN reveals to increase the import

Fig. 4.12 Changes on export (left) and import (right) (%)

4 How Do Production Networks Affect the Resilience of Firms …

99

Fig. 4.13 Changes on trade volume with partner country (%)

from Japan, Taiwan, China and Korea while changes are insignificant in other regions. TEQ seems to be the most rigidity import composite comparing with EEQ and MCH, implying Thailand’s TEQ competiveness. For EEQ and MCH, the simulation results show that R-ASEAN could develop a new production network rapidly by importing the key parts from leading countries. The TEQ production requires much more investment than other two sectors, indicating the production substitute could not be easily made.

100

M. C. Huang and A. Masuda

Fig. 4.14 Changes on welfare and GDP

The GDP of Thailand would reduce by 7%, other decreases show in Korea (2.8%) and ROW (0.1%). On the contrary, R-ASEAN shows vibrant growth by 7.5%, while increases could be found in ROA (0.6%), Japan and China by 0.3%. The welfare losses in Thailand (US$ 9,935 million) and R-ASEAN (US$ 2,912 million), whereas increases show in China (US$ 997 million), Malaysia (US$ 402 million) and Indonesia (US$ 552 million) respectively. Despite the GDP decrease in Korea, the welfare has improved by US$ 3,239 million. The ROW also enjoys a substantial welfare improvement for US$ 13,909 million. The disaster in Thai could result decomposition of production network not just in Asia and its consequences could also spare to other regions (Fig. 4.14).

4.4.3.1

Sectional Remarks

In the estimated flood disaster in Thailand (based on 2011 Thai Flood estimate), the simulation results show the impact and linkage respectively. It could be found that in EEQ and MCH, the highest interdependence shows in Thailand and Korea, with amplified intra- and interregional impact spared through the adjustment of production. The disaster shock in Thailand also affect the GDP in the region, however, the welfare is not correlated in Korea and R-ASEAN, implying that the rapid change of production may distort the function in the economy not necessarily in a positive way. As a great shirk occurred in the TEQ production, the supply showed a substitute mainly by Japan and Korea; Indonesia and Malaysia will increase the TEQ import to fill the gap caused by the import reduction from Thailand.

4 How Do Production Networks Affect the Resilience of Firms …

101

4.5 Conclusions 4.5.1 Impacts on Different Levels The research used a CGE model with GTAP database to simulate three potential large-scale natural disasters. The model visualized the disaster impact on the change of output, external trades, trade partner in three target sectors of Asian supply chain (Electronic products, EEQ; Transportation equipment, TEQ; Machinery products, MCH). These visualized impact caused by the assumed three large-scale natural disasters has shown the interdependence between countries, firms (sectors) with the flow of production and trade volume. CGE-GTAP simulation results show that disaster impact propagation differ among the countries and the economic sectors. The disaster impact is propagated to the trade partner countries through densely knitted supply chain network. The simulation results have clearly indicated the following. For country level, direct propagation dominates in the vertical integration among Japan, Korea and Taiwan; as to the country level, China and other ASEAN countries will have complementary effect which accelerate the production; for industrial level, the manufacturing sector is the major channel of the propagation while the service sectors show less degree of the integrity; for firm level, supply chain network may facilitate the recovery process from the supply chain disruption since the damaged factories or business points may easily get the intra-company supports in the supply chain. However, the simulation results demonstrated that the lean and complex network in Asia is serving as a devise to amplify the propagation of the disaster impacts at the aggregated level of the country-wide economy. The propagation effects can be positive or negative depending on the basic configuration of the production network.

4.5.2 Policy Implications The quantitative assessment serves as a vital and evidence-based instrument to optimize the policy response to the disaster. The natural disaster usually invokes emotional response since causality could have reduced if more extensive safety investment was implemented. A fair and reliable quantitative analysis is needed as a policy analysis tool to deliver the optimized response to the disaster. The policy implications are stated as following: (1) The disaster scenarios play a very important role in identifying potential risk and disruption on supply chain at the early stage. The simulation results also visualize the economic indicators and interdependence of the countries and sectors accounts for the entire supply chains. ASEAN countries should develop such models for disaster scenarios in order to have a more resilient industrial policy to make effective risk reduction planning in responding the large-scale

102

M. C. Huang and A. Masuda

natural disasters. The disaster preparedness for flagship industries is essential to sustain the economy. (2) The increased resilience to the disaster by the supply chain will contribute to the macroeconomic stability and welfare. The disaster impacts propagated through the supply chains act as regular noise to the national economy. Korea owns the capacity to increase the production when disaster occurred in other countries, where this does not substantially increase its welfare. Thus ASEAN countries should invest in its own capacity, such as to reinforce the building code, to diversify the manufacturing with its own subsidiaries, and to strengthen collaboration with other suppliers in the nearby region as business continuation plan. (3) The ASEAN countries are gradually playing important roles in regional supply chains. Philippines and Vietnam will soon become the next key players. Therefore, it is suggested that ASEAN could cope together for a disaster risk reduction (DRR) fund to invest more in building disaster-resilient supply chain. The reinforced capacity is expected to resume the production disruption sooner and thus to reduce the dependence from Korea and Japan, as well as the macroeconomic volatility. The industries in the ASEAN countries will also be granted opportunity to upgrade under the recovery and the steady growth. (4) The price fluctuation may have firms refrain from sudden price increase immediately after the disaster. Volume effects may be propagated through cascading quantitative adjustment in the production chain. Higher prices of the goods which is subject to the shortage may induce investment incentives on the damaged sectors. The CGE analysis confirms price adjustment in economy-wide response to the natural disaster. The ASEAN countries should encourage industries to improve the disaster awareness by implementing disaster preparedness, such as switch to other suppliers to limit the uncertainty and to avoid the supply chain disruption.

4.5.3 Research Limitations CGE models applied for the disaster impact assessment use the same IO table before and after the disaster assuming the production structure is totally unchanged even after massive destruction of the production facility. Due to the parameter given such as homogenous production and demand function, the simulation results in a static CGE model may display the ‘planned’ of ‘expected’ results despite its non-lineal function. It is an empirical question whether the IO tables needs to change after the economy experienced devastating destruction of the production facility. The model used in this research is as of the static nature. Consumption behaver is restricted only for one term in the current model scope; but in reality, it is likely to change particularly after experiencing the devastating disaster and inherited shortages of the goods. A dynamic and recursive model will enable us to estimate the response to

4 How Do Production Networks Affect the Resilience of Firms …

103

the disaster aftermath and recovery process that needs to be incorporated in consumer and corporate behaviour. If these assumptions and features could be added in the model, the simulation would be more accurate with the reality. Moreover, the IO table contained in GTAP v9.0 also has not fully updated especially with the ASEAN emerging members such as Vietnam, Philippines, Laos and Cambodia. Thus these countries are classified as R-ASEAN with Singapore. This may confine the individual overview of the ASEAN member countries.

