New labour-saving and energy-using technologies first originated in Britain, and then spread with a lag to continental Europe and North America. The initial impact was a Great Divergence in living standards between the West and the Rest. This divergence is now being eroded as developing economies rapidly industrialise (Allen 2009). But when did modern manufacturing first spread from the West to the Rest? Was it only when the ‘second globalisation’ began in earnest after the 1970s? Was it during the world de-globalisation which began after WW1? Or did the spread of modern manufacturing start even earlier, during the ‘first globalisation’ of the late 19th century?
In both globalisation periods, international economic integration may have helped developing countries import new technologies, exploit their lower labour costs, and import those raw materials with which they were poorly endowed. Alternatively, trade may have made it difficult for developing countries to compete with established industries in richer countries (Williamson 2011), creating an international division of labour which saw the industrial West export manufactures and import food and raw materials (Lewis 1978). In that case, the de-globalisation epoch may have favoured industrial growth in the poor periphery. So, which was it? Did globalisation favour and de-globalisation disfavour the spread of modern industry to the poor periphery, or was it the other way round?
In a recent book, we offer answers to these questions (O’Rourke and Williamson 2017). We explore the origins of modern industrial growth around the global periphery – Southern and Eastern Europe, the Middle East, Asia, Latin America, and sub-Saharan Africa, regions that fell behind the West during the Great Divergence. First, we document when modern manufacturing emerged in these poor regions, which industries were involved, and how industrial development evolved over time. We then explain these patterns, asking whether the timing of early industrial growth was the result of market forces or government policy. We show where factor endowments encouraged early industrial development in the Rest, and where they did not. We explore the importance of foreign markets, or, alternatively, the role of protected home markets. Finally, we assess alternative methods of technology transfer.
Factor endowments had a profound impact on industrial policy and performance. Labour-abundant and resource-scarce countries could enter at the bottom of the ladder, producing and exporting labour-intensive products (e.g. East Asia). Labour-scarce and high-wage periphery countries could not exploit that strategy, and thus relied on a tariff-protected domestic market (e.g. Latin America). Where the labour-scarce economy had only a small domestic market (e.g. Southeast Asia), industrial growth was difficult.
The poor periphery was characterised by widely varying factor endowments. Sub-Saharan Africa, Latin America, and Southeast Asia were resource-abundant and labour-scarce, and consequently found it difficult to develop modern labour-intensive manufacturing. Commodity export processing, and later import substituting industrialisation, were the typical routes to industrialisation for such economies. Over time, however, factor endowments evolved in ways that made it easier to sustain a competitive modern manufacturing sector. Populations grew fast, leading to falling relative wages, and bigger domestic markets.
The supply of educated labour seems to have been just as important as the supply of overall labour in improving the ability of poor countries to develop modern manufacturing. When modern manufacturing began its spread to the periphery, a lack of skills was often an important constraint. It follows that loosening that constraint was crucial in spurring industrial growth. From this perspective, European colonisers damaged their colonies: it was not until their colonies achieved independence that major progress was made towards providing universal primary (and later secondary) education. After independence, a literacy revolution took place almost everywhere around the periphery.
It was easier to overcome a shortage of financial and physical capital than human capital. Poor economies had limited supplies of capital, but in periods when international capital markets were working smoothly, financial capital was borrowed from abroad. This was certainly the case in the first global century up to WWI for Imperial Russia, colonial India, and Latin America. But borrowing from abroad was only possible when international capital markets were functioning properly, and this was not the case during the de-globalising years from the 1920s to the 1970s.
Even more important than financial capital inflows were imports of equipment and machinery, which embodied up-to-date technology. These imports were crucial for countries seeking to build a modern manufacturing sector. They had to be paid for with export earnings when international borrowing was difficult. Otherwise, balance-of-payments constraints suppressed investment, with serious consequences for domestic manufacturing.
Governments attempted to relax these constraints in various ways. The USSR, India, Yugoslavia, and Romania all built up their capital goods industries, thus supplying more of their equipment needs. The USSR famously encouraged high savings rates, attempting to dramatically shift the country’s factor endowment in a capital-intensive direction. Such policies boosted industrial output initially, but eventually ran into severely diminishing returns.
International context and luck
By the mid-19th century, the global periphery had become the commodity exporter to the industrial leaders, so world commodity price trends and their volatility were central to local manufacturing profitability and performance. That is, commodity price booms through the 1890s generated Dutch Disease: labour and capital rushed to commodity export sectors and fled domestic manufacturing (Williamson 2011). However, from the 1890s on, commodity prices fell and the relative price of manufactures rose, stimulating the growth of domestic manufactures in Asia, Africa, the Middle East, and Latin America up to WW2. On the other hand, quasi-non-tradable consumer goods industries in Latin America and elsewhere depended on local incomes, and thus grew during commodity booms.
