The Political Economy of Automotive Industrialization in East Asia
Richard F. Doner, Gregory W. Noble, and John Ravenhill
Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries.
Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above.
You must not circulate this work in any other form and you must impose this same condition on any acquirer.
Library of Congress Cataloging-in-Publication Data
Names: Doner, Richard F., author. | Noble, Gregory W., author. | Ravenhill, John, author. Title: The political economy of automotive industrialization in East Asia / Richard F. Doner, Gregory W. Noble, and John Ravenhill. Description: New York, NY : Oxford University Press, [2021] | Includes bibliographical references and index.
Identifiers: LCCN 2020037151 (print) | LCCN 2020037152 (ebook) | ISBN 9780197520253 (hardback) | ISBN 9780197520260 (paperback) | ISBN 9780197520284 (epub)
LC record available at https://lccn.loc.gov/2020037151
LC ebook record available at https://lccn.loc.gov/2020037152
DOI: 10.1093/oso/9780197520253.001.0001
1 3 5 7 9 8 6 4 2
Paperback printed by Marquis, Canada
Hardback printed by Bridgeport National Bindery, Inc., United States of America
Preface
All three of us have been studying the auto industry in East Asia for more than three decades. We began collaborating on research on the sector when the World Bank invited us to prepare a paper for its “post-East Asian Miracle” study (published as Doner, Noble, and Ravenhill 2004). Subsequently, we collaborated on other studies of the auto industry, including an article on the impact of accession to the WTO on China’s auto industry (Noble, Ravenhill, and Doner 2005) and a background paper (by Doner and Ravenhill) for the World Bank’s study of Malaysia and the middle-income trap (Yusuf and Nabeshima 2009). The present book is the culmination of the authors’ long experience in studying the East Asian auto industries.
In the period since we began working on autos, the industry has changed dramatically. As evidence, one need look no further than under the hood of the contemporary vehicle. A quarter of a century ago, someone with a modest aptitude for mechanics could confidently replace key parts such as starter motors or carburetors. Today, most parts are hidden from sight (and easy access)—and sophisticated computer equipment is needed to perform a routine service on a vehicle.
When we began our collaboration, in the run-up to China’s accession to the WTO in 2001, autos was the industrial sector in China that commentators identified most often as vulnerable to a loss of tariff protection (Harwit 1995; Lardy 2002). Predicted losses in employment in the industry ranged from half a million to close to 5 million jobs. Today, the Chinese auto industry is the largest in the world, producing more than 24 million passenger cars (up from 700,000 at the time of China’s accession to the WTO) and close to 5 million commercial vehicles annually. It accounts for nearly 30% of worldwide vehicle production (Statista 2016).
While we continue to be fascinated by the car industry, researching it poses a number of challenges. With the ever-increasing use of electronic components in vehicles, the boundaries of the auto sector are less and less distinct. To some extent, there has always been a problem in isolating auto components in international trade. For instance, wing mirrors might be recorded by customs authorities under the more general tariff heading for glass mirrors; in a similar vein, GPS devices for autos are categorized under the rubric of “radar apparatus, radio navigational aid apparatus and radio remote
control apparatus.” The lack of specificity is often compounded by inconsistency in reporting by national customs authorities. At one point, we invested substantial time in drawing up a list of auto components and classifying them by level of complexity. When we attempted to make use of this classification, however, we found that some national customs authorities would, in one year, provide detailed breakdowns of auto component imports only, in the following year, to lump everything together under the heading of “automotive components not elsewhere specified.”
We had hoped as far as possible to make use of standardized data from international organizations. But, because of incomplete and inconsistent reporting, this was frequently not available (and Taiwan’s absence from most international organizations caused its own problems). Consequently, for some of our tables, we have had to compile data from a variety of national sources. The most useful sources varied from country to country—in some instances, national statistical yearbooks were helpful; in others, we relied very heavily on data collected by industry associations.
The analysis in this book rests on multiple research trips to our case study countries over the last three decades. On these trips, we conducted interviews with company managers, government officials, representatives of industry associations, auto research agencies and institutes, journalists specializing in the auto sector, and fellow academics. Between us, we had capacity to interview in most of the national languages—the exception was Korea, where we employed a translator. Because we were interested in answering the same research questions across all our cases, the interviews were semistructured—but often wide-ranging when respondents were eager to go beyond our initial questions. As far as possible, we triangulated the responses across various sources—whether different interviewees or documentary materials. Obtaining access to the private sector is never easy—not least where researchers are regarded with suspicion (on one visit to Hyundai Motors, for instance, we were questioned why political scientists would be interested in the auto industry and whether, because one of us is affiliated with a Japanese university, we were working for Toyota!). Persistence pays off, however.
A grant from the Australian Research Council (DP0666673, “Responding to Globalization: Firms, the State and Upgrading in the Automotive Industry on the Western Pacific Rim”) generously supported this project. This funding enabled the authors to make multiple field trips to our case study countries. It also facilitated meetings of the authors to discuss chapter drafts—in Thailand, in Berkeley, and at Stanford University. We are grateful to John Zysman, then Director of the Berkeley Roundtable on International Economics, and Don Emerson, then Director of the Walter H. Shorenstein Asia-Pacific Research
Center at Stanford University, for accommodating these meetings. The Australian Research Council funding supported a generation of PhD students in the Department of International Relations, Research School of Pacific and Asian Studies at the Australian National University—André Broome, Yang Jiang, Jikon Lai, Sang-bok Moon, Rongfang Pang, and Jeff Wilson—whose research assistance is gratefully acknowledged.
At Emory University, Elvin Ong and Andy Ratto collected data for some of the tables in the first three chapters; Katharine Tatum did a great job in formatting the first draft of the manuscript to OUP’s requirements. A publications grant from the Balsillie School of International Affairs enabled us to employ Caleb Lauer, a PhD student at the School, to provide a final update to the figures and tables, and to compile the index. We appreciate both the grant and the outstanding work that Caleb did for us. Doner is grateful to Emory University’s Department of Political Science and to the International Labour Organization for funding that supported his research. Several terms as Visiting Professor in the S. Rajaratnam School of International Studies at Nangyang Technological University in Singapore provided Ravenhill with an excellent base in Southeast Asia: he thanks the then Director of the School, Ambassador Barry Desker, for hosting him.
