Breaking-in and moving-up: New industrial challenges for the bottom billion and the middle-income countries

Over the past 30 years industry has expanded rapidly in many developing countries, driven by the explosive growth of manufacturing trade. Yet, a substantial part of the developing world remains at risk of failing to establish a vibrant, competitive industrial economy. During 2008, researchers at CSAE have been working on the United Nations Industrial Development Organisation (UNIDO) Industrial Development Report (IDR). This report is about those who have been left behind by industrial development and discusses the opportunities and constraints faced by two groups of countries: the countries of the ‘bottom billion’ trying to break into global markets in manufactured goods; and those middle-income countries that are striving to move up to more sophisticated manufacturing. The report further highlights the unprecedented changes in the global economy redefining industrial development, opening some avenues and closing off others. The analytical focus is on three aspects of structural change in industry: as industrialization proceeds, what goods are produced, where is the industry located and where is its output sold?

Industrialization: A driver for growth and development
Manufacturing exports are likely to offer more scope for long-term productivity growth than either agriculture or natural resources. A basic physical difference between agriculture and manufacturing limits agriculture’s scope for scale economies: land is an essential input for agriculture but not for manufacturing. For a growing number of countries in the ‘bottom billion’ the main alternative to agricultural exports is natural resource extraction. However, this has proved a highly problematic route to development. An increase in the price of commodity exports triggers a brief phase of output growth, but this is usually followed by a long period of decline with output ending up below its initial level. Although the resource extraction sector itself generates income, it often undermines the rest of the economy. Of the various export possibilities, manufacturing, where it is feasible, appears to offer the surest route to development.

Scarcely any countries have grown without industrializing. Rapidly growing economies have rapidly growing manufacturing sectors, and given the significance of trade in manufactures, the direction of change most likely runs from manufacturing growth to economy-wide growth, not the other way. This is because structural change – the shift of resources from low-productivity activities to higher productivity sectors – is a key driver of economic growth. Industry is most often the leading high-productivity sector. UNIDO’s research shows that differences in levels of total factor productivity are the dominant factors accounting for the difference in levels of development; growth has lagged where countries have failed in shifting capital and labour from low- to high-productivity sectors. In short: what economies make matters for growth.

Ordinary people depend for their wellbeing primarily upon two economic pillars: access to waged jobs and access to public social services. As labour-intensive manufacturing-based development proceeds it creates jobs and, because much of manufacturing benefits from scale economies, the expansion of manufacturing jobs can be spectacular. Ordinary people benefit both through opportunities for formal wage employment and through rising wages. But there is no automatic link from successful manufacturing development to improvements in public services. Because manufacturing is highly competitive and increasingly footloose (shifting the location of production with relative ease), the scope for generating large tax revenues from corporate profits is becoming increasingly limited. Hence, governments that adopt strategies for industrialization through manufacturing will need complementary strategies to finance public services, such as broad-based tax systems.

Products, tasks and natural resources
The report provides new evidence on the relationship between industrial sophistication, structural change and growth. Our results confirm that diversifying and moving up the product sophistication ladder in industry are important drivers of development. Fast-growing, low-income countries both diversified their manufacturing base and raised their level of product sophistication. Fast-growing, middle-income countries shifted strongly in the direction of more sophisticated products, exiting low sophistication sectors and entering higher sophistication ones. In contrast, slow-growing, low- and middle-income countries moved in the opposite direction; production intensities narrowed toward the mid-range of product sophistication. In addition, slow-growing, middle-income countries experienced a decline in the intensity of high-sophistication manufacturing.

Why should industrial diversity and sophistication matter for development? One reason is that more diverse economies may be better able to take advantage of opportunities in global markets as they emerge. Industrial diversification leads export diversification, as economies build industrial competence in new activities and enter global markets. Another reason may be that diverse industrial structures facilitate the growth of globally competitive firms in an economy.

In some manufacturing activities a production process can be decomposed into a series of steps, or tasks. Many countries may be manufacturing the same product, but each working on a different step in the process. For countries that have failed to industrialize, task-based production and trade is a potential lifeline: it is considerably more feasible to specialize in a single task than in the entire range of tasks needed to produce a product. A common concern with trade in tasks is that it may reinforce poor countries’ specialization in unsophisticated industrial processes. We tested whether task-based production is less sophisticated than other developing country manufacturing activities by comparing the sophistication of countries’ total manufacturing production with the sophistication of their task-based production. There is no indication that task trade contributes to greater specialization by poor countries in less sophisticated activities. It is possible for low- and middle-income countries to move up the sophistication gradient in tasks, just as in products.