4.6 Recommendations for Future Work 4.6.1 Other Types of Natural Disasters In addition to earthquakes and floods, the climate change has contributed to greater exposure to typhoons/cyclones/hurricane with severer strength. The Katrina Hurricane in 2005, Sandy Hurricane in 2012 and Haiyan Typhoon in 2013 have greatly threatened the capacity in mega cities and less developed area with great massive losses. We may need to construct more estimates for modelling many other kinds of natural disasters for making better assessment and policy recommendation for disaster preparedness.

4.6.2 Dynamic and Stochastic Analysis With the preliminary simulation results in the research, the development of the dynamic recursive models with stochastic approach may become indispensable to capture the disaster recovery process under a more realistic platform. The dynamic CGE model will assist policy makers to estimate how long and how much it will take to recover. It will be of help for making a recovery policy package and aid-fund raising.

4.6.3 The Analysis of Labour Mobility After a large-scale natural disaster, there may be a substantial labour immigration for recovery or contrarily, to leave for another country. We may employ the mobile labour factor to observe labour immigration problem and its impact. Further, while decomposing labour endowment into skill and unskilled labour, it may help us figure out impact of labour mobility and for redesigning immigration policy package to boost economy after the disaster in the decrease of aggregate demand.

104

M. C. Huang and A. Masuda

4.6.4 New Parameter of Growth and Innovation By using the recursive dynamic analysis, the growth path of economy with international aid in certain sectors may stimulate domestic development with the spill-over effect in trade and R&D. If we are able to implement such parameters while distinguishing the tangible and intangible capital in knowledge stock, it will enable us to further unfold the supply chain upgrade in the sophisticated ICT and IoT (Internet of Things) society. A holistic economic plan for analysing supply chain is highly expected.

Annex 1. Region Categories See Table 4.5.

Annex 2. Sector Categories See Table 4.6.

Annex 3 Sector Impact Estimates In terms of capital loss assumption, following Huang and Hosoe (2016), the geographic building collapse rate is used to estimate the capital loss. In the r-th region, the study uses information of geographic concentration rates of the i-th industry in the r-th region from the IO table in prefecture level. Combining these two datasets, the regional and industrial building collapse rates can be calculated. As the proportion (ratio) of fully- and partially-collapsed buildings was approximately 1:1, the estimated capital damage could be assumed to be twice as large as the original damage. Finally, the capital losses of the i-th industry in the r-th region CL i,r as  C L i,r = X i,r × Cr × 2 could be computed. The total sectoral capital loss rate C L i,r is shown in Tables 4.7 and 4.8 We estimated regional labour loss rates r

in the total labour force (endowment) LL r by combining the regional employment share SL r with the national labour force affected by the building damage (C r × 2) as S L r × Cr × 2. For the sectoral capital stock damage assumption of Thai Flood, we mainly used the results estimated by Isono and Kumagai (2014). For energy sector of coal, crude oil, petroleum and natural gas, we used the statistics from mineral USGS (2014) and multiplied by the share of capital in the value-add part. For labour endowment loss,

4 How Do Production Networks Affect the Resilience of Firms …

105

Table 4.5 Region categories from GTAP v.9 Abbreviation

GTAP regions

TWN

Taiwan

CHN

China

KOR

South Korea

JPN

Japan

THL

Thailand

MAY

Malaysia

IDO

Indonesia

RASEAN

Brunei, Cambodia, Laos, Philippines, Singapore, Vietnam

ROA

Hong Kong, Mongolia, Rest of Asia, Rest of Southeast Asia, Bangladesh, India, Nepal, Pakistan, Sri Lanka, Rest of South Asia

ROW

Australia, New Zealand, Rest of Oceania, Canada, USA, Mexico, Rest of North America, Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Paraguay, Peru, Uruguay, Venezuela, Rest of South America, Costa Rica, Guatemala, Honduras, Nicaragua, Panama, El Salvador, Rest of Central America, Dominican Republic, Jamaica, Puerto Rico, Trinidad and Tobago, Caribbean, Austria, Belgium, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal, Slovakia, Slovenia, Spain, Sweden, United Kingdom, Switzerland, Norway, Rest of EFTA, Albania, Bulgaria, Belarus, Croatia, Romania, Russian Federation, Ukraine, Rest of Eastern Europe, Rest of Europe, Kazakhstan, Kyrgyzstan, Rest of Former Soviet Union, Armenia, Azerbaijan, Georgia, Bahrain, Iran Islamic Republic of, Israel, Jordan, Kuwait, Oman, Qatar, Saudi Arabia, Turkey, United Arab Emirates, Rest of Western Asia, Egypt, Morocco, Tunisia, Rest of North Africa, Benin, Burkina Faso, Cameroon, Cote d’lvoire, Ghana, Guinea, Nigeria, Senegal, Togo, Rest of Western Africa, Central Africa, South Central Africa, Ethiopia, Kenya, Madagascar, Malawi, Mauritius, Mozambique, Rwanda, Tanzania, Uganda, Zambia, Zimbabwe, Rest of Eastern Africa, Botswana, Namibia, South Africa, Rest of South African Customs, Rest of the world

we grounded the article Koser (2014) and multiplied the affected labour force by labour share in the value-add part. Generally speaking, the flood impact usually lasts for months, and the factory operation may as well as be disrupted. The flood recovery and aftermath may not take much time of years as it may comparing with earthquake. However, in the static analysis in this research, a one year disruption of supply chain may have similar impact as earthquakes. And thus we run the flood simulation as we do for earthquake simulation (Table 4.9).

106

M. C. Huang and A. Masuda

Table 4.6 Sector categories from GTAP v.9 Abbreviation

Industries in GTAP classifications

AGR

Paddy rice, Wheat, Cereal grains nec, Vegetables, fruit, nuts, Oil seeds, Sugar cane, sugar beet, Plant-based fibres, Crops nec, Cattle, sheep, goats, horses, Animal products nec, Raw milk, Wool, silk-worm cocoons, Forestry, Fishing, Minerals nec

ENG

Coal, Oil, Gas, Petroleum, coal products, Gas manufacture, distribution

FOD

Meat: cattle, sheep, goats, horse, Meat products nec, Vegetable oils and fats, Dairy products, Processed rice, Sugar, Food products nec, Beverages and tobacco products

MAN

Textiles, Wearing apparel, Leather products, Wood products, Paper products, publishing

STL

Ferrous metals, Metals nec, Metal products

EEQ

Electronic products

TEQ

Motor vehicles and parts, Transportation equipment

MCH

Machinery and equipment nec

TRS

Transport nec, Air transport, Communication

SRV

Electricity, Water, Construction, Trade, Financial services nec, Insurance, Business services nec, Recreation and other services, Public Administration/Defence/Health/Education, Dwellings

Table 4.7 Tokyo earthquake (Tokyo Bay North) Region

# of collapsed buildings

Share of collapsed building in areal existing buildings (C r ) (%)

Share of labor endowment (LL r ) (%)

Tokyo Metropolis

333,000

4.9

1.9

Kanagawa Prefecture

136,000

3.3

0.4

Saitama Prefecture

97,000

3.2

0.3

Chiba Prefecture

42,000

1.5

0.1

Ibaraki Prefecture

1300

0.1

0.0

Gunma Prefecture

90

0.0

0.0

Tochigi Prefecture

80

0.0

0.0

Labor Endowment Loss Source Calculated by Authors based on Japan’s Cabinet Office