In Korea, Jung-Hyun Choi (최정현) was enormously helpful in setting up interviews for us and, where needed, providing simultaneous translation. Were it not for her persistence and powers of persuasion, we would have had far less success in gaining access to the seemingly impenetrable auto sector in Korea. We thank Hun Joo Park, of the KDI School of Public Policy and Management, for introducing us to Dr. Choi and for his other help on the Korean leg of this project. In Indonesia, we received substantial assistance from the late Thee Kian Wee, Ridwan Gunawan, and several former officials of the Astra Corporation. In Thailand, important insights into the evolution of the auto industry were provided by Boonharn Ou-Udomying, George Abonyi, Archanun Kopaiboon, Somsak Tambunlertchai, Achana Limpaiboon, Thavorn Chalassatien, Krisda Suchiva, Akira Suehiro, Adisak Kongwaree, Parithut Bhandubhanyong, Narong Varongkriengkrai, and Anittha Jutarosaga. Doner is especially grateful to Laurids Lauridsen for his exemplary scholarship and to Chayo Trangadisaikul, who has gone out of his way to provide an understanding of all levels of the industry in Thailand, from problem-solving on the factory floor, to business association initiatives, to ministerial debates and decisions. In Malaysia, Jean Law (MACPMA), Kevin Wai (JEBCO), Chow Siew Hon and Terence Chow (Yokohama Battery), Winston Hock (Nakagawa Rubber), Herman Leong (Tong Yong Rubber), and Paul Low (Malaysian Sheet Glass) all provided insights into the development
Preface
of the industry and its contemporary challenges. Ravenhill is also grateful to officials from the Malaysian government and to senior executives at Proton, interviewed as part of the World Bank project.
For help with the China chapter, we would like to extend our thanks to Zhu Tianbiao, then at Beijing University, now at Zhejiang University; Kamiyama Kunio of Josai University, who took the lead in organizing several research trips to China; Shahid Yusuf, then at the World Bank, now at Georgetown University, for inviting us to conduct interviews at several innovative Chinese firms, including the leading privately-owned automotive parts firm Wanxiang; and last but not least to the University of Tokyo’s Institute of Social Science for support with travel funds. In Taiwan, special thanks are due to Feng Chi-tai, then Taiwan’s Representative to Japan; Yeh Kuang-shih, former Minister of Transportation; Hwang Wen-fang, then section chief of the Taiwan Transportation Vehicle Manufacturers Association (TTVMA); and Stephen Su, Jim Chung, Justine Cheng-mei Tung, and Chen Chih Yang of the Industrial Technology Research Institute (ITRI).
Finally, we are grateful to Kaoru Natsuda and John Thoburn for their advice and scholarship on the Southeast Asian auto industries, and thank the three readers for OUP who provided incisive comments and prompted us to clarify our arguments.
1 Introduction
Contrary to the expectations of convergence found in conventional growth theories, only a handful of countries have graduated from middle-income to high-income status in the last half century. Many developing economies appear to be mired in a “middle-income trap” in which they have been increasingly squeezed between the low-wage country competitors dominating mature industries and more developed country innovators dominating industries experiencing rapid technological change (Gill, Kharas, and Bhattasali 2007: 5). For most, the absolute gap between their per capita gross domestic product (GDP) and those of industrialized economies has increased (Kharas and Gill 2019). Lagging productivity is one, if not the fundamental, source of this gap (see, for example, Agénor and Canuto 2012: 3–4). Industrial upgrading is a key to closing this gap and thus a principal challenge facing developing countries.
Industrial upgrading, which we discuss in more detail in Chapter 3, occurs when domestic firms acquire the capabilities to move into higher value-added activities at global levels of efficiency. Our emphasis is on value added not simply through increases in the capital intensity of production (Krugman 1994) but rather as achieved by domestic firms producing at global levels of efficiency while absorbing advanced foreign technology and incorporating a widening range and higher level of local skills. This upgrading involves improvements in process, product, and functions, and in some cases shifts into new, related sectors (Barrientos, Gereffi, and Rossi 2011). If such changes are to occur in more than a few exceptional firms, they must involve what Lall (2000: 24) calls a “common national environment” that helps firms to develop new capabilities rather than merely doing more of the same thing or relying on foreign companies with advanced capabilities. In this book, our focus is on whether and how upgrading has occurred in the automobile industry in seven East Asian economies: China, Indonesia, Korea, Malaysia, the Philippines, Taiwan, and Thailand.
We are especially interested in going beyond what is usually labeled “industrialization,” a concept that masks substantial differences in economic transformation. We distinguish between two stylized but analytically useful
approaches to industrialization that lead to different levels of growth, each with their own challenges. In some countries, automotive industrialization involves extensive growth, consisting largely of vehicle and components assembly and, in some cases, exports, primarily under the aegis of foreign producers operating in global value chains. Extensive growth is principally characterized by economic diversification, i.e., changes in a country’s economic structure through inter-sectoral shifts in the allocation of labor and other resources, which provide employment opportunities and generate foreign exchange earnings. But productivity gains from sectoral shifts typically do not on their own lead to increases in local value added based on inputs from national producers and on national technical capabilities (Waldner 1999; Amsden 2001; Doner 2009). Such increases are the key elements of upgrading, which for us is the essence of intensive growth.
