Industrial Policy for Development 3

number 2, 2013

Overview New challenges and emerging paradigms Industrial Policy for have turned industrialization and industrial policy into one of the most hotly debated and interesting issues of the early twenty- Development first century. Both the role of manufacturing in economic development and the contributions of industrial policy are con- ew challenges and emerging paradigms have tested. In this Policy Brief we argue that Nturned industrialization and industrial policy into one of the most hotly the economic development of emerging debated and interesting issues of the early twenty-first century. In the UNU- economies still requires a type of indus- WIDER book Pathways to Industrialization in the Twenty-First Century, published trialization that can deliver high-quality by Oxford University Press (see Box), we address a number of policy-relevant employment, that is aligned with the inter- issues in this regard. These include (i) the need for industrialization in general, national division of labour, and that would and the specific role of the manufacturing sector; (ii) the comparative experiences not lead to autarky, or a reversal of global of countries across time and space; and (iii) new policy challenges such as the gains in establishing openness in trade. emergence of global value chains, climate change and entrepreneurship. Industrial policy can make valuable contributions in this regard if the lessons of The Need for Industrialization and Industrial Policy the past and the challenges of the future Manufacturing has historically been the driver of economic growth, structural are sufficiently taken into consideration. change and catch-up. In recent decades, from 1950 to 2005, the pattern of indus- Written by Wim Naudé and trialization has closely reflected changes in global patterns of development. Over Adam Szirmai this period the share of manufacturing in gross domestic product (GDP) in the © United Nations University, 2013 rising economies of Asia has doubled, while the share has been stagnant in Latin ISBN 978-92-808-3101-6 America and sub-Saharan Africa (SSA). These global patterns of change suggest ISSN 1814-8026 that manufacturing continues to be a potential engine for growth and catch-up. Licensed under the Creative Commons Manufacturing brings with it special opportunities for reaping economies of scale, Deed “Attribution-NonCommercial- engaging in technological progress and learning, profiting from spillovers to other NoDerivs 2.5” sectors and providing job opportunities for variously skilled levels of labour. Though The views expressed in this publication market services have undoubtedly become more important in economic develop- are those of the authors and do not ment, neglect of manufacturing in theory and policy would be a serious error. necessarily reflect the views of the Unfortunately, premature deindustrialization is a potential threat to develop- United Nations University. ment, especially in Latin America and Africa. Between 1985 and 2005 several developing countries experienced premature deindustrialization. In Africa especially deindustrialization has been one of the disappointing features of the post-1990 period. The problem – missing out on the benefits of a dynamic manufacturing sector – is compounded by the fact that reindustrialization may be more difficult to achieve than starting industrialization in the first place. Hence there is a need for carefully considered industrial policies to promote manufacturing in industrially lagging countries. www.unu.edu

