Economic Analysis Series No.155
Analysis of the Utilization of Older People's Human
Capital in the Labor Market of the Aging Society

October, 1998
(ERI, EPA)
Atsushi Seike (Senior Visiting Fellow)
Hitoshi Hayami (Visiting Fellow)
Masahiro Abe (Visiting Fellow)
Masahiko Tsutsumi
Atsuhiro Yamada
Osamu Ichinose
Masahito Nakajima

(Introduction)

[I] Research Purpose

To promote employment of older people is one of the most important measures to maintain social vitality of Japan in the coming 21 century. It means to have older people who are ever increasing in number actively participate in supporting the society as long as possible. It is vital to construct a system where older people who are capable and willing to work can play an active role.

As the society is aging, firms are also increasingly competing in a sophisticated manner. What is especially significant in this sense is to utilize older people with advanced working capacity (human capital). The current system, however, is not necessarily structured to fully utilize the human capital of older people. Systems and practices that prohibit utilization of older people with advanced working ability must immediately be improved.

In this study, econometric analysis using micro data will be conducted to find out how systems and practices prohibit effective utilization of older people' s human capital, based on such awareness of problems as above. Similar econometric analysis will be executed to find out how employment of older people will be affected by changes in the forms of business organization resulting from such factors as deregulation. From results of such empirical analyses, the authors intend to introduce policy implications that provide clues for the construction of a system that can utilize abilities of older people who are willing to work and are capable of working.

[II] Structure of Research Project

We will conduct four studies to accomplish the research purposes stated above. The first study is to analyze to what degree such public and quasi-public systems as public pension and mandatory retirement interfere with the utilization of older people's abilities. Specifically, how much human capital of older people is lost from the labor market through such systems will be measured.

The second study deals with the loss of human capital via the labor turnover of mid-age and older age groups. Young population which has been a pool of the inter-industry labor force redistribution will decrease in the coming years. Accordingly, the labor turnover of mid-age and older age groups will inevitably increase. The loss of human resources in the movement, therefore, will become a serious problem. We will specifically estimate how the probability of unemployment and successful turnover changes in accordance with age and the length of service.

While the first study focuses on the utilization of older people's human resources after the present mandatory retirement age, the second study emphasizes the utilization of human resources of the middle-aged and aged before the present mandatory retirement age. In the third study, human capital loss will be observed in terms of movement process, that is, the period leading up to the completion of the movement, while the second study deals with human capital loss in terms of labor turnover probability. Relations between the labor turnover and industry, duration of the inter- and intra-industry movements (unemployment), and duration of the movement (unemployment) based on such labor force descriptions, with age and tenure will be specifically measured.

Lastly in the fourth study, we will change our point of view from the preceding studies I to III, and investigate how changes in the form of business enterprise as a result of deregulation have an impact on the utilization of older people's human capital. We will exemplify the commercial sector where scale expansion, synthesization and diversification will advance by the deregulation of the Large-Scale Retail Store Law, and estimate how labor demand by descriptions such as age and gender changes in accordance with the expansion of firm size and diversification. Summary of the studies I to IV are described in the following.

[III] Effect of Mandatory Retirement and Public Pension Systems on Loss of Human Capital

The mandatory retirement system and public pension system not only promote retirement of older people, but are also closely related to high working ability (human capital) of older people. Because older people with higher working ability (human capital) tend to have mandatory retirement experience and to receive pensions of a large amount, the mandatory retirement and public pension systems tend to hinder utilization of the human capital of older people. Based on such viewpoint, we will examine the following three working hypotheses.
[Working hypothesis 1]  Retired people with public pension and mandatory retirement experience have higher human capital; therefore there is more potential of human capital loss from the labor market, which is associated with public pension coverage and mandatory retirement systems.
[Working hypothesis 2]  Public pension and the mandatory retirement experience reduce the degree of human capital utilization of older working people.
[Working hypothesis 3]  Public pension and mandatory retirement experience reduce the degree of human capital utilization in terms of reducing the return of human capital investment.

  • These hypotheses are basically supported by the following findings:
  • [1]  The distribution of imputed wages for retired men with public pension or with mandatory retirement experience is higher than for working men without public pension or mandatory retirement experience. (Working hypothesis 1)
  • [2]  The coefficients of human capital variables for explaining the probability of having public pension benefit or having mandatory retirement experience among retired people are positively estimated. (Working hypothesis 1)
  • [3]  The distributions of hours worked for working men with public pension or with mandatory retirement are shorter than those without public pension or mandatory retirement experience, and the proportion of workers who are working in the same occupation as they were at age 55 are lower among workers with public pension or with mandatory retirement experience. (Working hypothesis 2)
  • [4]  The coefficients of the dummy variables for having public pension or mandatory retirement experience are negatively estimated in the equations of hours worked and in the probability of having the same occupation as they did at age 55. (Working hypothesis 2)
  • [5]  The coefficients of the dummy variables for having public pension or mandatory retirement experience in the switch function of switching regression model are positively estimated, and the coefficients of human capital variables in the secondary market wage equation are smaller than those in the primary market wage equation. (Working hypothesis 3)

 Policy implications obtained from the above empirical findings are as follows. As for the public pension system, the current systems of employee's welfare pension and mutual pension are structured so that the scope of application and level of benefit are determined to be linked with professional career before retirement, that is, with the condition of human capital accumulation. What must be done is to cut the link and to not have utilization of high human capital unwittingly interfered with the public pension system.

 Accordingly, it is necessary to clearly identify the distinction between the parts proportionate to remuneration and to the basic pension of public pensions. Financially, pay-as-you-go system is to be adopted to cover the part of basic pension for older people with the young generation shouldering the pension tax burden. The minimum necessary income in old age is to be covered basic pension in part. As for the part proportionate to remuneration, the premium annuity based on a fully funded system is to be adopted. For the part other than a fixed amount of mandatory installment to prevent reverse selection, individuals are permitted to decide on the amount of installment. Older people who wish not to depend on pension as much as possible and to maintain comfortable living by being employed or running their own business can set the amount of installment for the part proportionate to remuneration as little as possible and to appropriate the remaining amount to invest in their own capabilities or to become independently employed so that they can increase their ability to earn in old age.

 In Japan where aging is progressing most rapidly among the industrialized countries, where the average life expectancy is the longest, and where work motivation of the aged is the highest, the mandatory retirement age of 60 years is just too early. It is necessary to either abolish the mandatory retirement age or raise it to at least 65 years by the year 2013 when the age to start receiving public pension benefit turns to 65. This does not mean that it guarantees employment up to that age, but rather, it encourages making it a rule not to implement mandatory retirement because of age up to the age 65.

[IV] Age, Tenure, and Loss of Human Capital
  • Here the following two working hypotheses are to be verified.
  • [Working hypothesis 1]:  The more techniques and skills workers have accumulated in a firm, the greater the tendency for the workers to avoid turnover. In general the length of service that has a high correlation with techniques and skills acquired within a firm and the probability of successful turnover have a negative correlation.
  • [Working hypothesis 2]:  The probability of successful turnover has a positive correlation with general usefulness of techniques and skills accumulated until then. The length of employment correlated with firm-specific techniques and skills that have little generality and the probability of successful turnover have a negative correlation.

 In the model, based on the viewpoints of the labor supply side, probabilities of quittance (turnover) and successful turnover are measured in an estimation equation explained by human capital variables represented in the length of service and age. Specifically, probabilities of quittance (turnover) and successful turnover are estimated by the Probit model, using micro data of the "Employment Status Survey" conducted by the Management and Coordination Agency.

  • The following analytical results have been obtained.
  • [1]  The probability of quittance (turnover) is lower for those who are older and have longer years of service.
  • [2]  The probability of successful turnover is lower for those who are older and have longer years of service.

 Policy implications denoted from the above are as follows. First, in the future when the mid-aged and older workers with longer years of service and with lower probabilities of quittance and turnover are increasing, inter-industrial labor turnover will likely decrease, which may become a factor to interfere with labor force redistribution functions. Second, the probability of unemployment will increase in the future when the middle-aged, aged and workers with longer years of service with a low probability rate of successful turnover increase. Third, a low probability of successful turnover of workers with long years of service that possess a large human capital accumulation may link to a decline in the potential growth rate. To avoid such problems, measures to promote a smooth re- distribution of labor force by "half-way" workers such as expanding employment exchange services and ability redevelopment projects will be needed.

[V] Labor Turnover, Unemployment Spell and Attributes of the Worker
  • The following two working hypotheses are to be investigated.
  • [Working hypothesis 1]:  Accumulation of special skills and expertise specific to the industry to which workers belong contributes to extend the unemployment spell-accompanying turnover to a different industry.
  • [Working hypothesis 2]:  Attributes of the labor force affect the length of unemployment spell.

 Such working hypotheses are to be verified on unemployment spell by econometric analysis. Micro data of the "Special Survey of the Labor Force Survey" conducted by the Management and Coordination Agency is adopted, which investigates the unemployment spell of individuals.