Since world markets for manufactured exports mattered, so did geography. Mexico’s manufacturing growth has always been favoured by its big and fast-growing northern neighbour; Southeast Asia, Taiwan, and Korea were favoured by Japan’s post-WW2 economic miracle; the same countries were favoured again by China’s economic miracle since the 1980s. And Central and Eastern Europe were favoured by the fast growth of Northwest Europe up to 1913; disfavoured by European disintegration after WW1 and their membership in the Soviet block after 1945; and were then favoured again by their re-connection to the EU from the 1990s onwards. These relationships not only fostered trade, but also FDI and technological transfer.
Location also mattered within countries. Modern industry in Austria-Hungary first emerged in the northwest – in what is today Austria and the Czech Republic – close to West European neighbours. It also appeared in the Italian northwest, also contiguous with a big West European neighbour. Factory production in China was initially located along the southeast coast, accessible to booming world markets, with a smaller cluster in Manchuria, close to booming Japanese markets. Modern Indian industry first appeared in port cities such as Calcutta and Bombay. In Turkey, modern factories initially clustered in Constantinople and Izmir in the west and Adana in the south, both coastal regions. These regional agglomerations were typically linked to trade connections with the rest of the world: port cities offered access to foreign capital, cheap raw material imports, entrepreneurship, and modern technology, or access to foreign markets. Factory production was more regionally dispersed where geography inhibited national market integration, as in Mexico, Columbia, and Chile.
Finally, luck mattered, both good and bad. Latin America dropped its trade barriers in the late 1970s only to have China flood world markets with manufactures beginning in the 1980s. This was very bad luck. Southeast Asia started its miracle in the 1970s when Japan shifted from labour-intensive to capital-intensive technologies and used FDI to move their older technologies to Malaysia, Thailand, and other Southeast Asian countries. The region was again favoured by a booming Chinese market starting in the 1980s after Southeast Asia had established a competitive manufacturing industry. Thus, the region was twice blessed with good luck.
Policy has always mattered, but the policies necessary to promote modern manufacturing varied across regions and over time. Tariffs spurred the growth of import-competing manufacturing, but they hampered the growth of export-oriented commodity processing, and early on this was the most important modern industrial activity in resource-abundant and labour-scarce regions such as Southeast Asia and Latin America. More fundamentally, the need for protection of manufacturing depended on underlying patterns of comparative advantage. In labour-abundant countries, labour-intensive manufacturing had at least a chance of getting off the ground without the artificial stimulus of tariffs: in labour-abundant China and India, for example, modern manufacturing first emerged during the late 19th century under conditions approximating free trade. When post-colonial governments in such countries decided to promote industrial growth, there was genuine growth potential. In labour-scarce countries, however, protection was probably going to be required if labour-intensive manufacturing was to get off the ground: industrialisation in peripheral Europe and Latin America typically originated behind tariff barriers, which made sense given their resource-abundant and labour-scarce factor endowments. However, such later import substitution industrialisation policies were explicitly working against their comparative advantage. Thus, when these countries liberalised in the 1980s and 1990s, many lost a good deal of the industry that had been built up under protection, as in East and Southeast Europe.
The impact of policy was particularly dramatic in those economies which turned to Communism: Russia, and especially Bulgaria and Romania, did not have a natural comparative advantage in manufacturing; but China did, at least when it came to labour-intensive activities. All promoted capital-intensive heavy industry for ideological reasons, which clearly went against their initial comparative advantage, and eventually suffered massively due to diminished efficiency. The experience following liberalisation in the 1980s or 1990s has differed greatly across these countries. Chinese central planners helped lay the foundation for the subsequent growth miracle, by changing factor endowments, importing technology, and providing a manufacturing base that would become much more efficient. East European countries have deindustrialised since 1989, while even Russia has reverted to more resource-exporting. India, which also pursued capital-intensive industrialisation strategies, saw its service sector expand dramatically after liberalisation. These examples suggest once again that comparative advantage determined the impact of liberalisation.
Modern manufacturing growth in the global periphery dates back to the interwar period, and in some regions much earlier. It depended on a complex interaction between factor endowments, the global context, economic policies, and luck. The effectiveness of policies depended on the geographical and historical context. There was not one unique recipe leading to the development of modern manufacturing.
-Kevin O’Rourke, Jeffrey Williamson