Automotive “industrialization” in East Asia is striking for its variation in both strategy and performance (Table 1.1). As described in subsequent chapters, China, Korea and Taiwan have pursued various forms of intensive growth strategies with considerable success, whereas Thailand, by completely relying on foreign assemblers, has become a champion of extensive growth, manifesting impressive expansion of production, assembly, and exports. Although these achievements were part of a worldwide trend toward the transfer of automobile assembly to less developed economies where the major growth in global demand is occurring (discussed in Chapter 2), they were by no means inevitable. Indeed, auto production in many other developing countries, including some in East Asia, has been characterized by lags or stalls. Indonesia has achieved a degree of success at extensive auto industrialization only after decades of ineffective and expensive intensive growth efforts. In the Philippines, the site of the region’s first automotive industrialization efforts, attempts to expand assembly and parts production through an extensive strategy under foreign auspices have largely stalled. Finally, Malaysia has persistently but ineffectively attempted to emulate Korea’s strategy of
Table 1.1 Cross-National Variation in East Asian Automotive Industrialization
Performance Strategy
Extensive
Intensive
Strong Thailand S. Korea, Taiwan, China
Weak Indonesia (gradual improvement) Philippines Malaysia
Note: Assessment based on performance data provided in Chapter 2.
intensive industrialization. Explaining this variance in both strategy and performance is the focus of this book. More specifically, through comparative case studies of automotive industrialization efforts, we highlight the impact of institutional and underlying political factors largely unaddressed by existing analyses.
The automobile industry is an obvious candidate for examining national upgrading efforts. For over a century, it has occupied a central position in the global economy. It has been a major source of employment and skill development in manufacturing, and one of the most important segments of international trade. The industry presents opportunities to local supporting industries by virtue of its myriad components and its powerful backward linkages to raw and intermediate materials such as steel, plastics, and rubber, as well as to machinery, computers, sensors and a wide range of capital equipment. In 2020, for example, the share of electronic components in the total cost of vehicles was estimated to be over 35%, nearly double its share at the turn of the century (Statista 2019a). The auto industry accounts for around 9% of the total revenues of the global semiconductor industry (Kendall 2018). It also creates substantial forward linkages to service industries such as sales, service and repairs, and insurance. And, outside of housing, cars are the biggest purchases most families ever make.
Given its economic significance, the automobile industry is also politically strategic. Not surprisingly, countries have long regarded success in the car industry as a cornerstone of industrialization and a symbol of national development. When the financial crisis of 2008–2009 hammered auto sales, governments across the world poured tens of billions of dollars into propping up local producers. More recently, the move toward hybrid and electric vehicles offers developing countries the prospect of “leapfrogging” to become pioneers in technologies that may sustain the next wave of industrial evolution.
Why then do the East Asian countries differ in their development trajectories in this crucial industry? Why, for example, has the Thai auto industry evolved largely toward extensive growth and done so with great success, whereas others have moved much more belatedly (Indonesia) or less successfully (Philippines) in this direction? Why have Malaysia, Korea, China, and Taiwan aimed at upgrading through intensive development, with Malaysia faring poorly compared to the latter three countries?
Resolving these puzzles requires us to ask questions such as: What kinds of national and sectoral capabilities are needed to help domestic firms develop the competencies required to manufacture and export vehicles or complex automotive components? How do such capabilities differ from those required
where development is limited to auto assembly by foreign producers? What do local firms need if they are to develop the technological and managerial capacities to design as well as to manufacture automotive components? What kinds of difficulties are involved in generating such competencies? What kinds of organizational arrangements and institutions are required to overcome these difficulties? These questions are especially important given the failures inherent in technology markets (Lall and Teubal 1998). More specifically, to what extent do we find institutions, such as sectoral institutes, specialized training programs, and testing centers devoted to the diffusion and deepening of technologies new to domestic firms (Shapira et al. 2015)? And, most critically, under what conditions do political leaders, typically more interested in satisfying close supporters and reacting to immediate political concerns, opt to spend the time and resources required to build such institutions?
The answer we propose in this book engages major debates on development at two levels. First, our approach departs from the analysis in conventional, neoclassical economics. We argue that stable macroeconomic conditions, clear property rights, and free flows of productive factors are, on their own, far from sufficient to overcome the market failures and imperfections inherent in economic development, especially under conditions of rapid technological change. These include shortages of qualified personnel due to fear of poaching by competitors, lack of information about technologies appropriate to local firms, the need for and costs of experiential knowledge through active, handson “tinkering,” long gestation periods with uncertain returns, a lack of complementary goods and services (e.g., physical infrastructure), and shortages of financing due to risk-averse lenders and imperfect financial markets.
By downplaying the significance of such challenges, mainstream economics accounts typically overlook the coordination and collective action challenges inherent in economic development, especially in upgrading. In doing so, they ignore institutions, whose function is precisely to overcome coordination problems by helping interdependent actors, especially firms, achieve joint gains (Hausmann and Rodrik 2002). As discussed in Chapter 3, development scholars’ and practitioners’ appreciation of institutions (beyond stable property rights) has grown significantly in the past few decades, as reflected in the emergence of “good governance,” “national innovation systems,” “developmental state,” and “global value chain” frameworks. These institutionalist approaches represent significant advances in the understanding of the development process, and inform this book.
But we go beyond these frameworks. First, we emphasize and explore differences in development challenges, especially those inherent in intensive versus extensive growth. Second, we explore the institutional implications
of challenges posed by these alternative strategies. We suggest that intensive development requires different kinds of institutions than those necessary to meet the challenges of extensive growth. Finally, we argue that institutions do not create themselves spontaneously simply because, in principle, they can promote efficiency gains among interdependent actors. Instead, we explore the conditions under which political leaders opt to support the development of such institutions. In sum, we propose a political economy explanation of the heterogeneity of East Asian auto industrialization as follows:
(1) Different forms of automotive industrialization require different kinds of competencies at both the macro (national or sectoral) level and micro (firm) level. These range from financial mobilization and basic infrastructure for extensive growth, to “quality infrastructure” such as testing and standards or technical training for the kinds of firm capacities required for intensive growth. The challenges of intensive growth (upgrading) are more demanding than those of extensive growth. Our arguments would be falsified if we found intensive growth, i.e., upgrading in national automotive sectors, in the absence of such competencies.