However one should beware of Indeed, a number of aspects of industrial bias and a disregard of the “received wisdom” of industrialization role of other sectors in economic devel- can be challenged. For instance, Asian opment. Biases against agriculture have development took off in the 1950s and been a prominent feature of post-war 1960s “largely in the dark through trial industrial policies, with very negative and error learning” and not through consequences. In recent decades there careful and artful prospective central are indications that the service sector planning. China’s famous rise as a man- may be becoming more important as a ufacturing giant involved a key role for driver of growth in developing coun- the state, which long predates the com- tries. The share of the service sectors is munist period, and its post-war devel- increasing and some countries, such as opment has involved the high costs of India, owe a part of their recent growth various “disastrous policy mistakes”. accelerations to dynamic service sectors In Indonesia, seen by many as a miracle such as software and business services. economy that confounded expectations Also the boundaries between manufac- in the 1960s when it was considered turing and services are becoming more an “economic basket case”, industrial “No country has been able to and more fluid, and the impact of the success was less due to selective and move successfully from a low- internet in enabling networked manu- sectoral policies than a “broad set of income agrarian economy to a high-income service-oriented facturing and niche production, are orthodox policies”. In Latin America economy without its govern- characterising what has been called industrial policy has been making a ment’s proactive use of indus- a “new industrial revolution”. This comeback despite the apparent short- trial policy to facilitate the strengthens the case for industrial comings of import substitution indus- economic transformation. policies that are broader in scope than trialization (ISI) policies in the 1960s However, industrial policies manufacturing alone, and requires and 1970s and the general rejection of failed in most countries in more research to deepen our under- industrial policy during the 1980s and the past. The book makes an standing of the intersectoral rela 1990s. invaluable contribution to the tionships between manufacturing, In the current debates about indus- emerging literature on rethink- technology and services in the struc- trial policy one can distinguish two ing industrial policy. It assembles tural transformation process of polar positions. The first position is the a group of experts to review industrially lagging countries. neo-liberal position, which continues to the reasons for industrial poli- focus on the shortcomings and ineffi- cies’ past successes and failures, Learning from the Comparative ciencies of the inward-looking ortho- and to draw useful lessons for Experiences of Countries dox industrial policies in the developing making industrial policy success- The bookPathways to Industrialization world in the period 1950 to 1980. This ful in the twenty-first century’s in the Twenty-First Century dissects a approach is generally critical of any multi-polar growth world. variety of comparative experiences kind of selective industrial policy. I highly recommend any student from which policy makers can learn. At the other extreme stand the neo- and practitioner interested in What these experiences suggest is that structuralists who argue for a revival of poverty reduction and sustained (i) industrialization is not an automatic industrial policies, including the option growth in developing countries process – history, policies and luck of protection for infant industries in to read this book.” matter; and (ii) that very different industrially lagging countries. The Justin Yifu Lin, Former Chief types of industrial policies are neces- neo-structuralists point to the perva- Economist, The World Bank, and sary in different contexts and different siveness of selective industrial policies Professor, Peking University times. Drawing simple “lessons” from and government interventions in the past country experiences may therefore successful Asian development experi- be difficult, as a number of contributors ences. They emphasize the disappoint- to the volume have made clear. ing experiences of Latin America and

2 Policy Brief www.wider.unu.edu

Africa in the period of structural Breaking into Global Value Chains adjustment, liberalization and Furthermore, despite the globalization deregulation. of trade, the difficulties of breaking We believe that both of these into global value chains remain daunt- positions are wrong, and that the most ing, but global value chains offer new appropriate policy with respect to opportunities for industrial policy. In industrial policy today is somewhere particular, the rise of global production between these two polar opposites. sharing and networked production has A more pragmatic, middle-ground radically changed the industrial policy position towards industrial policy instruments open to affect industrial requires that we should take into development. This is because successful account the lessons to be learned from industrial development will require recent experiences since the market countries to be competitive not in the reforms of the 1980s, but without complete production of some good, but forgetting the critical lessons of policy in the production only of a component failures in the post-war period of 1950 (“trade in tasks”). Integrating a coun- to 1980. There can be no return to try’s producers into global value chains policies of the past. Rather, policy may imply that the traditional focus of design should focus on the new industrial policy on “lumpy, complex challenges and new circumstances. industry” is no longer appropriate.

“Do not forget the critical lessons of policy failures of the post-war period”

New Challenges This is seen by some as positive since it may open up a wide range of Industrially lagging countries face a opportunities for poorer countries, number of new challenges. These which may be more able to find a niche About the Authors include the need for global financial in which to specialize rather than to Wim Naudé is Professorial reform, the competition that new be competitive along the entire pro Fellow at UNU-MERIT and at entrants to labour-intensive manu duction chain. In other words, finding the Maastricht Graduate School facturing production and exports face a comparative advantage in a “slice” of of Governance, University of from China and India, the competition the production chain may perhaps be Maastricht. He is also Dean that middle-income countries face easier than finding a comparative of the Maastricht School of from reshoring driven by accelerating advantage in the entire production Management. technological change and automation, chain and can be shaped by industrial Adam Szirmai is Professorial and the general need to avoid jobless policies. Fellow at UNU-MERIT and growth. The lack of sufficient employ- Professor of Development ment creation in manufacturing in Resource-Based Industrialization and Economics at the Maastricht countries with a youth bulge remains Climate Change Graduate School of Governance, a serious challenge for researchers One of the exciting new debates in University of Maastricht. and policy makers. industrial policy is that concerning

