Comment on “Working Elderly in Japan: Determinants and Outlook”

Kitao and Takeda (2025) offer a compelling analysis of the labor supply of the elderly in Japan, employing a simple dynamic model to examine how factors such as the pension system, health changes, preference heterogeneity, and income processes influence elderly employment. Their counterfactual simulations, derived from a structural model, yield a striking result: extending the pensionable age by 5 years or reducing the replacement rate by 20% could increase the labor force participation rate of Japanese men in their 60s by 12 percentage points. Given the scarcity of studies utilizing dynamic structural models to explore elderly employment and health issues, this research stands out for its methodological rigor and clear policy implications. In this comment, I will highlight several issues related to elderly employment in Japan that Kitao and Takeda (2025) have not fully addressed, which could enhance future analyses in this field. Kitao and Takeda (2025) assume that elderly individuals can work under conditions that match their marginal productivity. However, this assumption may oversimplify the labor market realities faced by retirees. According to the 2020 Survey on the Economic Life of Senior Citizens conducted by the Cabinet Office, nearly 20% of Japanese men aged 60 and over wish to continue working for as long as possible. When including those planning to work until age 70, this proportion exceeds 65%. While health constraints may limit some, the challenge of transitioning to new employment opportunities for elderly workers could be more significant than Kitao and Takeda's model suggests. One of the complexities in analyzing employment and consumption among the elderly is the significant heterogeneity within this demographic group. Figure 1 illustrates the differences in work types across age groups of Japanese males based on data from the Labour Force Survey 2023. In their late 50s, most employed individuals are wage earners, with the self-employed comprising only 6% of the population in this age group and less than 10% of the employed. However, in their late 60s, the self-employed account for 13% of the population and over 20% of the employed. Among those aged 75 and above, more than 45% of the employed are self-employed. Source: Labour Force Survey 2023 Although the number of startups in Japan has been increasing in recent years, the proportion of those initiated by the elderly remains very small. According to Sadaf (2024), the variances of (log) income and consumption among the self-employed are significantly higher than among wage earners, indicating substantial differences in income processes between the two groups. Furthermore, those who become self-employed as a result of starting a business likely possess different risk preferences, abilities, and assets compared to those who remain wage earners. If the results of Kitao and Takeda (2025) are to be extended to individuals in their 70s, or if a more detailed analysis of elderly employment is to be conducted, it will be necessary to analyze the self-employed and wage earners separately. Kitao and Takeda's (2025) model does not account for the retirement age set by companies, and factors related to labor demand are abstracted away. However, for those who continue working at a company, if they choose to stop supplying labor after reaching the company's retirement age, the impact of the pension reforms discussed in Kitao and Takeda on the elderly labor supply may not be as significant as their model suggests. Kitao and Takeda (2025) clearly demonstrate the usefulness of dynamic structural analysis in studying elderly employment in Japan, and I look forward to further advancements in this area of research.