In previous articles we have discussed how the advance of technology – and specifically the rise of automation technologies, robotics, and artificial intelligence – stands ready to disrupt developing Asia’s reliance on low-cost manufacturing to power early-stage economic development.
Indeed, preliminary results from a study by McKinsey Global Institute on the industries that could be affected by automation and artificial intelligence put manufacturing as the 2nd most automatable industry, after accommodation (hospitality) and food service, due to the significant amount of “predictable physical work” in both these sectors. The study, which covered 2000+ work activities across 800 sectors in the U.S., cited predictable physical work as highly susceptible to automation based on the current state of the technology.
This should be worrying for anyone who works with factories in developing Asia. Even in China, which leads the world in purchases of industrial robots, factories feature row and rows of workers engaged in “predictable physical work.” Granted, this is changing slowly, as the government tries to push industry up the value chain. As it does, the low-skill work has moved inland or offshore to Vietnam, Indonesia, Cambodia, or Bangladesh, among others, which have actively courted buyers and factory owners seeking an alternative to China.
This process demonstrates neoclassical trade theory at work, where countries move towards positions of natural comparative advantage vis-a-vis other countries, producing a balanced composite at the global level. Automation poses a challenge to this labor-cost offshoring model, by saying that it may be cheaper to automate inside the factory, rather than going offshore in search of low-cost workers. And, as we have discussed previously, there are strong “push” factors that support greater automation, including the need for speed and specialization, declines in order quantities, the wish to avoid labor unrest, and so on.
In either case, there will be displaced workers. Under the neoclassical trade theory, these displaced workers should be absorbed by new industries created as the economy converges towards a new point of comparative advantage. In other words, they get hired by other companies, and everyone ends up satisfied.
Reality tells us that this theory leaves a lot to be desired. A 2016 study by MIT’s David Autor and two colleagues tells us just how much.
Using a variety of regional economic, trade and census data, they studied the so-called “China shock”, that affected U.S. manufacturing industries and communities due to competition with China as the latter developed its manufacturing sector and acceded to the WTO, roughly from 1999 to 2011. The authors identified 379 sectors which were subject to competition from imports from China, and traced the impact on employment and livelihoods in the U.S.
The findings are sobering. Regions which were more exposed to import competition from China recorded larger declines in manufacturing employment, notwithstanding that manufacturing jobs were also declining nationally due to automation and other factors. Moreover, those employment declines then radiated up and down local supply chains, resulting in an estimated 2.4 million jobs lost to import competition or knock on effects.
Workers in the affected regions recorded higher levels of adjustment disbursements such as unemployment and disability insurance, but only enough to cover slightly more than 10% of the lost income from their jobs. Worryingly, workers whose sectors were more exposed to import competition from China in 1991, recorded lower earnings growth and more job churning in the subsequent period, when compared to others with similar demographic characteristics.
Critically, employment data showed that labor lost due to competition with China did not shift to other tradable sectors. In other words, labor markets proved to be rigid rather than fluid, and individual workers suffered a net loss, as did local economies. Remarkably, these outcomes emerged in the United States, which is not generally regarded as an economy with a high degree of labor market rigidity. While there were no obvious barriers to people moving to take jobs in other areas (presuming that other jobs did exist), this did not happen.
What does this all have to do with Asia?
Economic transitions happen all the time, and can be driven by any number of factors, such as the emergence of competitor nations, demographics (which determine labor supply), national regulation or international trade agreements, or, as we see now in Asia, the rise of automation, robotics and AI. In every instance, there will be job losses and job gains, but these do not cancel each other out. The Autor et al paper reminds us how damaging labor market rigidity can be, both economically and also socially.
This is a timely reminder, as many countries in developing Asia, including China, continue to promulgate changes to their employment and labor contract laws, to afford workers basic protections while promoting industry growth and flexibility. But governments, as well as the corporate sector, need to think beyond just labor law, to promote fluid labor markets through regulatory support, education and training, residential mobility, social services, and more. Moreover the digital age is changing the kinds of skills and attributes that employers desire and need from employees: for example physical mobility may matter less than digital literacy. In other words, how does the textbook understanding of labor market flexibility need to evolve for the age of automation?
The advent of automation will both speed up and amplify economic changes, which will become a continuous process rather than something that happens at a given point in time (as the “China shock” did). Labor markets will need to adjust to the same tempo. We should begin planning for this now.
The Mckinsey Global Institute study can be found here.
The study by Autor et al: