# Aging and labor participation in Japan and Korea

Comparative Study · Labor Markets
Population aging places pressure on labor supply, but the number of people above a conventional retirement age does not determine the amount of work available. Participation, employment opportunities, hours and the organization of later-life careers modify the relationship. Japan and Korea illustrate why demographic projections should be connected to those margins rather than translated mechanically into a proportional loss of productive labor.

The common comparison in this study is the contribution of each age and sex group to actual working hours. A worker count is retained as a separate measure. The institutional comparison then asks whether continued work occurs in the same employment relationship, under a different contract or through another type of activity. These distinctions matter for earnings and capability as well as for the volume of labor supplied.

Official Japanese annual observations for 2024 and a Korean supplementary survey for May 2025 provide evidence about different parts of that adjustment. They are deliberately not placed in a direct ranking: their age bands and periods differ. The numerical decomposition below uses a common, hypothetical age structure to establish the comparison logic without manufacturing a harmonized estimate from incompatible observations.

## Japanese evidence: participation has more than one dimension

Japan's Cabinet Office 2025 aging white paper reports employment among people aged 65–69 at 62.8 percent for men and 44.7 percent for women in its 2024 evidence. It also reports substantial non-regular employment among employees in that age band, excluding executives:67.8 percent for men and 83.2 percent for women. Continued employment and continuation in the same employment form are therefore different observations. [Cabinet Office, 2025 aging white paper, employment and income](https://www8.cao.go.jp/kourei/whitepaper/w-2025/html/zenbun/s1_2_1.html).

The figures should retain their denominators. The employment rate relates employed people to the relevant population. The non-regular share relates a particular type of employment to employees excluding executives. Applying the second percentage to every older resident would be incorrect. It would combine different populations and ignore self-employment and other statuses.

Working time adds another margin. MHLW's March 2026 basic policy, citing the Labor Force Survey, reports that 42.1 percent of employed people aged 65–69 worked at least 35 hours in the reference week in 2024, compared with 59.6 percent among those aged 60–64. These are historical hours-distribution observations, not a legal definition of full-time work or a direct estimate of total hours. [MHLW, basic policy on older-worker employment](https://www.mhlw.go.jp/web/t_doc?dataId=75ac0966&dataType=0&pageNo=1).

The implication is specific. A rise in the number of employed older people need not replace the same quantity of labor time lost through the retirement of younger full-time workers. It can nevertheless sustain valuable skills, earnings and activity. The volume effect and the character of the work should be assessed together.

## Korean evidence: the survey boundary matters

Statistics Korea's supplementary survey for May 2025 covers people aged 55–79. It reports a labor-force participation rate of 60.9 percent and an employment-to-population ratio of 59.5 percent for that group. Within the survey, people aged 65–79 had participation of 48.2 percent and employment of 47.2 percent. These observations identify meaningful later-life work, but their broader age band cannot be compared directly with Japan's 65–69 figures. [Statistics Korea, economic activity of older people, May 2025](https://www.kostat.go.kr/boardDownload.es?bid=11735&list_no=438137&seq=1).

The same publication examines employment histories and the longest-held job. This is an important institutional distinction: leaving a principal career position need not mean leaving all paid work. A worker can move to another employer, reduce hours or become self-employed. A statistic describing departure from a main job therefore does not independently establish the age of final labor-market exit.

For a country comparison, the appropriate record distinguishes current employment, previous career attachment and desired future work. An observed transition into a new job can preserve participation while changing earnings, hours and the use of accumulated skills. Treating all of those outcomes as a single retirement event would conceal the channel through which demographic adjustment occurs.

The Korean survey period is a particular month, whereas the Japanese evidence cited above refers to an annual framework. Seasonal work and the within-year timing of interviews can affect a comparison. This boundary is material even if the age groups could otherwise be aligned. A common annual or common-month comparison is required before interpreting a small difference as a structural feature.

## A common age-by-hours decomposition

For each age group, total hours can be written as population multiplied by the employment rate and average actual hours per employed person. Summing those group contributions produces the total. The identity keeps demographic composition separate from the behavioral and institutional margins affecting work.

Consider a hypothetical population of one thousand, divided into a younger and an older group. Initially, eight hundred are younger and two hundred older. Assume employment rates of eighty and forty percent, respectively, and weekly hours of thirty-five and twenty. These are invented parameters and do not estimate either country.

| Hypothetical group | Initial population | Employment rate | Weekly hours per employed person | Total weekly hours |
| --- | ---: | ---: | ---: | ---: |
| Younger | 800 | 80% | 35 | 22,400 |
| Older | 200 | 40% | 20 | 1,600 |
| Total | 1,000 | — | — | 24,000 |

Now shift one hundred people from the younger to the older group, keeping the total population and both employment and hours schedules unchanged. Younger hours become nineteen thousand six hundred, older hours become two thousand four hundred and total hours fall to twenty-two thousand. The demographic-composition effect is a decline of approximately 8.33 percent in working time under those fixed schedules.