(2) The development of various competencies in turn poses different kinds of challenges and degrees of difficulty (reflected, for instance, in the amount and type of information and the number of actors whose participation is required). Addressing these challenges is a task for institutions that facilitate different kinds and degrees of exchanges among firms, officials, and workers. The precise, optimal design (e.g., degrees of decentralization or delegation) of such institutions depends on sectoral and national contexts. For our purposes, however, the key proposition is the idea of “goodness of fit” between institutions and development challenges. More specifically, we hypothesize that extensive growth will be associated with institutions that focus largely on mobilizing capital and labor, whereas intensive growth through upgrading will feature institutions that facilitate the absorption and development of technologies new to a country’s firms.
(3) Politics—more specifically, pressures on leaders of political regimes— is key to institutional strength. Institutions with the capacities to promote upgrading will emerge to the degree that national regimes are compelled to address external threats and domestic unrest but lack easy access to resources enabling them to do so. Put more simply, tough conditions will, other things being equal, push leaders to deepen local industrialization through support of appropriate institutions despite
the vicissitudes of political and business cycles. They will be willing to make the hard choice of sacrificing current consumption to facilitate long-term development. We expect that only under such conditions will we find governments providing resources for upgrading institutions.
We introduce and assess our argument in the following steps: Chapter 2 begins by reviewing the lure and challenges of the global auto industry, and then establishes the book’s core empirical puzzle: how to explain cross-national variation in response to these conditions within East Asia. We address this puzzle through a two-pronged comparative research design in Chapters 3–9. The first, covered in Chapter 3, consists of theoretically informed, cross-case analysis. In this “most-similar-different-outcome” design, we examine variation in the strategies and performance of seven countries pursuing automotive industrialization within the same region and the same producer-driven value chain. We systematically assess the strength of a number of existing explanations (e.g., population size, open trade and investment policies, “good governance” institutions, global value chains, national innovation systems), and find that none of these constitutes an adequate explanation for performance variation. We then assess our hypothesized explanation. We first track the strength of institutions for industrial coordination, learning, and diffusion, finding that such arrangements correspond closely to variations in performance. We then move back in the causal chain by exploring the factors that account for these variations in institutional strength; we find a close correspondence between constrained natural resource endowments, external threats, and domestic political pressures (what we label “vulnerability”) on the one hand, and the strength of institutions for industrial diffusion and learning on the other.
Findings from such cross-case, “most-similar” comparisons are important for identifying variables that at first glance appear to be relevant but actually have little causal influence. For example, the fact that all seven of our countries have pursued automotive industrialization within the same global value chains and geographical regions suggests that these common factors are not relevant for explaining different outcomes. Similarly, the fact that both Indonesia and China have large populations and similar levels of per capita GDP but differ significantly in automotive industry performance suggests that population and income are far from determinant of industry performance.
On their own, such comparisons constitute an insufficient basis on which to draw strong causal conclusions because they typically require unrealistic and demanding assumptions to yield non-spurious conclusions. For our purposes, the key problem is the difficulty of assuming that we have identified
each factor that contributed causally to automotive industry performance (Bennett 2004). More specifically, it is difficult if not impossible to exclude the possibility that other factors, besides or beyond the presence of certain kinds of institutions and pressures on political leaders, account for the superior performance of, say, South Korea, relative to Malaysia.
Such weaknesses in cross-case analysis typically lead researchers to a second strategy, within-case analysis, which we pursue in Chapters 4–9. To facilitate further cross-case analysis, we cluster the cases into “intensive” and “extensive” strategy groups, thus allowing a comparison between more and less successful cases within each group. Such an analysis cannot conclusively resolve what is, in effect, a problem of causal inference. Through more indepth exploration of each national industry case, we provide a finer-grained assessment of performance and, most crucially, trace the temporal evolution of policies and especially institutions relative to such performance and the strength of our vulnerability variables.
2 The Lure and Challenges of the Automobile Industry
The automobile industry simultaneously entices and challenges developing countries. It is a leading employer and source of industrial output, a major trader, and a crucial integrator of manufacturing technologies. The industry has a wealth of backward and forward linkages. Autos also drive major financial and service sectors such as car loans, auto insurance, and networks of dealerships and repair shops. The industry holds unusual political and military significance, and serves as a marker of domestic and international prestige. Meeting the high requirements for quality and durability of automobile parts can help to raise the level of the entire manufacturing sector, including precision machining, multidimensional robots, computer-aided design and engineering, high-quality steel, injection molds, and other areas with important implications for the military-industrial complex. In this chapter we first briefly consider the significance of the automobile industry in global manufacturing and trade, discuss the challenges faced by firms in developing economies seeking to enter the industry, and then look at the evolution of production in the global industry and the specific performance of our case economies.
The automotive assembly and components industry employs more than 8 million people worldwide, more than 5% of the total global employment in manufacturing. In addition to these direct employees, the International Organization of Motor Vehicle Manufacturers estimates that about five times this number are employed indirectly in related manufacturing and service provision (http://www.oica.net/search/employment).
One reason for the attractiveness of the industry to developing economies is its role in international trade. Automotive products constituted 9% of total global trade in manufactures in 2018, trailing only office and telecom equipment, and chemicals: global exports of automotive products amounted to more than $1.5 trillion. The automotive industry also has a high positive income elasticity of demand: a 10% increase in income generates a 30% rise in demand for automobiles, a sharp contrast to most agricultural products,
for which demand typically rises by only 2% with a 10% increase in income (McConnell and Brue 2005: 368). A country lacking a substantial domestic auto industry will face a drain on its balance of payments from increasing auto imports as per capita income rises. In the past, vehicle ownership has increased dramatically when annual per capita income reached approximately $3,000–5,000 (in constant year 2000 $US) and peaked around $10,000 (Wang, Teter, and Sperling 2011: 3298). China and Thailand crossed the $5,000 threshold at the end of the first decade of this century, and by 2016 Indonesia reached nearly $4,000.