The current commodity boom is a huge opportunity for countries that possess valuable natural resources. Many of these countries are currently very poor and the revenues from the boom are their best chance for transformative development. However, in the past such opportunities have often not been taken. Resource-rich economies in general have had little success in converting resource-based rents into productive assets: after two decades the typical resource-extracting economy is actually producing less than it would have done in the absence of a commodity boom. How successful resource-rich economies are in transforming rents into productive assets depends on three important links between resource extraction and the economy: Dutch disease effects as a result of resource shocks can limit the potential of manufacturing production for growth and diversification. Furthermore, the use of natural resource revenues for infrastructure development can lead to construction booms and considerable cost rises in the construction sector, which may become a bottleneck for development, potentially frustrating the conversion of resource revenues into productive investment. Finally, downstream integration from commodity extraction is commonly a main focus of government attempts to broaden the economy from its extractive industries base; however, this is likely to fail due to the highly standardized nature of commodity outputs.

The location of manufacturing production
Manufacturing and service industries tend to cluster in concentrated geographical areas, most often cities. The reason for this is the presence of agglomeration economies that are external to the firm but internal to a group of firms concentrated in a specific geographical location. Among the advantages are access to a pool of specialized workers, quick access to supplies of inputs and access to knowledge relevant for the firm.

The econometric literature on high- and middle-income countries provides persuasive evidence of the existence of agglomeration economies, but we are less sure of what and how much of this evidence carries over to lower income settings. One econometric study – of Ethiopia – and ten surveys of dynamic industrial locations were commissioned for this report to start to fill that knowledge void. They show that industrial agglomerations are also important for developing countries. Productivity is higher if manufacturing firms cluster together.

The growing role for exports
World markets are changing the opportunities for industrialization in low- and middle-income countries: opening some paths to industrialization and closing off others. Exports of manufactures by developing countries reached nearly US$2.5 trillion in 2005, up from US$1.4 trillion in 2000. The evolving pattern of global trade in manufactures reflects three important trends. First, despite the commodity boom, trade in manufactures continues to grow much more rapidly than manufacturing output; second, developing countries are capturing an increasing share of the global market for manufactured exports; and third, East Asia dominates the success story in developing country manufactured exports.

The most important structural change in manufactured exports is the significant rise in the propensity to export across regions and products. This is of course the counterpart to the much faster growth of manufactured exports than manufacturing output worldwide. There was also an important shift in global manufacturing capacity away from the OECD and Latin America toward East Asia. In addition, the ‘deindustrialization’ of sub-Saharan Africa from an already very small base revealed by these decompositions is particularly worrying.

Successful developing country exporters have taken multiple paths in terms of the sophistication of their exports. China had an export structure that as early as the mid-1970s was quite sophisticated for its level of income, and it increased in relative sophistication as per capita income grew. Korea, on the other hand, began with an export structure that was close to that predicted by its income level, but by 1995 had upgraded its export structure to a high level of sophistication. Argentina and Brazil – two middle-income countries at approximately the same level of income – have strikingly different levels of export sophistication. Argentina’s is well below its predicted value, whereas Brazil’s exceeds the norm. Among low-income countries, Bangladesh is notable for its very low level of export sophistication, reflecting its heavy concentration in exports of textiles and apparel.

In the middle-income countries greater export sophistication boosted growth: fast-growers exited traditional, low sophistication export sectors and entered more highly sophisticated ones. Slow-growers moved in the opposite direction, toward specialization in low sophistication exports. Both fast- and slow-growing low-income countries had roughly similar patterns of structural change in their exports, increasing the intensity of their exports of low sophistication manufactures. A striking difference between fast- and slow-growers was that in both low- and middle-income countries production and export structures moved in the same direction in fast-growers; in the slow-growers they did not.
Data on trade in tasks are limited. For this report we have attempted to account for the importance of intermediate goods in several types of manufacturing production and trade. By our measures the growth of trade in tasks has been impressive. In 1986–90 imported intermediates constituted 12% of total global manufacturing output and 26% of total intermediate inputs. By 1996–2000 these figures had risen to 18% and 44% respectively. We also found that exports use a substantially higher share of imported intermediate inputs than production for the domestic market: a ratio of about 2:1.

