ERI Discussion Paper Series No.71
On Aging and Employment in Japan
AgeTenureand Loss of Human Capital

March 1997
Masahiro Abe
(Central Research Institute of Electric Power Industry;
Visiting Researcher, Economic Research Institute, Economic Planning Agency)
Masahiko Tsutsumi
(Economic Research Institute, Economic Planning Agency)


In a few decades, more than 40 % of the labor force will be shared by the generation aged 55 and over. It inevitably affects the dynamics of every part of the economy, including balancing government budgets, the social security system, and the labor market. These institutions or market structures are usually established to accompany the historical trends of exogenous factors and environments. In that sense, a structural change in demography simultaneously demands a reconstruction in these fields.

A reallocation of labor force has been carried on by the younger generation in Japan. Applying the thought of human capital theory, this is because younger generation does not accumulate the firm-specific capital so much that the loss caused by turnover is negligible. However, it will be a relatively small portion in aged society.

On the contrary, the elder generation, especially with long tenure, holds huge firm-specific human capital. It means that the flexibility or mobility of elder labor force will be less than that of young. Data proves that the labor participation ratio of elder worker is low, and they easily leave the labor market. It suggests that future growth potential will be at stake due to decline of total labor supply and increase of the mismatch in the labor market without reallocating them promptly.

In recognition above, this paper attempts to answer the questions:

  1. How much does the age affect the turnover rate and job-quit tendencies of laborers?
  2. How much does the tenure, a proxy of firm-specific human capital, affect the probability above?
  3. What do we prospect for the future labor market with empirical evidence?
  4. How do we understand the effect of the tenure, as a proxy of firm-specific human capital?
  5. What are the policy options to be implemented?

In our empirical analysis, we used the micro data from "Employment Status Survey in 1992" and estimated the Probit model. Theoretically, estimated functions are derivatives of utility function, consisting of income and leisure. Three stage probabilistic nests, quit, participation and turnover are designed. Main findings from these estimations are:

  1. A tenure has a negative effect on all three probabilistic nests with controlling other attributes of age, and education level.
  2. It is difficult to identify the negative effects to participation. Because the existence of the retirement allowance may discourage the workers to stay in the labor force. The retirement allowance in Japan has strong links to the tenure. As long as the tenure brings high allowance, it is not necessary to stay in the labor market.
  3. Effects on the quit and turnover stage are regarded to come from accumulation of human capital. Long tenure promotes accumulation of firm specific human capital, that is not productive in other firms. Therefore, it discourages quitting and increases the mismatch between the talent of labor and the needs of firms.

Based on these estimation, we conducted some indicative simulations. The expected probability of falling into unemployment pool is only 0.004 % for the worker, whose attributes are, large firm, with family, university graduates, aged 50 and tenure 25. This is mainly due to the low probability of quit. However, once he quits, the expected probability of reemployed is less than 50 %.

Implemented policies are surveyed in final chapter. Generally, these policies are in line with the recognition that the general human capital/skill will contribute to reduce the frictional unemployment, and increase the probability of turnover. Although the recognition and policies are correct, there is a insoluble problem incurred from the nature of the firm. "Product differentiation" of the firms inevitably require "Process and skill differentiation" generated by the "Firm-specific human capital".

Structure of the whole text

  1. page1
    I. Introduction別ウィンドウで開きます。(PDF-Format 159 KB)
  2. page2
    II. Features of Aging, Labor Market
    1. page2
      A. Aging Society
    2. page3
      B. Labor Supply
    3. page4
      C. Market Adjustment
    4. page5
      D. Unempolyment
  3. page8
    III. Empirical Analysis of the Quit and Turnover
    1. page8
      A. Model
    2. page10
      B. Data
    3. page11
      C. Empirical Results
    4. page14
      D. Simulations
  4. page18
    IV. Policy Issues
  5. page20
    V. Reference
  6. page22
    VI. Appendix
  7. page24
    Tables and Figures
    1. [1]別ウィンドウで開きます。(PDF-Format 1.0 MB) [2]別ウィンドウで開きます。(PDF-Format 1.0 MB) [3]別ウィンドウで開きます。(PDF-Format 1.0 MB) [4]別ウィンドウで開きます。(PDF-Format 644 KB)
  • 1-6-1 Nagata-cho, Chiyoda-ku, Tokyo 100-8914, Japan.
    Tel: +81-3-5253-2111