Industrial Policy for Development 3 www.unu.edu

resource-based industrialization. In of more sustainable energy sources) – the past, there was a general pessimism and second on utilizing the opportunities about the ability of resource-rich coun- for different patterns of industrializa- tries to industrialize and achieve struc- tion inherent in “green” or “low-carbon” tural transformation in light of the industrialization. This will require comparative advantages in primary much more research than what is avail- exports and the problems of “Dutch able at present to identify opportunities Disease”. But nowadays resource-based and risks for green industrialization; how manufacturing is shown to be a viable to generate the entrepreneurial innova- and technologically dynamic and trans- tions – both radical and incremental – formative option. that may be required to utilize or

“An important challenge is posed by climate change”

Another important new challenge – minimize these; and to understand the new in the sense of not being an issue when and how of regulation of industry when the existing industrial countries and technology better. Achieving these first engaged in industrialization – is difficult objectives may perhaps more posed by climate change. Both mitiga- importantly require a different tion of and adaptation to climate approach to industrial policy than in change will require greater global the past, requiring such policies to be coordination of industrial policies as more entrepreneur-focused, rather well as more emphasis on innovation than state-focused. within the content of industrial policies. The global asymmetries between Entrepreneurial Industrialization advanced and developing countries Unfortunately, the design of entrepre- make global policy coordination essen- neur-focused industrial policies is an tial. Pollution reduction should start in underresearched and complex chal- developing countries where environ- lenge. It is, however, a challenge that is mental efficiency is low and the greatest possibly central to the way in which gains could be realized at the lowest successful industrial policy will be cost. At least part of the funding for conducted in future. Entrepreneurial such efforts should be provided by the innovation is important for the reallo- advanced economies which have con- cation of resources from the traditional tributed so much more to the global (agricultural) sector to the modern pollution stocks in the past. manufacturing and service sectors and Both of these required shifts in the development of new activities within industrial policies will have to aim first these sectors. Recovery after the 2008 at reducing waste – both on the output global financial and economic crises side (through greater recycling) and on and the challenge of climate change the input side (through greater energy will require more, not less, entrepre- efficiency and the development and use neurial innovation. This implies indus-

