Regional Employment and Artificial Intelligence

         
Author Name HAMAGUCHI Nobuaki (Faculty Fellow, RIETI) / KONDO Keisuke (Fellow, RIETI)
Creation Date/NO. March 2017 17-J-023
Research Project Regional Economies in the New Era of Globalization and Informatization
Download / Links

Abstract

There is growing concern that human jobs will be replaced by the rapid technological progress of artificial intelligence (AI), robotics, and automation. This study investigates employment risk to computerization by using the probability of computerization and employment data in Japan. The new aspect of this study is to consider regional heterogeneity because the geographical distribution of occupations is uneven, which is observed particularly between male and female workers. Thus, this study analyzes the employment risk to computerization in terms of city size and gender.

Our results find that female workers are exposed to higher risks of computerization than male workers. Importantly, this impact becomes bigger as the city size increases. In addition, we find that occupations with high probability of computerization are, on average, low-educated and low-wage, thus implying that workers need additional human capital investment to obtain occupations with lower risks to computerization.

Our policy implications emphasize that, as mentioned by most of the previous studies, human capital investment is necessary to mitigate future employment risks to computerization. Moreover, we also find that even some high-skilled workers face high risks of computerization, and thus active labor market policies to facilitate their job mobility are necessary. In addition, the important concept is that AI can complement human activities, not replace them. Utilizing AI and robotics supports business efficiency and better work-life balance, which can solve issues on long working hours in the Japanese labor market. Therefore, it is important, especially, for female workers, to consolidate the employment system fundamentally.

The English version of this paper is 18-E-032.