Despite the popular picture of trade in tasks as mainly rich country firms outsourcing intermediate inputs to developing country suppliers, reliance on imported intermediate inputs has grown across all regions. In fact, our analysis reveals that East Asia, not the OECD, had the highest levels and fastest growth of trade in tasks.

Implications for industrial development
New entrants to manufacturing, the ‘bottom billion’, are no longer merely competing with the high-wage OECD, as China was when it broke into the market. They are competing with China. In effect, China now plays the role previously played by ‘the North’: it has the scale economies, which make it competitive against new entrants. One prospect is that there may be no room for new entrants into global manufacturing because East Asia is firmly established and able to reap scale economies from its clusters while still having low wages.

Fortunately, there are three reasons to think that the future is less bleak than this suggests:

  • Rising Chinese costs. China is growing so rapidly that it is likely to encounter rising costs in manufacturing production. One source of rising costs will be rising real wages, either through an appreciation of the currency or through a rapid increase in nominal wage rates. Furthermore, China has only a limited number of coastal cities. As these expand, they are likely to encounter diseconomies of congestion.
  • Trade in tasks. For the countries of the ‘bottom billion’, trade in tasks is a potential lifeline. It is considerably more feasible to specialize in a single task rather than in the entire range of tasks needed to produce a product. The extremely limited industrialization of the ‘bottom billion’ to date demonstrates that establishing vertically integrated industries has not been viable.
  • Supportive policies in developed countries. There is scope for developed countries to support late-industrializers through their trade and aid policies. Even if used to best advantage these policies are not sufficiently potent to conjure up competitive advantage where none exists; but they do have the potential to push countries that are close to the threshold of competitiveness over the threshold.

Structural change – the rapid growth of manufactured exports from low-income countries, the explosion of trade in tasks, and the very rapid upgrading of the manufacturing sophistication of the fast-growing, middle-income countries – is also putting intense pressure on the slow-growing middle-income countries (MICs). The slow-growing, middle-income countries stand out for how little their production and export structures have changed in the last 30 years and for the fact that such changes as have taken place have probably retarded their growth. Rather than broadening, the production base in the slow-growing MICs has been narrowing toward a specialization in goods of intermediate sophistication, and since 1990, there has been virtually no change in export intensity. In contrast, the fast-growing, middle-income countries have both diversified their production and export base and moved up the scale in terms of product sophistication.

Can the slow-growing, middle-income countries escape the competitive pressure in the middle? Possibly, but it will not be easy. These countries have industrial competencies and industrial agglomerations that with appropriate policies can perhaps be tapped to support the growth of dynamic new export sectors. This capacity to adapt was illustrated quite well by a study of the performance of the Buenos Aires automotive cluster included in the report. When national policies and the corporate strategies of the cluster’s multinational investors shifted from serving a limited local market to export-orientation, the skills and technological capacities of firms in the cluster were already well developed, spurring rapid expansion of exports and employment in the sector.

Changes in task-based production may help as well. Time is emerging as a critical factor shaping the global distribution of trade in tasks: in industries subject to short cycle times or uncertain demand, such as fashion and consumer electronics, time is an important determinant of industrial location. With short cycle times, shorter transport times may outweigh higher wage costs, leading to ‘reverse outsourcing’, as industries locate closer to customers. Middle-income countries located close to major markets for short-cycle products may be able to use this time-wage trade-off to break into export markets in tasks that were formerly closed to them, due to their relatively high industrial wage levels.

*The UNIDO Industrial Report 2008/9 was written by CSAE’s director Paul Collier and John Page (former Chief Economist (Africa Region) at the World Bank and a Visiting Fellow at St. Antony’s College, Oxford during Michaelmas 2008), with the support of researchers at CSAE, UNIDO and a number of universities in Latin America, Africa and East Asia. At CSAE the research team included Alberto Behar, Markus Eberhardt, Adeel Malik, Bilal Siddiqi, Justin Sandefur and Francis Teal. Tony Venables (OxCarre, OxIGED) also provided considerable input. The Report will be formally launched in early 2009. This overview is adapted from the Report summary by John Page.

The measure of the sophistication of manufacturing activities is adapted from that used by Hausmann, Hwang and Rodrik to measure the sophistication of exports: Hausmann, R., J. Hwang and D. Rodrik ‘What You Export Matters.’ Journal of Economic Growth, vol. 12(1), 2007.