The Challenges of the Contemporary Auto Industry
Despite the prominence and allure of automobiles, the car industry presents daunting obstacles to would-be new entrants in developing countries. Barriers to entry in assembly and production of the more sophisticated parts and components are extremely high. Besides the initial costs of investment, the challenges include the need for volume production to realize economies of scale, generating bargaining leverage in a market increasingly characterized by oligopsony, and (where appropriate) establishing a positive brand image.
New assembly plants typically require initial investments in the hundreds of millions if not billions of dollars, and require coordination of inputs from a vast array of suppliers. Large components require an even more substantial investment: the cost of developing a new family of engines is estimated at between $1.3 and $2.6 billion (Automotive News 2012). And emerging technologies are even more prohibitive: Hyundai announced in 2019, for instance, that it was investing $6.7 billion in the development of its hydrogen fuel system and a further $2 billion in autonomous driving technology (Song and White 2019; Song 2019).
New entrants also face the challenge of quickly ramping up output to reach efficient levels of production. For engines and transmissions, the conventional rule-of-thumb for minimum efficient scale was typically one million units per year; for foundry, pressing, and forging, it was two million plus. These figures are of such a magnitude that only a handful of manufacturers—General Motors (GM), Ford, Toyota, Nissan, Volkswagen (VW), Renault, and Groupe Peugeot Société Anonyme (PSA)—approximated the volumes required for minimum efficient scale for foundry activities in the 1990s. Early studies suggested that the penalty that manufacturers incur if producing at only 50% of minimum efficient scale (MES) is an increase in unit cost of approximately
6%. Hasan (1997) quotes data from Waverman and Murphy (1992) that suggest that this cost penalty rises to 20% when production is only 30% of MES, and to 34.5% at one-tenth of MES. Latecomers thus face a dilemma: maximizing cost competitiveness requires expensive investment in large plants, but if they are unable to sell all of that massive output, production will remain high-cost and inefficient.
The auto industry also presents imposing challenges of quality control and coordination. Unlike steel and commodity chemicals, where much of the technology is embedded in capital equipment available for purchase even by firms in developing countries, and the key to success is efficient use of production capacity, success in the auto industry requires integrating thousands of parts into an attractive and durable design. Market pressures and legal requirements for quality, durability, safety, comfort, drivability and performance, sound insulation, fuel efficiency, emissions controls, recyclability, and product liability are high and continue to increase. In 1998, American consumers surveyed by J. D. Power reported 176 problems per 100 vehicles. By 2009 the figure had fallen to 108, despite the increasing complexity and luxury of cars (J. D. Power and Associates 2010) (subsequently, the fall has been less dramatic, in part because of problems with more sophisticated electronics: in 2018, the industry average was 93 problems per 100 vehicles [J. D. Power and Associates 2018]).
Brand names matter in the auto industry, posing another barrier to entry. It is not just “petrol heads” that associate the name Aston Martin with luxury cars or that of Volvo with safe but rather staid vehicles. Moreover, it is far easier to damage a positive image than to overcome a negative brand image. Korea’s Hyundai, for example, struggled for years to overcome the negative image that its early auto exports to North America generated. And, in the public mind, all exports from a particular country may become “tarred with the same brush” because of the problems of one manufacturer—hence the aversion to “Korean” cars in the late 1990s.
Production of vehicles and components has become increasingly consolidated in the hands of a few companies in the “global North.” Very few countries have successfully launched and sustained a domestically owned assembly industry in the seven decades since the end of the Second World War; Japan was the first to do so, followed by Korea. Malaysia enjoyed some short-term success in exporting cars assembled by domestically owned companies but, as we discuss later in this volume, it failed to sustain its national champion assembler. Finally, China’s independent companies, while still facing formidable challenges, seem to be on the verge of breaking into international markets as well as competing effectively within their domestic market, now
the world’s largest. But elsewhere around the globe, independent producers have disappeared or have been acquired by the major global players. Famous brands such as Volvo, Land Rover, MG, and Jaguar have all changed hands— some on several occasions. The liberalization of trade in auto products has enabled consumers to choose the advanced offerings of the global automakers and shun inferior local products.
As a prominent study concluded, the auto industry this century has been undergoing a “profound transition” from one historically based on exporting from home country–based exporters to a network-led industry organized around regional and global supply chains through which each major firm produces in all of the significant markets (Sturgeon and Lester 2004). This account slightly exaggerates—most global auto firms remain highly dependent on their home markets and at most one other region, and virtually none derives revenues evenly from around the world. Six of the top 10 auto components suppliers, for instance, all receive more than half of their revenue from sales in one regional market (Automotive News 2019). Nonetheless, it vividly summarizes important trends. Mergers and acquisitions have by no means uniformly succeeded, as Daimler’s failed takeover of Chrysler and many other examples in the supply industry suggest, but the broad trend toward globalization and consolidation remains unchecked. The top eight global assemblers are estimated to have a 90% share of global revenues; the top 50 components companies command more than 80% of the global market (Automotive News 2017b).
Components
Whereas consumers usually think of the auto industry in terms of assembled cars, two-thirds or more of value-added in an automobile actually resides in the myriad parts that go into it. The annual value of auto components manufactured globally in 2018 exceeded $800 billion, based on the sum of global sales of the top 100 Original Equipment Manufacturers (OEM) parts suppliers (Automotive News 2019). Comtrade data indicate that the value of exports of auto components in 2018 approached $375 billions. The spread of lean production techniques has produced a reversal of the vertical integration so doggedly pursued by Henry Ford, and in a more moderate form by Alfred Sloan at GM. In order to increase flexibility, enhance transmission of price signals and market competition, and cut labor costs, assemblers spun off many of their parts divisions, led by Delphi (GM) and Visteon (Ford), which immediately became two of the largest parts firms in the world.
