Since the Industrial Revolution began, every country that hasn’t been lucky enough to have huge oil deposits has gotten rich the same way: by getting good at manufacturing. Countries like the U.K., Germany, the U.S, Japan, and South Korea all became world-class manufacturers long before their economies began to shift toward services. More recently, a new crop of almost-developed countries including Malaysia, Poland, Turkey, China, Romania, Thailand and Mexico have also strengthened their manufacturing capacity.
Economists like Ha-Joon Chang and Dani Rodrik and writers like Joe Studwell have argued that intentional promotion of manufacturing exports is crucial for this kind of development. In a 2008 paper, Rodrik summarized much of the empirical and theoretical case in favor of industrial policy. “Development,” he writes, “is fundamentally about structural change” toward producing high-value exportable products -- most of which tend to be manufactured goods. Competing in export markets also forces a country’s producers to increase productivity and enables them to adopt advanced foreign technologies.
A number of poor countries have been trying to put this idea into practice. Two prominent examples are Vietnam and Bangladesh, which have experienced exponential growth in recent years.
Their industrial profiles look just right for countries on the first rung of the manufacturing ladder. Vietnam’s top exports are electronics and clothing, while Bangladesh’s are clothing and textiles.
But worryingly, Rodrik has been arguing for several years that the window for successful manufacturing-driven development has closed. In a 2016 presentation entitled “Is the Age of Growth Miracles Over?”, he argued that countries like China and Malaysia would be the last to use manufacturing to make the leap from poverty to wealth.
Rodrik’s argument was mainly an empirical one. While most countries eventually shift from manufacturing to services after they get rich, Rodrik observed that in recent decades, this shift has been happening earlier and earlier -- so that many countries now actually start shifting away from manufacturing before they fully industrialize.
Other research seems to have documented a similar phenomenon. For example, a 2015 paper by economists Douglas Gollin, Remi Jedwab and Dietrich Vollrath found that in many developing countries, urbanization now means poor people moving off of farms to work in local services rather than in factories.
For countries like Vietnam and Bangladesh, this is bad news. If Rodrik is right, their current burst of manufacturing prowess may evaporate, and they’ll find themselves following in the footsteps of countries like Indonesia, Nigeria, and Brazil, where manufacturing has been declining as a share of GDP.
What’s the reason for this deindustrialization? One obvious explanation is automation. If robots are replacing the lower end of the supply chain, then the poor, unskilled workers who traditionally power low-end manufacturing industries, like garment, toys, and electronics assembly, no longer have a comparative advantage. Poor countries would presumably then specialize in natural resource exports and low-value service exports like call centers, leaving advanced countries with machine tools and robots to build all the physical stuff.
But there are good reasons to doubt that this is happening. The link between automation and employment is tenuous; some studies find that industrial robots replace human workers, but others find that they increase employment, including among the low-skilled.
More importantly, the timing is wrong for technology to be the story. Much of the premature deindustrialization Rodrik documented -- including in Nigeria and Brazil -- happened in the 1980s and 1990s. At the exact same time, China was ramping up its manufacturing capacity in labor-intensive industries and beginning its own growth miracle. So it seems likely that something else was going on in the countries where industrialization reversed.
Many of these countries may have simply had a dysfunctional, premature industrialization. In Africa, for example, a number of countries used a policy of import substitution, closing themselves off to trade and trying to manufacture their own goods, often using inefficient state-owned factories. Brazil used a similar approach. This is very different than the kind of export-focused, productivity-boosting industrial policy that Rodrik, Chang or Studwell would recommend. And then when these industrialization drives left countries poor, they opened their borders to imports, which decimated inefficient local industry.
In addition to policy mistakes, China’s own industrialization may have played a role. That huge country’s incredible manufacturing prowess may have caused multinational companies to turn away from slower-footed competitors like Indonesia, stifling their own nascent industries. Notably, Rodrik found that on average, East Asian countries hadn’t experienced premature deindustrialization like their counterparts in other regions of the world.
An economic theory created by Paul Krugman, Masahisa Fujita, and Anthony Venables may hold the key to why this happened. They predicted that as the world economy develops, regions industrialize in spurts, one after another. If this is right, then other countries would have had no choice but to wait for China to finish its own growth miracle before starting their own.
That time may now have come. Thanks to rapidly rising Chinese costs, the US-China trade war, and multinational companies’ desire to diversify in the face of threats like coronavirus, it might be time for countries like Vietnam and Bangladesh to take over much of labor-intensive manufacturing. Even countries that deindustrialized, like Indonesia, might now get a second chance. Poor countries shouldn’t give up on the dream of manufacturing-based development just yet.
Bloomberg
