Cheap automation raises risk of ‘premature deindustrialisation’
We’ll send you a myFT Daily Digest email rounding up the latest San Francisco news every morning.
Historically, technology has shaped the economic trajectories of corporations, cities and even nations.
While the Industrial Revolution made the west rich and created the “great divergence” in incomes between the west and the rest, the adoption of western technology has more recently spread the fortunes of industrialisation to such places as South Korea, Turkey, and China.
Yet this process of rapid economic convergence risks coming to a halt, as labour-saving technologies put developing and emerging economies at risk of “premature deindustrialisation”.
Over the past 20 years, China has become the “factory of the world” while western economies have seen a decline in manufacturing employment. The growth of such cities as Shenzhen, where the iPhone is assembled, is the inverse of the decline of old US manufacturing centres such as Buffalo, Cleveland and Detroit — now part of the “rust belt”, since they have failed to produce new industries.
Nevertheless, China may be one of the last nations to ride the wave of industrialisation to prosperity.
As shown by Dani Rodrik of Harvard University, over the course of the 20th century, peak manufacturing employment has steadily declined in emerging economies. In Britain, the first country to industrialise, manufacturing employment peaked at 45 per cent before the first world war. By contrast, Mr Rodrik suggests manufacturing employment in Brazil, India and China has already peaked below 15 per cent.
This is in part because manufacturing processes are more automated. China in particular is not only the fastest-growing market for industrial robotics, it has replaced the US as the biggest market for automation.
In China’s 12th Five Year Plan, workforce automation is a strategic area the country’s leaders hope will sustain its competitive manufacturing edge.
While China’s push has been driven by rising wages and concerns over a “peak-out” of its working-age population, similar trends elsewhere are cause for more concern. Since the 1980s, low-income countries in sub-Saharan Africa have seen declines in their manufacturing share, as have middle-income countries in Latin America, making it less likely manufacturing will provide a route for workers seeking to escape poverty.
Automation alone cannot explain this trend; many emerging economies have seen falls in manufacturing output. An alternative, albeit complementary, explanation for the failure of low and middle-income economies to achieve industrialisation comparable to that of the west is also globalisation itself.
While the transport revolution in general, and the container ship in particular, paved the way for emerging economies to export cheaper, increasingly sophisticated goods, countries with a comparative disadvantage in manufacturing started to import deindustrialisation, as they became more exposed to production costs abroad.
These declining costs are largely driven by technology. For example, although the employment share of manufacturing in the US has fallen sharply, the output share has remained roughly constant over the past 50 years.
Thus, the US remains a competitive manufacturing location, but without producing many new production jobs. Even in low-cost locations, this will gradually become the case. Estimates by Citi Research show the payback period for industrial robots in China is now less than two years.
While many key technologies of the 20th century — the telephone, the container ship and the computer — helped companies co-ordinate cross-border production, shifting production to cheaper locations, developments in robotics and additive manufacturing have incentivised western companies to bring production back to automated factories.
In emerging middle-income countries, as well as low-cost destinations, labour will struggle to compete with ever cheaper technologies. In the absence of industrialisation as a path towards economic development, countries will need to discover new growth models.
Although service-led growth is one option, many low-skill services are today equally automatable. So the best hope for developing and emerging economies is to churn out more highly skilled workers. This is the lesson from the western experience.
Even though Ford and General Motors have some of the world’s most modern factories, Detroit has failed to modernise and produce employment in new industries. By contrast, some older manufacturing cities — New York, Boston and San Francisco — have made the transition to become innovators in services, offering financial and legal services, advertising, data processing and computer systems design.
While emerging economies such as China have been able to industrialise by leapfrogging technologies, the challenge for the next generation of emerging economies is that they will have to leapfrog industrialisation itself.
Models of economic development will need to feature more investment in education, faster implementation of new technologies and — most importantly — higher rates of local innovation. Managing this process is, however, easier said than done.
The writer is co-director of the Oxford Martin programme on technology and employment and Oxford Martin Citi fellow at Oxford university