Shigeyuki Abe: This paper starts with a concise and yet comprehensive survey of literature on female labor participation and the authors make their position of the analysis clear. Their contribution is to shed light on the role of ICT and offshoring as determinants of female labor demand by estimating a system of labor demands, controlling for the effects of skill-biased technological change and offshoring simultaneously.

To the extent that the authors use the data set of ICT capital deepening and labor education level as quality of labor, their modeling endeavor is successful. Their most interesting finding is that the ICT capital stock has significantly negative effects on the demand for middle-low-skilled female workers, whereas it has significantly positive effects on the demand for low-, middle-high-, and high-skilled female workers. The use of ICT such as computers and other high-tech equipment can increase the demand for female workers in general. This is because skill-biased technology may change physically demanding tasks to less physically demanding tasks. The demand for female workers would thus increase as a result of technological change.

Their finding is in line with Weinberg (2000), who concludes that increases in computer use accounted for more than half of the growth in demand for female workers in the United States. The question, however, is why the ICT capital stock has significantly negative effects only on the demand for middle-low-skilled female workers. The explanation to this question unfortunately cannot be found in the paper, although the authors argue that governments should not hesitate to enhance ICT to expand female employment because the other subgroups show positive effects and this negative group makes up only a small share of employment (5.1 percent in 2011). In this regard, Kiyota and Maruyama might need to explore the data of middle-low-skilled female labor because they pick up the peaks of a zigzag trend every five years, starting from 1980 (Figure 3).

The same ICT capital data are used for different types of labor in this paper because of the unavailability of such data. If one can find such detailed ICT capital data and use them in the model estimation, then we can analyze the exact impacts of ICT on each type of female labor demand. Common sense suggests that middle-low-skilled female labor can be replaced by robots, and quite likely this labor will not be equipped with ICT capital. Using sectoral total ICT capital for this labor is thus not appropriate. This might be an explanation of the negative coefficient in their regression.

Another point relates to the service sector. Weinberg (2000) treats industry as a whole, inclusive of the service sector. In Japan, the share of manufacturing has been declining as well as the share of manufacturing employment. Female workers often choose the service sector in Japan mainly because the workplace is generally close to home. It might be better to analyze industry as a whole rather than manufacturing only. In addition, the share of female part-time workers is high in general in the service sector.

Using their data set simple correlations are calculated here in Table 1.

Table 1.
Simple correlation between share of female workers and ICT capital across manufacturing sectors
Year
 1980198519901995200020052011
(1) and (2) 0.571 0.627 0.669 0.743 0.778 0.801 0.687 
(1) and (3) 0.120 0.228 0.213 0.121 0.078 0.062 0.019 
(2) and (3) 0.036 0.067 0.113 −0.046 −0.107 −0.212 −0.044 
Year
 1980198519901995200020052011
(1) and (2) 0.571 0.627 0.669 0.743 0.778 0.801 0.687 
(1) and (3) 0.120 0.228 0.213 0.121 0.078 0.062 0.019 
(2) and (3) 0.036 0.067 0.113 −0.046 −0.107 −0.212 −0.044 

Source:Author's calculation based on the Kiyota and Maruyama paper.

Note:(1) share of female workers, (2) share of part-time workers, (3) ICT capital.

Correlations are high between the share of female workers and that of part-time workers across manufacturing sectors, but correlations are low between share of female workers and ICT capital and between part-time workers and ICT capital share.

Based upon these observations my two suggestions for their future work are as follows.

  • A)

    For the analysis of ICT capital on female labor, it might be better to include the service sector as well. This is simply because a large number of female workers is employed in the service sector. Japanese characteristics of female labor should be explored more carefully and the differences from that of the United States should be analyzed.

  • B)

    The offshoring argument is interesting but it requires input–output data. It is a tradeoff between using less data without the service sector for the ICT capital and offshoring analysis and using full data to deepen the analysis of ICT capital on female workers. Because the effects of offshoring are not substantial, it is worthwhile to try to use full time series and service sector data.

Reference

Weinberg
,
Bruce A.
2000
.
Computer Use and the Demand for Female Workers
.
Industrial and Labor Relations Review
53
(
2
):
298
308
.