## Land of the Rising ROBOTS

## Source details

**Canonical URL:** [Land of the Rising ROBOTS](https://www.imf.org/-/media/files/publications/fandd/article/2018/june/schneider.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/fandd/article/2018/june/schneider.pdf.md)
- [Structured JSON version](/-/media/files/publications/fandd/article/2018/june/schneider.pdf.json)

---

### Demographic context and labor constraints
- Japan’s estimated population fell by a record-breaking 264,000 people in 2017.
- Deaths outnumber births by an average of 1,000 people a day.
- Japan’s birth rate currently stands at about 1.4 births a woman; 2.1 births a woman are needed to sustain growth.
- Nearly a third of Japanese citizens were older than 65 in 2015; research suggests that number will rise to nearly 40 percent by 2050.
- The UN Population Division estimate shows Japan’s population will dip below 100 million shortly after the middle of the 21st century.
- By the century’s end, Japan stands to lose 34 percent of its current population.
- Japan’s domestic labor force (those ages 15–64) is projected to decline by some 24 million between now and 2050.
- Immigration is unlikely to rise enough to compensate for this dramatic decline anytime soon.

### Japan’s robotics and automation landscape
- Historical and manufacturing facts:
  - About 700,000 industrial robots were used worldwide in 1995; 500,000 of them were in Japan.
  - Japan exported some $1.6 billion worth of industrial robots in 2016—more than the next five biggest exporters combined (Germany, France, Italy, United States, South Korea).
- Robot density (number of robots per 100 workers) reported in Chart 2:
  - Korea: 6.31 robots
  - Singapore: 4.88 robots
  - Germany: 3.09 robots
  - Japan: 3.03 robots
  - Sweden: 2.23 robots
  - World Average: .74 robots
  - China: .68 robots
- Japan has been a leader in robotic production and industrial use; Kawasaki Robotics started commercial production of industrial robots over 40 years ago.

### Productivity patterns and potential of automation
- The services sector accounts for 75 percent of GDP and has seen little annual productivity growth—only about half that of the United States.
- Labor productivity changes since 1970:
  - Manufacturing: roughly tripled since 1970.
  - Nonmanufacturing: improved by only about 25 percent since 1970.
- The most productive manufacturing sectors in Japan—automotive and electronics—are heavily reliant on automation.
- Automation and AI are spreading beyond manufacturing into transportation, communications, retail services, storage, health care, financial services, and other services (examples: self-checkout registers, touch-screen order terminals, robot chefs, hotel staff).

### Empirical findings and simulated scenarios
- IMF staff calculations (using an approach pioneered by Acemoglu and Restrepo 2017, with prefectural-level data from Japan) found:
  - Increased robot density in manufacturing is associated with greater productivity and local gains in employment and wages.
  - These findings exclude crisis periods and differ from similar exercises based on US data.
- IMF staff simulation across the Group of 20 industrialized countries points to risks under substantial transition costs (unemployment, lower wages) as automation substitutes for existing human labor:
  - Risk of declining labor shares.
  - Income polarization.
  - Rising inequality.
- Applying the same approach only to Japan yields different results:
  - With a shrinking labor force, even fully substitutable automation could boost wages and economic growth.
  - Automation and robotics could fill the labor gap and result in higher output and greater income rather than wholesale replacement of the human workforce.

### Distributional and social risks
- Polarization risk: a relatively small proportion of workers may have the training and education needed to fully leverage productivity from robotics.
- Female labor force vulnerability:
  - The female labor force has swelled in the past five years but is particularly vulnerable to displacement.
  - Women are heavily concentrated in nonregular jobs (temporary, part-time, or other positions outside the mainstream of Japan’s lifetime employment system), whose tasks are more susceptible to automation (Hamaguchi and Kondo 2017).
- Declining quality of services:
  - Research (Morikawa 2018) shows quality of services is eroding as a result of labor shortages.
  - Most critically affected: parcel delivery services, hospitals, restaurants, elementary and junior high schools, convenience stores, and government services.

### Policy implications and recommendations
- Policymakers should accept that change from artificial intelligence, robotics, and automation is coming and prepare accordingly.
- Key policy responses emphasized in the source:
  - Strengthen and maintain strong and effective social safety nets to manage inevitable disruption of traditional labor and social contracts.
  - Invest in education and skills development to enable more people to take advantage of jobs in a high-tech world.
  - Work to bring greater equality into the labor force—between men and women, between regular and nonregular employees, and across regions—so that the benefits and risks of automation can be more equally shared.
- Legal and infrastructure hurdles to address for wider deployment of robotics and AI include:
  - Consumer protection.
  - Data protection.
  - Intellectual property.
  - Commercial contracting.

*Source: Schneider, Todd; Gee Hee Hong; Anh Van Le, “Land of the Rising ROBOTS,” Finance & Development, June 2018.*

---


_Source: https://www.imf.org/-/media/files/publications/fandd/article/2018/june/schneider.pdf_