4 Policy Brief www.wider.unu.edu

trial policies where the relationship be incorporated into policy. It implies between government and entrepreneurs that one-size-fits-all policies for indus- Pathways to Industrialization (the private sector) is important. trial development are unlikely to work, in the Twenty-First Century: As Ricardo Hausmann and Dani and they require more research on devel- New Challenges and Rodrik have pointed out, entrepre opment and the use of better tools for Emerging Paradigms neurial entry in developing countries measuring and studying entrepreneur- Edited by Adam Szirmai, generates information on the possible ship across various levels of development. Wim Naudé and latent comparative advantage of a coun- The focus on entrepreneurship Ludovico Alcorta try. Thus industrial policy becomes a draws our attention to important UNU-WIDER Studies in process of “self-discovery” of what the new elements in the industrial policy Development Economics economy might be good at producing. debate, namely the orientation towards 2013 Because leading/early entrants absorb learning, experimentation and self- (hardback, the costs (but not necessarily the ben- discovery. In the past, industrial policy ISBN 978-0-19-966785-7) efits) of early entry, entry itself may be was often heavily state-oriented and Oxford University Press suboptimal. Policies to promote such based on top–down planning. Today yy Timely contribution on man- self-discovery could be support for industrial policy needs to be interactive ufacturing and industrializa- innovation, including the establishment and experimental. Entrepreneurial tion, in light of the financial and promotion of national innovation effort, innovation and the knowledge crisis and the re-emergence systems; support for new firm start-ups sector fulfil pivotal roles. Hence policy of industrial policies and con- (e.g. by reducing regulations and require- should seek to create a learning envi- cerns about manufacturing ments and/or providing subsidized ronment in which capabilities are yy Accessible and clear struc- credit); and support for the integration upgraded and complemented. Policies ture, highlighting common of domestic small firms into global value also need to be more experimental, themes and helping the chains. More research is needed in these willing to quickly phase out activities reader to interpret, compare regards – for instance on establishing that turn out not to be promising, and critically analyse the national innovation systems given the while expanding support to activities issues facing industrial policy nature of firms’ positions in global value that turn out to be successful. China today chains and linking national innovation provides some encouraging examples yy Contributions and insights systems with multinational firms and in this respect. from leading scholars in the their outsourcing to indigenous firms. field These considerations imply that Technological Innovation yy Result of a joint project one should be careful in arguing for Authors such as Peter Marsh and between the United Nations industrial policy to be merely focused Chris Anderson have recently argued University and UNIDO with a global reach that reflects on an industrially lagging country’s that the world is at the start of a “New global concerns specialization based on its comparative Industrial Revolution” wherein the yy Unique focus on trends and advantage. They also imply that unlike interaction of the Internet, social media paradigms that are neglected in the past where industrial policies and new production technologies such elsewhere, as well as on as 3-D printing is making scale econo- were either focused on creation and climate change and industrial growth of state-owned firms or alterna- mies less important for production and policy, and on entrepreneur- tively consisted merely of broadly func- the tailoring of niche-market products ship and industrial policy tional policies without consideration to consumer tastes and preferences for firm or entrepreneurial specifics, more important. For instance, archi- http://www.wider.unu.edu/ the requirement now is for industrial tects in the Netherlands are planning publications/books-and- policy to be a nuanced partnership to “print” a designer house by 2014 journals/2013/en_GB/ between entrepreneurs and the state. using a 3-D printer that can manufac- Pathways-Industrialization/ The difficulty is that such industrial ture 6 × 9 meter panels. Astronauts policies will require heterogeneity on expect in the future to “print” food the country, firm and regional level to from a diverse menu when on space

Industrial Policy for Development 5 www.unu.edu

missions. In New York, MakerBot has enterprises in China, a remarkable fea- This Policy Brief is based on been building 3-D printers at ever ture over the past decade has been the Pathways to Industrialization in the decreasing costs. This “New Industrial relative decline of Greenfield FDI in Twenty-First Century: New Chal- Revolution” will make technological total gross fixed capital formation – a lenges and Emerging Paradigms innovation and human skills even more sign of the importance of domestic edited by Adam Szirmai, Wim important in industrialization and investment. The lessons are that tech- Naudé and Ludovico Alcorta re-industrialization. nology is vital for industrialization, (Oxford University Press, 2013). In the most successful industrializ- that industrial policy should aim to It is a result of the UNU- ing country of the past century, China, obtain technology both from abroad as WIDER project “Promoting there has indeed been in recent years well as from domestic sources, and that Entrepreneurial Capacity” an increased emphasis on the role of focusing on the complementarities directed by Wim Naudé and the industrial policy to foster indigenous between foreign and domestic innova- outcome of a joint UNU-MERIT, absorption capacity and technological tion is likely to be most productive. UNU-WIDER and UNIDO innovation for the development of workshop held in Maastricht Loss of Policy Space in 2009. manufacturing, and to underpin this by robust investment in supportive Finally, in all of the above, a central business infrastructure. Since 1998 new challenge in industrial policy for- there has been a veritable explosion in mation for developing countries is the international patent registrations from loss of policy space due to globalization Chinese companies: between 2004 and and the current architecture of inter 2008 more than 7,000 patents were national institutions. We have learned registered by Chinese companies at the from past experiences that there are United States Patent and Trademark few examples of successful industriali- Office; R&D expenditure as a share of zation that did not involve a phase of GDP increased in China from 0.6 protection of new industrial activities. per cent in 1998 to over 1.4 per cent by Whatever the inefficiencies of ISI poli- 2008. This technological prowess has cies, especially when continued for too given Chinese firms a lead in the pro- long, it seems on balance that they have duction of wind turbines: four of the indeed contributed to capability build- top ten producers of wind turbines ing in economies such as the Republic today are Chinese firms – Goldwind, of Korea, Taiwan, India, Indonesia and Sinovel, United Power and Mingyang. China, which enabled (existing or new) Such advances in research in China firms to compete in later more liberal have been accompanied by greater policy environments when policy investments in human capital: China turned outwards. One response to cur- rose to be amongst the top five coun- rent policy constraints is to engage in tries in the world in terms of university more regional and decentralized policy enrolment in 2007. In the past the initiatives. Another possible response is adoption of policies to attract return to revisit the notion of non-reciprocity, migration of skilled workers, to trans- which under the General Agreement on fer surplus labour from rural to urban Tariffs and Trade allowed poor coun- areas, and the policy of requiring joint tries latitude for protection, while ventures with foreign firms have played reducing barriers to trade in the global a highly significant role in making tech- economy. This might be relevant for the nology transfers from multinational poorest developing countries embark- enterprises more effective. Moreover, ing on structural change in the face of despite the huge role of foreign direct Chinese and advanced economy com- investment (FDI) by multinational petition. The notion of non-reciprocity