As with assembly, a major consolidation has occurred in the components industry in the last two decades. According to one study, auto parts manufacturers in North America numbered 30,000 in 1990, but had shrunk to 10,000 by 2000 and to 8,000 by 2004. By 2010, there were only 5,784 firms in the industry (Bureau of Labor Statistics 2011a). From 2014 to 2017, the combined value of supplier mergers and acquisitions averaged $50–60 billion; in 2018 it jumped to a record $97.5 billion, driven in part by a desire to enhance capabilities in technology for autonomous vehicles and connectivity (Automotive News 2019).
For many major parts, such as engine control systems or seats, a handful of global companies such as Bosch and Johnson Controls supply the vast majority of global demand. In order to facilitate just-in-time production, firsttier suppliers of modules and parts typically invest in facilities located near or even inside those of the final assembler. The components industry in East Asia has been deeply affected by these trends, particularly since the financial crises of 1997–1998 and 2008–2009. The entrance of Western assemblers, intensified competition for the region’s growing markets, and shorter product cycles have pushed global automakers to expand parts manufacture in East Asia as part of increasingly integrated regional and global production networks.
Meeting the performance expectations of assemblers requires ongoing innovation by producers, another challenge faced by domestic companies in East Asia. Innovation in the auto industry is generally incremental and cumulative, based as much on tacit skills as formal research and development. Leapfrogging incumbents is much more difficult than in the many segments of the electronics industry that are periodically reshaped by radical innovations (although the move to electric vehicles may have a similar impact on the auto industry). Successful production of auto parts requires a combination of independence and tight linkages with suppliers. Some parts, such as tires, wheels, and batteries, are relatively independent of the design of the overall vehicle and from other parts, but most are tightly integrated. Auto assemblers, which might more accurately be termed system integrators, are enmeshed with component suppliers and lower-level parts producers in complex networks that balance cooperation and competition.
Several other developments within the global industry have further raised entry barriers to component producers that aspire to sell directly to assemblers. The most notable of these has been a move to common platforms (generally understood as a common underbody and suspension (with axles), although the term is often used freely in the industry). Since the mid-1990s, assemblers have increasingly produced a range of cars off the same product platform. In April 2012, for instance, VW announced that it intended to
introduce a modular platform that would be used in 40 different models worldwide, involving a total production of 40 million units (Rogers 2012). The initial investment costs for developing these platforms are very high. But once established, platforms offer assemblers an opportunity to realize substantial economies of scale and scope as parts are standardized across a number of models. Korth (2003) cites the example of a major assembler that in 1995 used 45 different trunk lock mechanisms across its various vehicles. By 1999, it had reduced this total to 12. Assemblers have returned to the idea of producing “world cars” off a single platform. Ford’s Fiesta, for example, is assembled in China, Germany, Mexico, Taiwan, Thailand, Venezuela, and Vietnam (http:// corporate.ford.com/company/operation-list.html#s1f30; accessed August 25, 2017).
For aspiring first-tier parts suppliers, the use of platforms poses several challenges. First, they have to work closely with the assemblers in the development of products that will function effectively across a range of vehicles. Second, they typically need to have a global presence to be able to supply to the various plants that are using the same platform. Third, they have to be able to produce in substantial volumes.
Related to the move to common platforms is a second trend, led by European carmakers: modularization. Instead of ordering and assembling all of the thousands of parts that go into an automobile, global automakers provide first-tier suppliers with rough specifications and then rely on the suppliers not only to produce the modules cheaply and reliably, but also to invest heavily in product innovation and to collaborate with them on the development of new vehicles. For the assemblers, this particular form of outsourcing reduces their capital requirements, forces others to share risks, and enables them to capitalize on expertise outside the firm. For parts producers, modularization can facilitate a move up the value chain and also enable them to develop technologies that can be applied across several manufacturers (realizing economies of scale and sometimes of scope). But it also involves very substantial investments. In 1996, parts firms devoted an average of 2.6% of revenues on research and development; by 2003 that figure had increased to 4.2%, nearly as high as that for the average global assembler (Misawa 2005).
Coupled with the downward price pressures that are endemic to the industry (in our interviews, we regularly heard stories of assemblers requiring at least an annual 5% reduction in price from component suppliers, and in some instances the figure was as high as 20%), the requirement for higher levels of capital outlays has encouraged a process of consolidation among first-tier suppliers similar to that among assemblers. All the largest first-tier suppliers— Robert Bosch (Germany), Denso (Japan), Magna (Canada), Continental
15 (Germany), ZF Friedrichshafen (Germany), and Aisin Seiki (Japan)—in 2018 had revenues above $30 billion (in comparison, the 2018 revenues for GM were $147 billion and for Ford $160 billion) (Automotive News 2019; General Motors 2019; Ford Motor Company 2019). Just as very few assemblers from developing economies have survived in this competitive industry, so very few parts companies from outside the industrialized world have grown to be major players. Of the top 100 global suppliers in 2019, only 18 companies came from outside the Triad of Europe, North America, and Japan (Table 2.1) (Automotive News 2019). Fourteen were from China and Korea. The smallest of these companies, Anhui Zhongding Sealing Parts, from China, had sales of $1.6 billion in 2019.
This failure to date of companies from other parts of Asia to grow into competitive global players stands in marked contrast to their success in electronics—although rapid growth by some Chinese suppliers may change this story.
Component Suppliers from Outside the Triad in the World Top 100
Source: Automotive News, 29 June 2020 <https://www.nxtbook.com/nxtbooks/crain/an8097364512SITPF_ supp/index.php#/p/SIntro> (accessed 31 October 2020)
Table 2.1
In short, domestic companies in Asia face formidable challenges if they aspire to break into the ranks of leading first-tier component producers. Although the size of the industry provides scope for entry at lower tiers of the supply chain, the risk is that companies will be operating in a relatively low-technology segment where downward pressures on prices are most formidable. We turn now to the shifting range of policy instruments available to Asian governments seeking to navigate the challenges of the contemporary industry.
Constrained Policy Instruments
What policy instruments are available to Asian governments seeking to build their domestic capabilities in the automotive industry? Here we find a rapidly changing landscape. The trade and investment policy options available to governments are increasingly, but not entirely, constrained by commitments that they have entered into in bilateral, regional, and global trade regimes.
The automotive sector historically has been among the most heavily protected in manufacturing. Governments’ desires to promote what has been regarded as a sector of strategic significance have led them to use the full panoply of protectionist instruments to attempt to develop local capabilities. The ongoing political sensitivity of the sector has had a powerful effect in shaping the emerging division of labor in the industry (Sturgeon and van Biesebroeck 2011).
Traditionally, most developing countries erected high barriers against imports of motor vehicles and components, including quotas, local content requirements, and extremely high tariffs, often exceeding 100%. They hoped that enticing investment, usually by global auto firms in local assembly operations, with promises of protection against imports, or by licensing technology from producers in advanced countries, would spur the creation of a local parts industry. Even after many rounds of trade liberalization under the World Trade Organization (WTO) and its predecessor, the General Agreement on Tariffs and Trade (GATT), regional agreements, and bilateral Preferential Trade Agreements (PTAs), autos and parts at the dawn of the 21st century attracted tariffs much higher than those in electronics or most other manufacturing sectors (Table 2.2). After acceding to the WTO in 2001 and completing a six-year transition period, China continued to levy a 25% tariff on imports of motor vehicles and 10% tariffs on auto parts. Even Thailand, famously accommodating toward foreign investment, imposed surprisingly
Table 2.2
2015 Tariffs on Assembled Vehicles and Auto Components
Assembled Vehicles* Auto Components**
Australia 5%
Canada
China
Indonesia
Japan
Korea
Malaysia
3%–6%
3%–5%
3%–6%
25% 9%–14%
27%–50%
5%–25%
0% 0%
8% 8%
15%–24%
Philippines 20%–30%
13%–30%
1%–16%
Taiwan 17.5% 0%–15%
Thailand 40%–77% 10%–30%
* Motor Vehicles listed under HS 8703 tariff heading.
** Auto Components listed under HS 8708 tariff heading.
Source: World Trade Organization Tariff Download Facility, http://tariffdata.wto.org/ (last accessed May 24, 2017). Data for Taiwan from Government of Taiwan, “Tariff Database Search,” https://portal. sw.nat.gov.tw/APGQ/GC451 (last accessed October 20, 2019).
high tariffs on autos and their parts: as of 2013, tariffs on motor vehicles varied from 40% to 77%, while tariffs on auto parts ranged from 10% to 30%.
Non-tariff barriers and explicit or implicit local-contents requirements were also used extensively to protect the auto industry. Promotional measures included preferential allocation of financing, investments in research and development, and subsidies for testing, training, and technology diffusion. Even seemingly neutral (WTO-compliant) regulations often served to promote a particular policy or industry sector. In the case of Thailand, tax policy gave a huge preference to purchasers of one-ton pickups, thus encouraging consolidation within the pickup sector. Eventually Thailand became the world’s second-largest producer and a major exporter of pickups, even while production and especially export of passenger cars at that time remained modest and heavily protected. In Japan, Korea, and Malaysia, differential rates of tax, dependent upon engine size, provided an advantage to domestic producers whose output traditionally was dominated by small cars.
Yet the trade regimes through which developing countries seek to promote their domestic auto industries have evolved, constraining the policy options that are available to them (Wade 2003). Most notable here are the WTO’s Trade-Related Investment Measures (TRIMs) Agreement, and the proliferation of bilateral and minilateral preferential agreements. The TRIMs agreement, negotiated during the Uruguay Round of GATT talks, which concluded in 1994, marked the most extensive internationally agreed curtailment of
industrial policy tools. The brevity of the agreement (six pages, including an Annex that illustrates the type of measures that are prohibited) belies its significance in outlawing many of the instruments that governments had previously used to promote local industries, including requirements for local content, trade balancing, and export performance measures. In subsequent cases that have come before WTO Dispute Settlement Panels, the agreement has been interpreted in an expansive way (and it is perhaps no coincidence that a majority of the cases considered by panels under TRIMs have involved the auto industry). Although developing economies were permitted to apply for an extension before they had to comply in full with the agreement, all TRIMs had to be removed by 2004. The Uruguay Round also saw the negotiation of the “Agreement on Subsidies and Countervailing Measures,” which further limited the scope for governments to promote domestic industry in a discriminatory manner.
This is not to suggest that it is impossible for a creative government to devise ways to pressure foreign investors to comply with certain government objectives. But it is far more difficult for governments to impose formal requirements than in the past. China was obliged to amend its Policy on Development of Automotive Industry (Order No. 8 of the National Development and Reform Commission, May 21, 2004) following a WTO Dispute Panel finding that upheld complaints from Canada, the European Union, and the United States that the policy, which applied the tariff applicable to assembled vehicles to imported components if they exceeded a certain threshold value, violated the WTO’s national treatment principle and obligations that China had accepted under its terms of accession. On the other hand, for countries with large domestic markets, various informal pressures can be brought to bear on foreign investors that have the effect of realizing the same goals sought through TRIMs—and this has certainly happened in China, where provincial governments have persuaded potential investors of the desirability of complying with specific local content targets that they have established. Even the smaller economies in the region have succeeded in carving out some “policy space” within the new WTO regime (on Malaysia and Thailand, see Natsuda and Thoburn 2014).
Most recently, the move toward bilateral and minilateral PTAs has subjected developing country governments to a new source of pressure for liberalizing their trade regimes. The heavy protection traditionally afforded the automotive sector has made it one of the principal targets in PTA negotiations. In East Asia, the main pressure on developing countries has come from the Japanese government, which has responded to calls from Japanese car makers not just to liberalize access to markets for assembled vehicles (including, for
instance, removal of “luxury” car taxes), but also to make it easier for local subsidiaries of Japanese assemblers to import raw materials and components from Japan (see, for instance, Manger 2005). In bilateral trade agreements with Thailand and Malaysia, for instance, Japanese officials succeeded in negotiating a phased removal of the duties on car parts—over eight years in Thailand, and over three years in Malaysia. Sometimes these pressures were supported by local assemblers: in Malaysia, the national car companies, Proton and Perodua, shared an interest with assemblers of Japanese vehicles in unimpeded access to imported components. The (much smaller) domestic companies supplying components resisted. Duties on imported steel were to be phased out over a longer period. Neither Thailand nor Malaysia, however, made significant concessions on imports of assembled vehicles in these bilateral agreements.