2.7

4 How Do Production Networks Affect the Resilience of Firms …

107

Table 4.8 Taipei earthquake Region

# of collapsed buildings

Share of collapsed building in areal existing buildings (C r ) (%)

Share of labor endowment (LL r ) (%)

Taipei City

23,971

12.6

2.8

New Taipei City

80,681

12.4

4.2

Taoyuan County

9,073

1.9

0.3

Hsinchu County

88

0.1

0.0

Hsinchu City

96

0.1

0.0

Keelung City

1,588

2.2

0.1

5

0.0

0.0

512

0.4

0.0

Miaoli County Iilan County Labor endowment loss

7.4

Source Huang and Hosoe (2016)

Table 4.9 Thai flood

Sector

Assumed sectoral capital stock damage (%)

Agriculture

−17.6

Coal, Oil, Gas, Petroleum

−23.4

Food manufacture

−15.6

Textile and printing

−11.1

Steel and metals

−18.0

Electronic products

−15.0

Transportation equipment

−19.8

Machinery and manufacture

−15.6

Transportation

−11.1

Service

−2.8

Labor endowment loss

−3.8

Source Isono and Kumagai (2014), USGS (2014), Koser (2014), recalculated by authors

References Adam C (2013) Coping with adversity: the macroeconomic management of natural disasters. Environ Sci Policy 27S:S99–S111 AON (2013) Annual Global Climate and Catastrophe Report. AON Benfield. http:// thoughtleadership.aonbenfield.com/Documents/20130124_if_annual_global_climate_ catastrophe_report.pdf

108

M. C. Huang and A. Masuda

Armington P (1969) A theory of demand for products distinguished by place of production. Int Monet Fund Staff Pap 16:159–178 Haraguchi M, Lall U (2015) Flood risks and impacts: a case study of Thailand’s floods in 2011 and research questions for supply chain decision making. Int J Disaster Risk Reduct 14:256–272 Hertel T (1996) Global trade analysis: modeling and applications. Cambridge University Press Hosoe N, Gasawa K, Hashimoto H (2010) Textbook of computable general equilibrium modeling: programming and simulations. Palgrave Macmillan Huang MC, Hosoe N (2016) Computable general equilibrium assessment of a compound disaster in northern Taiwan. Rev Urban Regnal Dev Stud 28(2):89–106 Isono I, Kumagai S (2014) Long-run economic impacts of Thai flooding on markets and production networks: geographical simulation analysis. Resil Recover Asian Disasters 18:155–169 King A (2012) Economy-wide impacts of industry policy. New Zealand Treasury Working Paper 12/05 Kirdruang P (2013) Impacts of the 2011 flood on the employment sector in Thailand. Thammasat Econ J 31(3):32–67 Koser K (2014) Protecting non-citizens in situations of conflict, violence and disaster. In: Martin S, Weerasinghe S, Taylor A (eds) Humanitarian crises and migration: causes, consequences and responses. Routledge, London Moriguchi C, Abe N, Inakura N (2015) Short-term impacts of the great east Japan earthquake on consumption and price: empirical analysis through high frequency data. In: Saito M (ed) Disaster and economy Rose A, Guha G (2004) Computable general equilibrium modeling of electric utility lifeline losses from earthquakes. In: Okuyama Y, Chang S (eds) Modeling spatial and economic impacts of disasters. Springer, Heidelberg, pp 119–141 Rose A, Liao S (2005) Modeling regional economic resilience to disasters: a computable general equilibrium analysis of water service disruptions. J Regnal Sci 45(1):75–112 Sue Wing I (2011) Computable general equilibrium models for the analysis of economyenvironment interactions. In: Batabyal A, Nijkamp P (eds) Research tools in natural resource and environmental economics. World Scientific, Hackensack, New Jersey, pp 255–305 Todo Y, Nakajima K, Matous P (2015) Merit and sin of the supply chain network in the process of the recovery from the natural disaster. J Regnal Sci 55(2):209–229 UNESCAP (2014) Statistical Yearbook for Asia and the Pacific 2014. UNESCAP, Bangkok. http:// www.unescap.org/sites/default/files/ESCAP-SYB2014_0.pdf UNISDR (2015) Global assessment report 2015. UNISDR, Geneva. http://www.preventionweb.net/ english/hyogo/gar/2015/en/gar-pdf/GAR2015_EN.pdf USGS (2014) The mineral industry of Thailand, U.S. Geological Survey Minerals Yearbook-2012, 25.1-25.6 Warren D, Vargas V, Loose V, Smith B, Vugrin E (2010) An input-output procedure for calculating economy-wide economic impacts in supply chains using homeland security consequence analysis tools. Presentation at North American Regional Science Council, 10–13 Nov 2010

Chapter 5

Economic Risk Characteristics of an Indonesian Palm Oil Value Chain and Identifying Sources of Uncertainty in Policy Making A. Faroby Falatehan and Budi Indra Setiawan Abstract Palm oil industry is one of the main contributors to the economic development in Indonesia. It has experienced rapid growth in recent decades, and become a significant contributor to the world market for vegetable oils. Oil palm plantations produce crude palm oil (CPO) used as raw material by downstream industries to provide many types of derivative products such as in food and oleochemicals. In consequence, managing value chains are vital to gain sustainable palm oil development considering the increasing competitiveness and needs to maintain mutual partnerships among stakeholders. This chapter analyzed possible external shocks that would give potential impacts to value chains of the palm oil industry in Indonesia and recommends what the policymakers should do to anticipate the negative implications. We found that some variables show elastic values represented by the world price of CPO, the export price of Indonesian CPO and the exchange rate to the export quantity of Indonesian CPO. Based on our simulation, the export tax policy of palm oil has caused negative impacts both to the small scale (farmers) and the big scale industries in the form of decreasing FFB price and exported quantity of CPO, respectively. Keywords External risks · Value chain · Palm oil · Tax export JEL Classification F15 · F23

This research was as a part of the project of Economic Research Institute for ASEAN and East Asia (ERIA) ‘Economic Risk Characteristics of an Indonesian Value Chain and Identifying Sources of Uncertainty in Policy Making.’ The opinions expressed in this paper are the sole responsibility of the authors and do not reflect the views of ERIA. A. F. Falatehan (B) Faculty of Economics and Management, IPB University, Kampus IPB Darmaga, Bogor 16610, Indonesia e-mail: [email protected] B. I. Setiawan Agricultural Infrastructure to the Minister of Agriculture, the Government of the Republic of Indonesia, IPB University, Bogor, Indonesia © Springer Nature Singapore Pte Ltd. 2020 V. Anbumozhi et al. (eds.), Supply Chain Resilience, https://doi.org/10.1007/978-981-15-2870-5_5

