The Technology-Employment Trade-Off : Automation, Industry, and Income Effects

New technologies can both substitute for and complement labor. Evidence from structural vector autoregressions using a large global sample of economies suggests that the substitution effect dominates in the short-run for over three-quarters of economies. A typical 10 percent technology-driven improvement in labor productivity reduces employment by 2 percent in advanced economies in the first year and 1 percent in emerging market and developing economies (EMDEs). Advanced economies have been more affected by employment-displacing technological change in recent decades but the disruption to the labor market in EMDEs has been more persistent. The negative employment effect is larger and more persistent in economies that have experienced a larger increase, or smaller fall, in industrial employment shares since 1990. In contrast, economies where workers have been better able to transition to other sectors have benefited more in the medium run from the positive ``income effect'' of new technologies. This corresponds with existing evidence that industrial jobs are most at risk of automation and reduced-form evidence that more industrially-focused economies have tended to create fewer jobs in recent decades. EMDEs are likely to face increasing challenges from automation as their share of global industry and production complexity increases