Comment on “An Assessment of Abenomics from the Labor Market Perspective”

Abstract

Kawaguchi et al. (2021) use data from the Labour Force Survey of Japan to examine the changes in employment rates from 2000 to the present. They show that Japan's employment rate has risen by 1.1 percentage points from 2000 to 2019, despite the continued population aging. According to their decomposition results, had the participation behavior stayed constant at the 2000 level, the employment rate would have decreased by 5.2 percentage points in 2019. Kawaguchi et al. demonstrate that the primary contributors to this rise were females aged 25–64. Their reserach further shows that: (i) most of the rise in female participation is a shift from out-of-the-labor-force to employment, not so much from unemployment to employment; (ii) there was increased participation of women in the health care and social welfare sector; (iii) the large rise in female employment rates is not explained by various family policies (such as expanded child daycare or the public provision of elderly care); and (iv) the share of temporary employment did not increase during Abenomics. Kawaguchi et al. also examine the impact of the COVID-19 pandemic on employment rates, and conclude that the fall in employment was modest. As Kawaguchi et al. show, Japan has faced a gradual shrinkage of its labor supply, which poses a labor shortage pressure. According to Japan's population census and its predictions (Statistics Bureau of Japan, 2020), the number of people aged 15 or over was 108 million in 2000, 111 million in 2013, and 111 million in 2019. Therefore, the total number of those aged 15 or over has not decreased much from 2000 to the present, because of increased longevity. In contrast, the number of people aged 15–64 was 86.4 million in 2000, 79.0 million in 2013, and 75.0 million in 2019. During the time of this supply pressure, the overall employment rates (the denominator of which is the population aged 15 or over) did not fall, thanks to the significant increase in female participation rates. During the period of Abenomics, the baby boomers have retired, so supply pressure had been present. Baby boomers were in their early 50s in 2000; they were in their mid-60s at the start of Abenomics; and they reached around 70 in 2019. Even with the demand shock due to the COVID-19 pandemic, the Japanese labor market did not become so weak, possibly because of the underlying supply pressure. I have two major comments. The first is about the nature of women's participation behavior. In the decomposition, the fact that female participation has risen over time could represent two different behaviors. First, later cohorts of women may participate more in the labor market in the same age range than did earlier cohorts. If later cohorts of women are less likely to quit the labor market upon marriage or childbirth than earlier cohorts, that would increase age-specific participation rates over time. Second, the same rise could happen if more women who were out of the labor force start working (re-enter the labor market) as they age. These two behaviors result in the same age-specific rise in female participation rates. Yet, in terms of individual behaviors, they mean different things. Section 2.4 of Kawaguchi et al. reports the transition from out-of-the-labor-force to employment increased among those aged 15–34 from 2013 to 2020. In contrast, the same transition was unchanged for those aged 35–54. This pattern might suggest that the re-entry to the labor market has increased after 2013. However, the increase for those aged 15–34 could come from other channels. For instance, if more students take part-time jobs than before, the inflow to employment increases. My second comment is about the long-term trend versus the impact of Abenomics. The rise in female participation is a long-term phenomenon that started before 1980. In Kawaguchi et al.'s figure 4, the rise in employment rates of females aged 25–44 seems to have been following the long-term trend, rather than any structural change taking place after 2013. In contrast, the rise in employment of females aged 45–64 seems to have accelerated after 2013, during the period of Abenomics. The reasons why we see these differences across age groups, and what their implications are, may be worth further investigation. If Kawaguchi et al. wish to assess Abenomics' impact, then perhaps the long-term trend (that is, the changes that would have happened without Abenomics) should be separated out.