109

110

A. F. Falatehan and B. I. Setiawan

5.1 Introduction The world CPO production has grown relatively faster compared to other oil-yielding crops. In 2001–2005, the world CPO production grew on average 8.78% per year (Soeherman et al. 2006). CPO production in Indonesia though has been constantly lower than that in Malaysia. Export growth of Indonesian CPO is attributed to three factors: world demand, product and market, and competitiveness. The world demand factor reflects growth in exports that can be attributed to rising international demand, i.e., the stronger global import demand, the stronger the country’s export growth (Susila 2004). The palm oil industry is one of the main contributors to the economic development in Indonesia. It has been proven to be successful in reducing poverty as this industry has changed the lives of small-scale farmers positively. As an essential product in the Indonesian economy, palm oil is one of the most significant exported commodities with its an averaged growth of 11.92% from 2001–2015 (Fig. 5.1). Palm oil is a vegetable oil derived from the fruits of the palm tree used for both food and non-food consumption. The total global production of palm oil is estimated over 45 million tones in which Indonesia and Malaysia are the major world producers, as well as exporters with the major importers, are India, China and the European Union (Fig. 5.2). Oil palm plantations produce crude palm oil which is used as raw material by its downstream industries as derivative products such as in food industry (cooking oil and margarine, and shortening) and oleochemicals (fatty acids, fatty alcohols, and glycerin). In the country, CPO plays an essential raw material to fulfill the need of domestic downstream industries (such as palm cooking oil, oleochemical, margarine, soap, and biodiesel). Consequently, managing value chain is vital for sustainable palm oil plantations considering the increasing competitiveness and mutual partnership among stakeholders–suppliers and processors. Managing effective value chains will

Fig. 5.1 Indonesia palm oil exports by year (Statistics of Indonesia 2016)

5 Economic Risk Characteristics of an Indonesian Palm Oil Value …

111

Fig. 5.2 Principal importer of palm oil in the world 2015 (Index Mundi 2016)

support the distribution of products from the estate to intermediate production up to the world market. Nevertheless, there are many issues concerning palm oil developments in Indonesia covering over its sustainability on the aspects of economy, social and environment. To mention a view, from an economic perspective, there was a financial and economic crisis in the world. In the social aspect, there are still problematic relationships between the company and its neighboring community. Meanwhile, in the environment aspect, even though it is not allowed, there is still land/forest burning method for land preparation to open new plantations. As a strategic commodity, the role of government is essential to solving the problems including how to come up with a sound regulation system. During the economic crisis in 1997, the IDR–USD exchange rate decreased significantly, and selling CPO in the international market was more profitable than selling it domestically. At that time, CPO export increased sharply leading to a significant reduction in domestic supply. At early 1998, following the economic crisis and political turmoil in Indonesia, the Indonesian government banned CPO export and its products from December 1997 to March 1998. However, as the export ban decreased national income significantly and led to smuggling than in July 1998, the government lifted the ban and introduced an export tax of 60%. Since then the export tax of CPO has been fluctuating following world CPO price in the international market (Puteri et al. 2008). There is still an unharmonized agreement between farmers and government in determining the export tax of CPO. In one hand, the farmers wish to have a shallow level of export tax while on the other hand, the government wants to push the export tax at the highest level. The stakeholders do not easily solve this problem. Though, the government should make political decision appropriately to avoid adverse impacts on the value chain.

112

A. F. Falatehan and B. I. Setiawan

This chapter analyzed external shocks that would give potential impacts to value chains of the palm oil industry in Indonesia and recommends what the policymakers should do to anticipate the negative implications. Focused external shocks herewith are the economic, financial crisis and natural disasters that deemed have given impacts to the palm oil value chains. This chapter was mainly aimed to identify actors in supply chains and to quantify the effects of external shocks to the supply chain of Indonesian palm oil.

5.2 Palm Oil Production Oil palm is the most efficient oilseed crop. In one hectare, oil palm plantation can produce up to ten times more oil than other leading oilseed crops (Fig. 5.3). Oil palm is the most efficient tree that may achieve yields as high as eight tons of oil per hectare. Among the ten important oilseeds, oil palm is accounted for 5.5% of global land use for cultivation, but it produced 32.0% of global oils and fats output in 2012. Oil palm requires relatively low fertilizer inputs (about 1 MT of fertilizer per planted ha) with a productive cycle until 25 years. The supply of palm oil or fresh fruit bunch (FFB) in Indonesia comes from three kinds of the supplier such as (1) Smallholders (farmers); State-owned company and (3) Private company. As shown in Table 5.1, palm oil plantation grew steadily with its growth rate of 6.05% from 2010–2015 where Farmer is the highest rate. While the growth rate of CPO production from 2010 to 2015 was 6.59% where Private was the highest growth rate (7.58%). As shown in Fig. 5.4, the area of oil palm plantation increased from 1990 to 2015 its growth rate 9.90% where the most significant growth rate is Farmers which

Fig. 5.3 Oil palm efficiency versus other major oil crops (Oil World in Sime Darby 2014)

5 Economic Risk Characteristics of an Indonesian Palm Oil Value …

113

Table 5.1 Palm oil area and CPO production by category of producer (Ministry of Agriculture 2014) Year

Area (1000 ha)

CPO production (1000 ton)

Farmers

State

Private

2010

3,387

658

4,503

2011

3,782

692

2012

4,137

734

2013

4,356

2014

Total

Farmers

State

Private

Total

8,548

8,458

1,921

12,116

22,496

4,657

9,102

8,797

2,154

13,043

23,995

5,261

10,133

9,197

2,133

14,684

26,015

727

5,381

10,465

10,010

2,144

15,626

27,782

4,551

748

5,656

10,956

10,683

2,156

16,504

29,344

2015

4,739

769

5,935

11,444

11,312

2,201

17,434

30,948

Growth (%)

6.99

3.19

5.74

6.05

6.00

2.86

7.58

6.59

Fig. 5.4 Oil palm plantation in Indonesia (Statistics Indonesia 2016)

is 12.04% while Privates grew 11.30%. The total area in 2015 was 11.4 million hectares in which the largest is owned by Privates 6 million ha (51.44%), Farmers 4.7 million ha (38.19%) and State 0.77 million ha (10.36%). As shown in Fig. 5.5, production of palm oil increased significantly attaining 31 million tons in 2015 in which Privates produced the most 17.5 million tons (52.08%), Farmers 11.3 million (34.54%) and State 2.2 million tons (13.38%). From 2006–2008, the total production was relatively stable due to replanting activities both done by State and Privates. The average growth from 1990 to 2015 was 11.07% in which Farmers gained the most substantial 15.67% while for Privates 13.72%. Table 5.2 shows private and state-owned companies who are the major players in the palm oil industry in Indonesia. There are not only Indonesian companies,

Fig. 5.5 Palm oil production in Indonesia (Statistics of Indonesia 2016)

114 Table 5.2 Palm oil company in Indonesia (Ministry of Agriculture 2014)

A. F. Falatehan and B. I. Setiawan Company

Owner

Areal (ha)

Percent (%)