6 Policy Brief www.wider.unu.edu

emphasizes the important point that and on the amassed literature on the we want to avoid creating obstacles to topic. The debate should focus less on the growth of world trade – one of the whether or not there should be any obvious dangers of resurgent protec- industrial policy and more on making tionism – while creating opportunities existing policy instruments more for poor developing countries to effective and crafting new policy embark on structural change and entry instruments that take into account into global trade. entrepreneurship, level of development

“Development prospects in the early 21st century depend on job-creating industrialization”

Concluding Remarks of a country or region and the changing The development prospects of develop- relationship between state and private ing countries in the early twenty-first sector. Policy makers and academics century depend on a type of industri- need to be aware of the recent trends, alization that can deliver high quality challenges and emerging paradigms in employment, which is aligned with the the world economy and to understand international division of labour, and how these shape the crafting of new which would not take place in autarky. industrial policy instruments and We believe that industrial policy can the effective application of existing make valuable contributions to struc- instruments. While industrial policy is tural change and industrialization, if perhaps even more urgent than ever, it the lessons of the past and the chal- may also be more difficult to practically lenges of the future are sufficiently implement than before. These aspects taken into consideration. should not be overlooked in the policy Developing countries can benefit dialogue and should increasingly form from the debates on industrial policy part of the scholarly agenda.

Industrial Policy for Development 7 www.unu.edu www.wider.unu.edu

INSIDE: Policy Brief Industrial Policy for UNU World Institute for Development Economics Research (UNU- Development WIDER) is a research and training centre of the United Nations University. UNU-WIDER was established by the United Nations Industrial policy can make University (UNU) as its first research and training centre and valuable contributions to started work in Helsinki, Finland, in 1985. The Institute undertakes structural change and applied research and policy analysis on structural changes affecting industrialization, if the the developing and transitional economies, provides a forum for the advocacy of policies leading to robust, equitable and environmentally lessons of the past and sustainable growth, and promotes capacity strengthening and training in the challenges of the future the field of economic and social policy-making. Work is carried out by are sufficiently taken into staff researchers and visiting scholars in Helsinki and through networks consideration. While of collaborating scholars and institutions around the world. industrial policy is perhaps even more urgent than ever, it may also be more difficult to practically implement

than before.

www.wider.unu.edu

Finland

Helsinki

Katajanokanlaituri 6 B FIN-00160 B 6 Katajanokanlaituri

Economics Research Economics

Development for Institute World United Nations University Nations United

8 Policy Brief