| 摘要 | Oxford Economics, a UK-based economic forecasting consultancy, published a report last week on the global, national and sub-national impacts of industrial robotics. It has some eye-popping projections.
According to the authors, the disruptive effects of robotics have thus far been small compared to those just over the horizon. Since the late-1990s, a total of 2.25 million industrial robots have been deployed around the world. By 2030, we can anticipate that as many as 20 million such machines will be in use. China alone will have nearly 14 million manufacturing robots, 20 percent of the global total up from .1 percent in 2000.
The projected employment impacts are profound. Oxford Economics estimates that since 2000, the industrial robots have displaced 1.7 million manufacturing workers world-wide, with each robot replacing about 1.6 human workers. China has lost the largest share (550,000), over twice as many as the U.S. (260,000). However, as a percentage of its manufacturing workforce, the U.S. impact has been twice that of China, 2 percent versus 1 percent losses in the two countries’ respective workforces. Overall, under the high-investment scenario, global manufacturing employment would decline sharply: we could see as many as 30 million manufacturing jobs disappear by 2030.
The report shows these displacements would be unequally distributed. Lower-income areas (internationally and domestically) are far more vulnerable to automation than higher-income areas. The introduction of a robot into low-income area displaces 2.2 workers compared to 1.3 workers in a higher-income area. The authors argue this difference is task-driven: lower-income areas produce goods while higher-income areas house those who manage production; management tasks are more insulated from automation. As we’ve seen, this disproportionate impact on “makers” means workers in industrialized but comparatively poorer areas will continue to bear the brunt of automation, a recipe for accelerated populism.
China plays a central part in this story. Over the past 20 years, it has drawn on its vast reserves of human capital to become the world’s low-cost manufacturing hub. Oxford Economics estimates that close to 14 million Chinese manufacturing workers can expect to be displaced over the next decade by automation. That figure is more than the entire U.S. manufacturing sector which raises some interesting challenges for a nominally communist workers’ regime. The new wealth generated by higher productivity will undoubtedly create new jobs but the question is how China will manage a transition that occurs quickly on an unprecedented scale.
There are some interesting twists and turns in the robotics story for the U.S. economy as well. A diverse group of states are significantly exposed to job displacement including Oregon, Washington, California, Texas, Indiana and North Carolina. The “left Coast” issue deserves special note. States like Oregon went through a wrenching transition from extraction industries like timber and are now have large high-tech manufacturing workforces in the Silicon Forest outside Portland. It’s as if the Luddites had invented the steam-driven loom: the high-tech manufacturing workforce is vulnerable to its own industry’s advances. Similar effects may occur in California and Washington State. It’s anyone’s guess what the highly combustible mixture of job displacement and progressive politics will produce.
The key question is where those relieved of their manufacturing jobs will go. In the past, three sectors – transportation, construction and maintenance, and office/administration work – absorbed more than half the workers who left manufacturing. “Ominously,” says Oxford Economics, “our analysis found that [those] three occupational areas are among the most vulnerable to automation over the next decade.” As the report notes, nations need to resist the temptation to protect industries and jobs against automation as doing so would result in their workers and industries falling even further behind the automation and competitiveness curves. Instead, the authors argue for shared responsibility in responding to the projected changes in labor market demand: a genuine commitment to life-long learning among workers, investments in up-skilling by employers, and a better balance in training between technical and non-technical skills (i.e. soft and noncognitive skills) that prepare workers for sectors that are harder to automate.
To those suggestions, I would add a fourth: we need to look for alternatives in the way we finance and organize public programs that help dislocated workers re-skill, up-skill, and find employment in a rapidly evolving economy. A haphazard collection of federal programs across multiple agencies, such as the cumbersome, under-utilized, and ineffective Trade Adjustment Assistance entitlement are just not up to the challenge of the employment transitions workers experience. A flexible and adaptable system that combines individual responsibility for skills development with comprehensive, worker-directed training resources, improved, public-private economic planning partnerships, and better real-time labor market information would be good place to start the discussion. |