Sinarmas Group

Indonesian

278,400

11.45

Astra Agro Lestari

Indonesian

272,994

11.23

Indo Agri

Indonesian

230,919

9.50

Sampoerna Agro

Indonesian

114,827

4.72

Bakrie Sumatera P

Indonesian

103,288

4.25

PP London Sumatera P

Indonesian

106,407

4.38

Sime Darby

Malaysian

289,422

11.91

Guthrie Berhad

Malaysian

221,685

9.12

Kuala Lumpur Kepong

Malaysian

98,792

4.06

Tabung Haji Plantation

Malaysian

82,147

3.38

Golden Hope P

Malaysian

12,883

0.53

Wilmar Group

Singaporean

186,623

7.68

Bumitama Agri

Singaporean

113,383

4.66

PTPN IV

State-owned

136,737

5.63

PTPN V

State-owned

77,064

3.17

PTPN III

State-owned

105,290

4.33

2,430,861

100.00

Total

but also foreign companies belong to Malaysian and Singaporean. The largest area belongs to Malaysian Sime Darby (289,422 ha) located in Kalimantan and Sumatera Islands followed by Indonesian Sinarmas Group (278,400 ha) and Astra Agro Lestari (272,994) located in Sumatera and Kalimantan Islands. These two islands are indeed the main destinations of the investors to expand their plantation. The next destination now is Papua Island. Land productivity of palm oil fluctuated with planters (Fig. 5.6) with the average value in 2015 was 2.70 ton/ha, and there is an increasing trend. In general, Privates gained more productivity (2.94 ton/ha) compared to State (2.86 ton/ha) and Farmers (2.39) indicating that Privates are more intensive and efficient in managing their plantation.

5.3 External Shocks Rahmat (2016) explored the economics of the palm oil industry in Malaysia considering its positive and negative impacts of externalities. He found out that in the long term because of industrial developments and their close-link with global markets, Privates enjoyed gross margins higher than Farmers. It is evident than Privates, or

5 Economic Risk Characteristics of an Indonesian Palm Oil Value …

115

Fig. 5.6 The productivity of palm oil in Indonesia. Source Statistics Indonesia 2016

large-scale landowners are the key actors in the Malaysian palm oil industry. Meanwhile, Jakfar et al. (2015) indicated that there were 3 essential supply chains that influence sustainable development of palm oil industry in Aceh, Indonesia: (1) from bunches of palm fruit to its processing plants; (2) roles of stakeholders on determining volumes, profits and value added; and (3) factors that influence performances and competitive advantages which are land productivity, cost allocation for investment and operation, capacity of processing plants and CPO rendement rate. Natural disasters are potential shocks of externalities to supply chains in Indonesia. Setiawan and Nuryantono (2015) described the recent eruptions of Mount Sinabung in North Sumatra which caused damages of resources and infrastructures resulted in significant impacts on food productions of rice, corn, and soybean not only within Karo Regency but also in its neighboring regencies and the province. During 2009– 2014, food productions on the provincial scale decreased significantly as the harvested area decreased. Changes of cropped area and food production for rice were −6.7 and 2.9%, for corn, were −19.0 and −0.6%, and for soybean were −56.3% and −59.8%, respectively. In the Karo Regency and its neighboring regencies, annual production rates for rice were not affected due to increasing productivity but for corn and soybean decreased significantly attaining −26% and 88%, respectively. Food price for rice was relatively stable due to humanitarian supplies, but the rate for corn and soybean increased significantly. They recommended that accelerating actions to enable farmers to cultivate new farmlands is very crucial to provide the food supply for the evacuated people and reduce dependence on humanitarian aids. Furthermore, Riffin (2014) intended to analyze the effects of crude palm oil export tax on export and prices. Utilizing vector Auto-regression or vector Error Correction, it was concluded that the CPO export tax did not affect CPO export and domestic prices during the implementation of the progressive export tax. While Novindra (2011) reported that the local cost of CPO was more responsive to changes in domestic demand quantity of CPO than export demand of CPO, moreover, the development of domestic downstream industries of CPO will increase the amount of demand for

116

A. F. Falatehan and B. I. Setiawan

CPO and raise the prices received by CPO producers. Domestic policy in the form of restrictions on CPO exports along with its tax rate of 20% can improve the net welfare more than with that of the policy of national quota and the policy of export quota. On the other hand, increasing the domestic supply of CPO gives a negative impact on net welfare. Maryadi et al. (2004) conducted a study on the pricing of palm oil fresh fruit bunches managed by farmers in South Sumatra. They used PAM methodology to examine private and social profits of palm oil based on farm budget analysis. They gave recommendations on how to improve FFB prices and to increase the prosperity of Farmers. Based on calculated K-index, they found that all components of the Kindex were entirely controlled by the nucleus estates that did not act transparently. Furthermore, they stated that the floor price of the FFBs could be established relatively and benefited for farmers if the index parameters were controlled and audited regularly by a neutral agency.

5.4 Conceptual Framework In this chapter, the external shock we focused on was the natural disasters and the economic crisis. The natural disasters were referred to mount eruption, Aceh tsunami in 2004 and forest fires in 2015. The economic crisis was related to the financial crisis in Indonesia, the global financial crisis and the instability of crude oil price. Our hypotheses that these external shocks would give impacts to the palm oil value chain considering also the implementation of the moratorium on the opening of new plantation form forest and peatland as stipulated by the Presidential Instruction (INPRES No. 10/2011). Simulations were also done for the 1997/1998 economic crisis in Indonesia subjected to the implementation of CPO export tax and to the global financial crisis that has escalated crude oil prices in the world market. Furthermore, analyses of value chain were carried out in each level based on specific parameters. In the farm level, we used price and quantity of fresh fruits bunch (FFB); in the industry level, we used price and volume of exported CPO. Figure 5.7 shows the systematic relationship between the external shock and palm oil value chain. Table 5.3 shows the targets, methods, and data used for this chapter. The first target was to identify actors in the global supply chain of Indonesia palm oil and the second target was to recognize the impacts of the external shocks to the supply chain. Methods to use were supply chain analysis, literature review and mathematical equations describing multi-relationships among concerned variables such as export tax, the price of FFB, the cost of world CPO, Indonesian population, harvested area, the quantity of Indonesian FFB and exchange rate.

5 Economic Risk Characteristics of an Indonesian Palm Oil Value …

117

Palm oil industry in Indonesia

Palm oil value chain

External shocks

Economic crisis

Natural Disaster

Mount eruption and tsunami

Forest fire in Indonesia

Indonesia Financial Crisis

World Financial Crisis

Decreasing of GDP

Moratorium Palm Oil Area

Increasing Export Tax of

Increasing of oil crude price

Export Price of CPO

Price of FFB at Farm

Quantity of FFB in

Export quantity of Indonesian

Recommendation to Indonesian palm oil industry Fig. 5.7 Conceptual framework Table 5.3 The relation between objectives and analysis No.

Targets

Methods

Data

1

Identified actors in the global supply chain of Indonesian palm oil

Supply chain analysis Literature review

Primary and secondary data

2

Recognized impact of external shock to supply chain of Indonesian palm oil

The simultaneous equation of 2SLS

Palm oil export tax, the price of FFB, the price of world CPO from 1990–2015, Population of Indonesia, harvested area, the quantity of Indonesian FFB, Exchange Rate

118

A. F. Falatehan and B. I. Setiawan

5.5 Value Chains of Palm Oil 5.5.1 Value Chains and Added Values Figure 5.8 shows the value chains of Indonesia palm oil which consists of 4 main segments: (1) Upstream (2) Midstream (3) Downstream Processing (4) Consumer Products. In each segment, there are related activities and their resulted products. Products are more diversified as the value chains flow to from the Upstream to the Consumer ends. In the overall, palm oil products and its derivatives are in the form of food (80%) and non-food (20%). Such as shown in Table 5.4, referred to CPO price, the added values are larger as more derivatives are produced. Cosmetic products are

Fig. 5.8 Palm oil value chain and application (Sime Darby 2009)

Table 5.4 The value added of various derivative products of CPO (Industrial Ministry of Indonesia (2009) in Nova 2010) No.

Product

Added values (compared to CPO) (%)

1.

Olein and Stearin

20

2.

Fatty acids

50

3.

Ester

150

4.

Surfactant/Emulsifier

200

5.

Soap

300

6.

Wax

300

7.

Cosmetic (lotion, cream), talcum powder, shampoo

600

5 Economic Risk Characteristics of an Indonesian Palm Oil Value …

119

on the top of the added value attaining 600% while the added values for olein and stearin are about 20%.

5.5.2 Patterns of Supply Chains Michael Porter introduced the term value chain in his book entitled Competitive Advantage: Creating and Sustaining Superior Performance. The value chain analysis describes activities that an organization performs and links them to its competitive position. The value chain of palm oil commodity in Indonesia stretches from the farmers, processing units, industry, consumer and foreign countries. Indonesian palm oil industry is dominated by large and often vertically integrated plantation companies that manage 58% of the plantation area. Indonesian law regulates that plantation companies shall allocate 20% out of their concession shared with their neighboring Farmers. Besides, the Farmers cultivate about 40% independently and cooperating with Privates. A smallholding farmer does not require a business license if their land is not more than 25 ha. In general, this farmer has limited access to capital and manages the land by his own family. Farmers with the much smaller area are dominant producing a limited quantity of fresh fruits (Alliance 2016). Falatehan (2015a, b), such as shown in Fig. 5.9, indicated the pattern of the value chain of Indonesia CPO in Pelalawan, Sanggau, and Landak of Riau Province, and characterized the actors that play roles in the value chain from the source of FFB and processing (Fig. 5.10).

5.5.3 Source of FFB In general, FFB is produced by independent farmers, plasma and plantation company who do not have a processing unit. Independent farmer sells FFB through local agent or trader. Glenday et al. (2015) classified the farmers into two models: (1) Partnership models divided into: a. Farmer-managed cooperatives b. Individual partnership scheme (company plasma model) c. Company-managed, smallholder farmer-owned, plasma plantations. (2) Independent models divided into: a. Small-scale independent farmers (generally ~ 2–5 ha plantations) b. Larger scale independent farmers (usually ~ 10+ ha plantations). Mill company plantation which is independently managed with no investment in midstream to downstream processing is highly reliant on business relationships with midstream mills to ensure its FFB be purchased. A plantation company commonly

120

A. F. Falatehan and B. I. Setiawan

Smallholdin g Farmers

Large Farmers

Local Agent

Smallholder Farmers

Plasma Farmers

Plasma Farmers (Land lease)

Cooperative Plantation

Mill Company Plantation

Company Managed SHF Plantation

Local Trader Broker

Other Third Party Company Plantation

Mill

Refinary

Processor

Optional mid & downstream integration

Consumer

Export

Fig. 5.9 The pattern of the supply chain of Indonesia CPO. Adapted from Glenday et al. 2015 Payment Payment Village Seller

Farmer

Delivery Order from Supplier/Agent

Supplier

Payment Acceptance Letter of FFB from Factory

Factory

Fig. 5.10 Payment model of palm oil marketing in farmer level (Falatehan 2015a, b)

5 Economic Risk Characteristics of an Indonesian Palm Oil Value …

121

manages 25–100 ha. If the company manages the area more than 1000 ha then under Ministry of Agriculture Regulation No. 98/2013, it requires to set up an integrated plantation up to the milling units (Glenday et al. 2015). Oil palm fruit pricing is a critical issue. There have been many complaints about the price range of FFB. Until now, there is no settlement on this issue between the government, farmers and the companies. FFB prices offered to any farmers differ depending on the quality of FFB, prices set by individual collectors and governmental programs. It is also common that input prices to the value chain vary occasionally. FFB price determination process is carried out every month by a price setting team consisting provincial government, forestry and plantation offices, the nucleus firms, the farmer associations, research centers, and other relevant institutions. Based on interviews with each member, it was revealed that the nucleus firms play dominant roles in the determination of FFB. It is because they are the parties who have better knowledge of the supply and value chains. Ministry of Agriculture (Reg No: 395/Kpts/OT.140/11/2005) uses the following formula to determine the buying price of FFB: HTBS = K(HCPO × RCPO ) + (HIS × RIS ) where, HTBS is FFB price to be accepted by farmers; K is the proportion index to be obtained by the farmers (%); HCP is the average price of CPO (Rp/Kg); RCPO is the conversion of FFB into CPO (%); HIS is weighted prices of palm kernel oil (Rp/Kg); and RIS is the conversion FFB into palm kernel oil (%). K index is determined by considering several components, such as price of crude palm oil (FOB price), tax costs, marketing costs, transportation costs to the exporting board, conversion, FFB average price, percentage of selling volume, processing costs, depreciation, FFB value at processor weighting scale, administration costs, and TBS value at manufacturer.

5.5.4 Processing Units In the midstream segment of the supply chain, the extraction mill processes Oil Palm Fruit Bunches (OPFB) into CPO, palm kernel oil, shell (fibers), bunch ash, and sludge. Some of the extraction mills export their products to buyers aboard. There are some models of the plantation mills business as follows:

122

A. F. Falatehan and B. I. Setiawan

(1) Independent Mills Independent mills involve investment at any single point within the supply chains. Ministry of Agriculture (Reg 98/2013) requires mills to absorb FFB at least 20% from their own linked plantations. Though there are some exceptions if the land availability is limited, or there are already numerous independent mills. It is also common that independent mills do not have a link with the upstream chains and are heavily reliant on the third parties to provide FFB to be processed. (2) Partly Integrated Most companies engaged in palm oil industry invested in this model. It is often referred to as Grower model since the primary focus of business operations is on the upstream or plantation side. The companies own mills with sufficient capacity to process FFB harvested from their lands and any associated companies and smallholder farmers. They are independent operators though at any particular times receive additional FFB from the third parties and smallholder farmers when they have spared milling capacity. They are reliant on the third parties in the downstream refineries and processors to off-take their CPO and CPKO (Glenday et al. 2015). Companies following this model hold various sizes of lands are commonly ranging from 1000 to 30,000 ha. Some operators even have the lands over 400,000 ha among others are Bakrie Plantation, Lonsum and Austindo Nusantara Jaya. (3) Fully Integrated Larger-scale, multi-national company groups often follow this model. This model involves the integration of companies from upstream plantations to midstream mills and downstream refining. Under this model, the companies generally also produce branded consumer or industrial products ready to be delivered to supermarkets or industrial users. These companies usually have plantation land banks over 100,000 ha in multiple large sites spreading across Indonesia. This model has a relatively equal capacity throughout their supply chains and has limited reliance on the third parties for FFB and CPO/CPKO supply. Trading operations executing both physical and financial trades are also integrated into their systems. Research and development facilities may also be part of these larger agroindustry business models (Glenday et al. 2015). These companies usually have high added values as they involve full value chains among others are Wilmar, SMART, Salim, Sime Derby, and Musim Mas.