If the older group's employment rate instead rises to sixty percent, its contribution becomes three thousand six hundred hours. Total hours reach twenty-three thousand two hundred, reducing the decline to approximately 3.33 percent. If its hours per worker also rise to twenty-five, total hours become twenty-four thousand one hundred, slightly above the initial total.

These alternatives do not show that higher employment or hours can be achieved without cost or constraint. Health, preferences, care responsibilities, occupations and employer demand determine feasibility. The exercise establishes that aging alone does not uniquely determine the labor-hours outcome. It also identifies the scale of the assumed institutional or behavioral adjustment required to offset it.

## Participation and employment are separate margins

An older person who wants work but has not found it can enter the labor force without immediately adding to employed hours. Participation can therefore rise before employment rises. Conversely, a person withdrawing from search can lower measured unemployment without adding any work. The participation and employment rates should remain visible together.

For planning, a scenario that raises participation while holding unemployment and hours fixed should identify that assumption explicitly. It is a labor-supply scenario, not an estimate of how many people employers will hire. A projection of filled jobs additionally requires demand for the relevant skills and a mechanism matching workers with those positions.

An age-standardized comparison can help separate those margins. Apply each economy's age-specific employment and hours schedules to the same population weights. The standardized total asks how much the differing schedules contribute under a common demographic structure. Report the actual total as well, because it describes the real quantity of work available rather than the counterfactual composition.

Sex-specific analysis is equally important. Aggregating men and women can conceal different employment histories, care responsibilities and job opportunities. A policy that expands access for one group can improve the aggregate without changing another group's employment. The Japanese figures demonstrate why a single older-person rate omits relevant variation.

## Employment form and the use of capability

Retaining a worker in employment does not establish that the worker's prior skills continue to be used fully. A transition to a shorter-hour or different occupation can be beneficial, constrained or both. The relevant evidence includes the job's tasks, pay, security and the worker's preferences, rather than only the number employed.

For firms, the economic issue is often the preservation and transfer of capability. A flexible schedule may retain specialist knowledge and support training even when it adds fewer hours than a standard position. Alternatively, a poorly matched role can preserve a headcount while making little use of accumulated experience. Neither outcome can be inferred from the age distribution alone.

Pay systems influence this transition. If compensation, responsibilities and retirement practices are linked closely to career stage, continuing work may require a different contractual arrangement. Assessing the effect requires the actual rules and outcomes in the industry being studied. It should not be assumed that one national employment category has the same economic meaning as a similarly named category elsewhere.

Self-employment requires separate treatment as well. Business owners may remain active beyond the age when employees leave a career job. Their hours and income can be volatile, and the productivity of a small enterprise depends on its capital and customer base. Including them is essential to total labor input, while excluding them can be appropriate for a narrowly defined employee-wage comparison. The choice should follow the question.

## Labor time and productive output

A given quantity of hours can generate different output depending on capital, organization and task allocation. Replacing a departing worker's hours with a different schedule is therefore not an exact replacement of productive services. The relevant capabilities, equipment and opportunities to transfer knowledge matter alongside the time total.

An output projection should state separately whether it assumes unchanged productivity within each age group. Assigning an automatic productivity penalty to age would be unsupported without evidence about tasks and performance. Assigning unchanged output to every additional hour would also be an assumption. The hours decomposition identifies a labor-input channel and leaves the productive contribution to a separate, testable account.

## Hours are an input, not a welfare target

Maintaining total hours is not independently a desirable social outcome. Some older workers value additional work; others face financial pressure or health constraints. An increase in participation can reflect improved opportunities, insufficient retirement resources or a combination. The direction of the rate does not identify which explanation dominates.

A welfare assessment should examine disposable income, pension and other resources, work preferences, physical demands and access to suitable jobs. Reducing involuntary exclusion can improve welfare even if the aggregate labor effect is modest. Requiring more hours from people unable or unwilling to supply them would have a different interpretation.

The same distinction applies to fiscal projections. More employment can affect taxes, contributions and benefit payments, but the budget effect depends on pay and institutional rules. It is not equal to the additional hours multiplied by an unchanged average tax receipt. A fiscal estimate needs a separate, consistent income and policy account.

## Implications for the Japan–Korea comparison

The official evidence shows that later-life work in both economies must be assessed through more than a conventional retirement threshold. Japan's age- and sex-specific observations identify differences in employment form and working time. Korea's supplementary survey demonstrates substantial employment within a broad older population while preserving a distinction between a career job and current activity. These findings describe different dimensions of adjustment rather than a harmonized ranking.

The common benchmark is the age-by-sex contribution to actual hours, with the employment relationship and earnings analyzed alongside it. The hypothetical decomposition shows that shifts in participation and hours can materially modify the effect of aging, while requiring explicit assumptions about feasible work and demand.

For business and public policy, the relevant objective is to connect willing workers' capabilities with suitable production opportunities. Demographic pressure remains real, but its effect on labor input depends on the institutions through which people continue, change or leave work. A careful comparison identifies those margins and their limits before equating a larger older population with an inevitable proportional reduction in economic capacity.

---

American Economic Senate