Master's Thesis · Graduate School of Economics · Waseda University · 2026
This research examines how Japan's ageing population is associated with wage trajectories within the regular employment sector, with particular attention to whether early-career workers experience a systematically different wage response to the growing presence of elderly regular workers.
The study combines individual-level panel data from the Japanese Panel Study of Employment Dynamics (JPSED) for 2016–2024 with education-group-specific demographic measures constructed from the Labour Force Survey. This makes it possible to examine demographic labour-market pressure while simultaneously controlling for individual worker characteristics and persistent differences between individuals.
The results reveal two contrasting wage associations. A one-standard-deviation increase in the education-group-specific share of elderly regular workers is associated with approximately 1.5 percent lower wages for the general regular workforce. This pattern is consistent with structural wage compression resulting from the adjustment of Japan's internal wage-setting system to population ageing.
Workers aged 20–29, however, experience a substantially different response. For early-career regular workers, the same demographic variation is associated with a net wage differential of approximately +8.2 percent. This finding is consistent with a cohort-scarcity mechanism in which shrinking youth cohorts strengthen competition among firms for scarce young entrants at the point of labour-market entry.
Taken together, the findings suggest that population ageing does not have a uniform effect on wages across the career hierarchy. Instead, it is associated with general wage compression among the established regular workforce while simultaneously generating a more favourable wage differential for young workers entering the regular employment system.
Keywords: Ageing Population · Wage Trajectories · Cohort Size · Labor Market · Japan
JEL Classification: J11 · J13 · J21 · J82
Csiszár, C. (2026). The Effect of Japan’s Ageing Population on Wage Trajectories of Young Workers. Master's Thesis, Graduate School of Economics, Waseda University, Tokyo, Japan.
Does growth in the education-group-specific share of elderly regular workers reduce wages across Japan's general regular workforce?
And, within the same demographic environment, do early-career workers aged 20–29 experience a systematically different wage response?
The research connects two previously separate strands of the literature: long-run compression of Japan's age–wage profile and cohort-size effects concentrated at labour-market entry.
It directly examines whether these mechanisms generate asymmetric wage responses across different stages of the career as elderly regular employment expands within Japan's internal labour-market system.
A one-standard-deviation increase in elderly regular employment share is associated with approximately 1.5% lower wages for the general regular workforce.
Workers aged 20–29 show a positive youth-specific differential, resulting in a net wage association of approximately +8.2%.
The empirical analysis combines two complementary data sources. Individual-level information comes from the Japanese Panel Study of Employment Dynamics (JPSED), covering the period from 2016 to 2024. The panel provides information on income, age, gender, education, working hours, firm size, industry, occupational position, and employment status.
The analysis is restricted to regular employees aged 20–64, corresponding to the segment of the Japanese labour market in which long-term employment relationships and internal wage-setting structures are most relevant.
The second data source is the Labour Force Survey Detailed Tabulation, from which education-group-specific measures of elderly regular employment are constructed. Elderly regular workers are defined as those aged 60–64, while the core working-age regular workforce used in the denominator consists of workers aged 20–59.
The resulting elderly regular employment share is calculated separately for each education group and year. To isolate changes within educational groups rather than permanent differences between them, the measure is orthogonalized with respect to education-group fixed effects and subsequently centered and standardized.
The dependent variable is the natural logarithm of real annual labour income. Income is adjusted for inflation using the Consumer Price Index, allowing estimated coefficients to be interpreted approximately as percentage differences in wages.
The main empirical specification is a linear mixed-effects model with individual-specific random intercepts. This structure accounts for persistent unobserved differences in earning capacity and takes advantage of the longitudinal nature of the JPSED data.
A binary indicator identifies workers aged 20–29. Its interaction with the elderly regular employment measure tests whether young workers exhibit a different wage association from the broader regular workforce.
The specification additionally applies a Mundlak correlated random-effects correction. Individual-specific means of selected time-varying variables are included to separate within-individual wage changes from persistent between-individual differences in career trajectories, firm sorting, hierarchical position, and labour-supply behaviour.
Individual-level controls include gender, age, education, weekly working hours, firm size, industry, and occupational position. Standard errors are clustered at the education-group-by-year level, corresponding to the level at which the demographic explanatory variable varies.
The final estimation sample contains 33,713 observations from 5,653 individuals, with individuals observed for approximately six survey waves on average.
The full master's thesis is available below.