5.5.5 Palm Oil Value Chains In the theory of supply chain, performance measures and the role of each actor are determined by four things: inventory, transportation, facilities, and information. Performance inventories at traders in the village depend on the size of FFB production managed through a group of farmers who are suppliers. It is scalable from the volume

5 Economic Risk Characteristics of an Indonesian Palm Oil Value …

123

of supply and continuity of production. Traders in the village have transportation systems ranging from vehicle yard and trucks driven to the mills. Facilities owned by village traders in the form of warehouse and devices such as scales and tools to FFB determine FFB ripeness. In general, the traders in the village as the first chain in the supply system do not have sufficient access to information on FFB conditions from the factory but are engaged by contracts to ensure FFB supply. Thus, traders in the village have constraints to know about the volume and price-setting of FFB from the mills (Jakfar et al. 2015). On the contrary, large-scale farmers regularly have direct access to know the size and prices FFB received by the factory. Large-scale farmers have supply facilities and active transport systems, and consequently, their performances are always better than the small-scale farmers and village traders. Figure 5.9 shows the global value chain of palm oil model. Smallholder, large and plasma farmers produce and sell Fresh Fruit Bunches (FFB). They sell FFB to the mill through a local agent, broker, cooperative and mill company plantation. The mill extracts FFB to take out the palm oil. Then, the palm oil is processed by a refinery to become Crude Palm Oil (CPO) and Crude Palm Kernel Oil (CPKO). Processors will be responsible for a further step to process those products into food (margarine and cooking oil) and energy (biodiesel). Globalization has increased competitive pressure and changing market opportunities for smallholders in developing countries. To strengthen their business, Indonesian smallholders tend to make cooperative. It is essential for smallholders to do upgrading in the agricultural sector to encounter competitive pressure (Humphrey and Schmitz 2000). In Fig. 5.9, there are seven models of palm oil supply chain from FFB supplier to the mills: (1) Smallholders (farmers) → local agent → broker → mill The smallholders sell their FFB to the local agent. Ministry of Agriculture Indonesia classifies that the small farmers are the ones who own the planting area of 2–5 ha. After collected FFB from farmers, the local agent sells it to the mill. (2) Smallholders → cooperative plantation → mill Smallholders are united in a plantation cooperative and sell their FFB to the collective which will they sell the FFB to the mill. (3) Large farmers → local trader → mill Large farmers own more than 10 ha, has significant capital and sell their FFB to the local trader who has contracts with and delivers the FFB to the mill. (4) Plasma farmers → mill company plantation → mill Plasma farmers generally have contracts with and deliver their FFB to the mills or plantation companies. (5) Plasma farmers (Land lease), → company, managed SHF plantation → mill A company leases plasma plantations belong to the smallholders who will receive money in advance from the company. (6) Other third-party company plantation → mill

124

A. F. Falatehan and B. I. Setiawan

A company who does not own mills sell their FFB to the mills directly. The company has a large stock of FFB but are still not enough to fulfill minimum quantity for a mill to be built. So instead, they deliver their FFB to other companies who already have mills. The closer delivery distance of goods and products between producer and seller, the more effective channel of distribution that would have following advantages: the producers could pay close attention with customers and be aware of their’ thoughts and ideas that were shared with them, and no intermediaries that would reduce the profit of a company resulting in potential loss while the delivery time be shortened without any obstacles form the middleman etc.1 From this point of view, Model 6 above is the most effective channel of distribution because there are only two distribution channels which are the third party plantation company and the mill. Herewith, FFB supplier becomes the plantation company. Naturally, the plantation company and the mill would have an agreement on the minimum amount of FFB to be supplied, the pricing and the source of FFB. As shown in Fig. 5.10, the farmers make payment through the agent or village seller through their FFB was delivered directly to the mills because the agent of the village seller has had an agreement with the mills. The farmers rarely have a deal with the mills but not for the big plantations. In general, the payment from the agent to the farmers is made the same day of delivery; meanwhile, the payment from the mills to the agent is made about two weeks later.

5.6 External Shocks and Their Impacts 5.6.1 Aceh Tsunami 2004 The 2004 Indian Ocean earthquake with a magnitude of 9.2 Richter triggered a tsunami in Aceh, Sumatra Island and its vicinities on 26 December 2004 killed more than 68,000. World Bank reported the tsunami lowered national GDP rate by 0.1–0.4 points in 2005. Sumatra island homes the largest area of oil palm plantation and CPO production in Indonesia. The first large-scale plantations were established around 1911 in the provinces of Aceh and North Sumatra. Palm oil sector in Aceh developed slower than those in the rest of Indonesia though it has high rainfall and fertile soils that offer ideal growing conditions. There were 214,847 ha of oil palm plantations in 1999 and expanding only to 257,970 ha in 2006. CPO production concentrates in northern and western coasts of North Aceh, East Aceh, Aceh Tamiang, Nagan Raya and Singkil districts. About half of the oil palm plantations are smallholders (86,065 ha). The tsunami devastated about 28,000 ha palm oil plantations located in the coastal area. Up to this time, the government succeeded to rehabilitate around 50% of the devastated plantation area. 1 https://en.wikipedia.org/wiki/Marketing_channel.

5 Economic Risk Characteristics of an Indonesian Palm Oil Value …

125

5.6.2 Forest Fires An annual forest fire is another disaster that affected the development of oil palm plantations. World Bank estimated that Indonesia loosed 16 Billion USD because of the forest fires occurred in 2015 that also sent haze to South-eastern Asia, and that more than double of the sum needs to rebuild the damages caused in Aceh alone. The economic losses due to haze varied including the disruption of air, land and sea transportations. These financial losses and infrastructure damages caused severe negative impacts on the economic growth rate in other affected provinces and somewhat delayed government’s efforts to reduce poverty in the hardest-hit regions of Riau, South Sumatera and Central Kalimantan. Indonesian climate is strongly influenced by El Niño Southern Oscillation (ENSO). El Niño has made the dry season longer with the annual rainfall can be a half less than in reasonable condition. Severe El Niño events have long been associated with land and forest fires such as happening in 1972–73, 1982–83, 1987, 1991–92, 1994, 1997–98, 2002 and 2006. In 1997–1998, forest fires devastated 9.7 million hectares causing economic losses between 8.8 and 9.7 Billion USD and the occurring haze covered Singapore, Malaysia and the Southern parts of Thailand for an extended period. To prevent forest fires from occurring, the government issued a Presidential Instruction (Inpres No. 10/2011) on 20 May 2011 announcing a moratorium on opening or granting new areas from forest and peatland for two years ahead. During this period, it was expected to come up with better planning and regulating system on forest governance covering the enhancements of institutional capacity, coordination processes, data collection and regulation improvements (Murdiyarso et al. 2011). The moratorium was extended for another two years started from May 2013, the moratorium was continued for an additional two years to facilitate reductions on illegal deforestation and greenhouse gas emissions, and to strengthen the forest governance (Murdiyarso et al. 2011). The forest fire is not easy to stop and in fact, in the event of extreme climate in 2015 forest fires outraged some parts of Sumatra and Kalimantan islands and delivered heavy haze to the neighboring countries. Though not all flames are coming from peatlands, the President issued another instruction on 24 October 2015 to prohibit any development on unutilized peatlands including those yet uncultivated given to the investors and to start restoration programs on degraded peatlands (Mongabay Haze Beat 2015, November 13th).