Preferential trade agreements nonetheless provide opportunities as well as posing new challenges for countries that are able to establish themselves as regional export bases. A notable example is the agreement between Thailand and Australia, which allows (subject to local value-added requirements) dutyfree access to the Australian market for vehicles assembled in Thailand and for components manufactured there. After the implementation of the agreement, Australian imports of assembled vehicles from Thailand tripled in value from $A1 billion in 2004 to $3.2 billion in 2007. Both Honda and Nissan switched their sourcing for vehicles imported into Australia from Japan to Thailand. As of 2019, thanks largely to the bilateral trade agreement, Australia remained Thailand’s single largest market for vehicles, accounting for approximately one-quarter of total exports.
The implementation of the ASEAN Free Trade Area (AFTA) also had a profound impact on the industry in Southeast Asia. Although Indonesia and especially Malaysia had sought to exempt their auto industries from the liberalization of intra-regional trade, pressure from their ASEAN partners eventually prevailed. By 2008, even Malaysia lowered its tariffs on imports from its ASEAN partners in compliance with the AFTA agreement (see discussion in Chapter 7). As with agreements with extra-regional partners, assembled vehicles continued to enjoy protection through differential excise taxes and various non-tariff barriers, but trade in components was substantially freed up.
Recent developments outside of East Asia show how contemporary trade agreements can also provide new instruments for protection. The rules of origin in the recently negotiated revised NAFTA, the US-Mexico-Canada Trade Agreement, are a classic example. These not only increased the overall valueadded requirement for vehicles to enjoy originating status (increasing from
66% for passenger vehicles in 2020 to 75% by 2023). They also added productspecific stipulations, including a requirement that by 2023 each vehicle should have a labor value content composed of at least 25% of high-wage material and manufacturing expenditures, and that at least 70% of the vehicle’s steel be produced within North America (https://usmca.com/rules-of-origin-usmca/ ). Contemporary preferential trade agreements are more about creating rents than promoting trade liberalization (Ravenhill 2017). In the most advanced of regional schemes, the European Union, the single market does dismantle border barriers, but as Markiewicz (2019) argues, deep integration also makes new developmental tools available, not least through generous financing available through the bloc’s Structural Funds.
New Opportunities for Developing Economies
If the trends toward consolidation and scale in the industry, coupled with the pressures for trade liberalization, present formidable barriers to entry for domestic companies into the global industry, the story is by no means uniformly bleak. Booming local markets for vehicles, coupled with vertical disintegration, have increased the opportunities for developing economies to insert themselves into regional and global supply chains, often securing substantial volumes of local assembly and associated component manufacturing, with positive effects on both exports and employment. In most instances, however, few local firms have participated extensively in the new domestic industries.
The vast majority of the increase in demand for autos globally is coming from developing economies, particularly in Asia, and is being met overwhelmingly from locally produced vehicles. In addition (although the overall effect is of a smaller magnitude), a substantial regionalization of production is occurring, with economies on the periphery of the main markets of Western Europe and North America emerging as significant bases to supply assembled vehicles to these markets. The consequence of these trends is a dramatic change over the first decade of the 21st century in the distribution of auto production. If we confine our focus for the moment to assembled vehicles, whereas Western Europe and North America accounted for 56% of global auto production in 1999, by 2016 their share had fallen to 24% (Japan’s share in the same period fell from 20% to 11%) (data from http://www.oica.net). In other words, by 2016, developing and emerging economies accounted for more than one-half of global car production (of which China accounted for
close to half). Whereas many industrialized economies experienced absolute declines in output of assembled vehicles, that in other parts of the world jumped substantially. The move of the industry toward developing economies is likely to continue as incomes rise: in China, under 3% of the population owns a car, in contrast to between 70% and 80% of the population in the United States (BBC News 2012).
Much of the recent success of some developing economies in the auto industry is a reflection of the fragmentation of production within the auto value chain. The relatively low value-weight ratio of components, coupled with the need for producers to be located close to assemblers and, in turn, the desire of assemblers to be in proximity to final markets, have caused “off-shoring” of production in autos to lag behind that of other sectors, most notably electronics. Nonetheless, fragmentation has afforded new opportunities. As Richard Baldwin (2014, 2016) notes, developing economies can now join existing supply chains rather than having to build them from scratch. Moreover, the transfer of know-how, capital equipment, and management have made it possible for advanced manufacturing plants to be created in developing economies in a remarkably short time period.
This potential for development through global and regional automotive supply chains contrasts with the 1960s or 1970s, when foreign direct investment (FDI) in auto production in developing countries typically led to a proliferation of inefficient, heavily protected import-substituting plants. Today, however, global assemblers and first-tier suppliers have made substantial investments, even in relatively low-income countries such as China and India. The resulting plants in these countries are capable of reaching world standards of efficiency and quality surprisingly quickly (Sutton 2004). Car assembly has become increasingly robotized, with a consequent reduction in labor inputs. The average car is estimated to require only 15–25 person hours for assembly (Zacks.com 2008).
Labor costs, including healthcare, pensions and other benefits, nonetheless, can still be a decisive factor in determining the location of assembly facilities. It is not just wages but indirect costs that have proved to be such a burden to the major manufacturers in advanced countries. According to the US Bureau of Labor Statistics (2011b), the hourly compensation for blue-collar workers in the US automotive industry in 2009 was composed of $22.19 in wages plus $14.79 in benefits plus $4.76 in healthcare—a total of close to $42 per hour. Then CEO of GM Rick Wagoner testified to Congress that healthcare costs added $1,525 to the cost of every GM vehicle produced in the United States, with the company spending more on healthcare than