5.6.3 Asian Financial Crisis 1997/1998 Indonesia experienced a financial crisis in 1997/98 as also happening in Asian countries. Fast capital flights out of the state led to significant IDR depreciation against the US dollar followed by a banking crisis (high-interest rate) and ended up to the economic crisis. During this period, the government increased from 4 to 10 of CPO

126

A. F. Falatehan and B. I. Setiawan

products and derivatives to be taxed. Besides, from January to April 1998 the government banned the exportation of palm oil products because of its scarcity in the local market. The magnitude of export tax ranged between 4.8 and 378 USD per ton CPO. The export tax even reached 60% when the check price was 630 USD per ton CPO in September 1998. In this study, the proxy of this external shock is a tax tariff for export from 1994 to 1999.

5.6.4 Global Financial Crisis 2008–2015 This global financial crisis is considered as the worst ever since the Great Depression in the 1930s. Many large financial institutions collapsed, and the government made any efforts to ease amounted tensions among others by introducing a bailout program to affected banks but not to avail since the stock markets worldwide continued to drop. Many industries suffered resulting in evictions, foreclosures, and prolonged unemployment. The crisis failed key businesses causing the declining of consumer wealth amounted to trillions UDS and down-turning economic activity leading to the global recession from 2008 to 2012 and contributing to the European sovereign-debt crisis. Up to 2008, some countries in Southeast Asia (Thailand, Malaysia, Singapore, Philippines, and Indonesia) showed resilience towards the crisis however in the first quarter of 2009, the economic performance of these countries except Indonesia and the Philippines were deteriorating. Both countries managed to keep economic growth although at declining rates. In the first quarter of 2009, Indonesia achieved 6.2% growth and in the last quarter decreased to 5.2%. This external shock plunged the CPO price to the lowest 60 USD during 2006–2014.

5.7 Impacts of External Shocks 5.7.1 Model Formulation To analyze the impacts of external shocks, we formulated econometric models for the Indonesian palm oil industry. The models were in the form of simultaneous dynamic equations consisting of 7 behavioral equations: (1) Indonesian Production of FFB (QIFB); (2) Indonesian Domestic Consumption (DCII); (3) Indonesian Export Price of CPO (PXC); (4) Indonesia Price of FFB (PIFB); (5) Indonesian Export Quantity of CPO (XIC); (6) World Price of CPO (PWC); and (7) World Export of CPO (WEC). Figure 5.11 shows the scheme of simultaneous models. Model identification was performed using the order of condition criteria and was estimated using 2SLS method validated to a time series data from 1990 to 2015. We used 17 variables such as shown in Table 5.5. Relationships among the variables are stated in the following equations:

5 Economic Risk Characteristics of an Indonesian Palm Oil Value …

ERI

QIFB

PIFB

DCII POP

AIF

PXC

GD

PIC

PRB

XIC

PIC

PWC

PWCO

127

PWOI

TA

WEC

Note: = exogenous variable = endogenous variable Fig. 5.11 The scheme of simultaneous models Table 5.5 Variables and their descriptions No.

Notation

Description

No.

Notation

Description

1

QIFB

The quantity of Indonesian FFB

10

PWC

Price of World CPO (USD/tones)

2

PWCO

Price of World Coconut Oil (USD/tones)

11

AIFB

Area Harvest (1000 ha)

3

WEC

World Export of CPO

12

PWOIL

Price of World Crude Oil (USD/barrel)

4

ERI

Exchange Rate (USD–IDR)

13

GDP

GDP per capita Indonesia (IDR)

5

DCII

Domestic Consumption (1000 MT)

14

PXC

Price of CPO Export Indonesia (USD/tones)

6

PIFB

Price of Fresh Fruit Bunch (IDR)

15

PIFB

The quantity of Fresh Fruit Bunch (tones)

7

XIC

Export quantity of Indonesian CPO (MT)

16

PRB

Price of Rubber

8

PICO

Price of Indonesia Cooking Oil (IDR)

17

TAX

Export Tax of CPO

9

POP

Population of Indonesia

128

A. F. Falatehan and B. I. Setiawan

1. QIFB = A0 + A1 ∗ AIFB + A2 ∗ PIFB + A3 ∗ DCII + A4 ∗ PXC + A5 ∗ XIC + A6 ∗ PICO + A7 ∗ PRB A1 , A2 , A3 , A4 , A5 > 0 > A6 , A7 < 0 2. DCII = B0 + B1 ∗ POP + B2 ∗ GDP + B3 ∗ PICO + B4 ∗ PXC B1 , B2 , B3 > 0; B4 < 0 3. PXC = C0 + C1 ∗ PWC + C2 ∗ TAX + C3 ∗ XIC C1 > 0; C2 , C3 < 0 4. PIFB = D0 + D1 ∗ PWC + D2 ∗ ERI + D3 ∗ TAX + D4 ∗ DCII D1 , D3 > 0; D3 < 0 5. XIC = E0 + E1 ∗ PXC + E2 ∗ ERI + E3 ∗ TAX E1 , E2 > 0; E3 < 0 6. PWC = F0 + F1 ∗ PWOIL + F2 ∗ PWCO + F3 ∗ WEC F1 , F2 > 0; F3 < 0 Table 5.6 shows a matrix that relates variable inputs and outputs as previously stated in the equations above. To solve the equations above, we used SAS program to find model coefficients of Ai , Bi , Ci , Di , Ei and Fi subjected to minimize Root Mean Square Error (RMSE), and evaluated the precision using the resulted determination coefficient (R2 ) and standard error, and observed the sensitivity of variable inputs using elasticity test. RMSE can measure how far the values of endogenous variables estimation results that deviate from the actual values. We also used Theil’s Inequality Coefficient (U-Theil) to evaluate the ability of the models for historical analysis. Tables 5.7, 5.8 and 5.9 showed the result of model constants, standard error, and elasticity, respectively. All signs of the constants are consistent with those stated in the models above. As shown in Table 5.7, the resulted R2 are sufficiently large (>0.7) except that for XIC (0.49). The produced F values for all models have probability values higher than the level of α = 0.01. RMSE for all variable inputs are relatively low (