  •  Analytical results are stated as follows.
  • [1]  Labor turnover or movement of labor in most cases occurs within the same industry. Even when inter-industry movement occurs, it tends to be a movement between industries with strong relations such as from precision machinery to electric machinery. Although movements from other industries to the construction industry are prevalent, movements from the construction industry to other industries are not many. The construction industry has a characteristic that is easy to enter but difficult to exit.
  • [2]  While 85 percent of workers is re-hired within a year in the intra-industry movement, the re- employment rate stays at 20 percent for the inter-industry movement. When a worker chooses an inter-industry movement, the probability of unemployment over one year increases by 49 percent. We can state that the inter-industry movement extends the unemployment spell without question.
  • [3]  As for the relationship between unemployment spell and age, the intra-industry movement becomes shorter in accordance with age, but longer for workers in the 50s or older. On the other hand, the inter-industry movement consistently becomes longer in tandem with aging, and aging by one year increases the probability of unemployment over one year by 2.8 percent. As one's age becomes higher, the impact of extended unemployment spell caused by labor turnovers to different industries becomes more severe.
  • [4]  Differences in unemployment spell depending on labor force descriptions demonstrate that the unemployment probability of over one year becomes higher for female than male, for university graduates in terms of educational background, for voluntary quittance, and for the married.

The above analytical results indicate the following policy implications. First, because cross- industrial labor turnovers resulting from changes in the industrial structure exacerbate human capital loss in the form of extended unemployment spell, policies are required to improve information and ability redevelopment functions of the labor market to deal with the situation. Second, the fact that the age factor over which workers have no option has a positive correlation with the extended unemployment spell means that as long as the employment system dependent on age is maintained, social losses accompanying unemployment will increase. To evade such a situation, the employment system must be reformed to become more age-neutral.

[VI] Changes in Corporate Form and Employment of the Elderly

In this analysis, labor demand functions by gender and age are estimated. Its purpose is to confirm what description of labor force will be demanded in accordance with changes in the form of business organization such as the size of business and diversification. Demand trends of the aged can be attained from this.

The first characteristic of the analysis is that a micro data set by commercial establishment has been prepared with the "Establishment Census" of the Management and Coordination Agency as a base, which is matched with the "Census of Commerce" of the Ministry of International Trade and industry and the "Basic Survey on Wage Structure" of the Ministry of Labor. Using this data, labor demand functions by labor force description is measured, and fixedness (volume to be employed at zero production) and elasticity (rate of variability of employment in accordance with fluctuations in production scale) of labor have been investigated by parameter constellation and labor demand share by description.

  • The findings of the measurement are compiled as follows.
  • [1]  In general, older workers cannot be called especially fixed in terms of labor fixedness.
  • [2]  In the retail industry, the proportion of young female workers increases in tandem with the expansion of production scale, and the proportion of older male workers increases in tandem with business diversification.
  • [3]  In the wholesale industry, the proportion of young workers increases in tandem with the expansion of production scale, and the proportion of male workers increases along with business diversification.

Taking the commercial sector as an example, we have demonstrated that the kind of effect changes in business management environment such as deregulation and market liberalization would bring to the employment of the aged. If business establishments should expand their size of business as a result of deregulation concerning large-scale retail stores, we consider that labor demand may change in a direction to increase the number of female workers in the commercial sector. If, however, business diversification should also be promoted because of deregulation, labor demand for older male workers is considered to increase in a relative term.


Structure of the whole text(PDF-Format 6files)

  1. page3
    Introduction別ウィンドウで開きます。 (PDF-Format 447 KB)
  1. (First Study)
  2. The Impact of Mandatory Retirement and The Public
    Pension System on Human Capital Loss
    1. page9
      Abstract
    2. page9
      1. Introduction
    3. page11
      2. Overview and A Descriptive Analysis of Retirement Patterns of Japanese Elderly
    4. page13
      3. Design of Experiment
    5. page16
      4. Empirical Results
    6. page20
      5. Discussion
    7. page21
      6. Conclusion

  1. (Second Study)
    1. page47
      Abstract
    2. page47
      1. Introduction
    3. page49
      2. Labor Supply
    4. page55
      3. Empirical Results
    5. page62
      4. Summary and Remained Issues
    6. page79
      appendix別ウィンドウで開きます。 (PDF-Format 264 KB)

  1. (Third Study)
    1. page85
      Abstract
    2. page86
      1. Introduction
    3. page87
      2. Prior Analyses on Unemployment Spell
    4. page89
      3. Statistics in Use and their Characteristics
    5. page90
      4. Analysis of Movement
    6. page91
      5. Analysis of Unemployment Spell
    7. page95
      6. Econometric Analysis
    8. page97
      7. Summary of Analysis
    9. page113
      figures and appendix別ウィンドウで開きます。 (PDF-Format 212 KB)

  1. (Fourth Study)
    1. page123
      Abstract
    2. page123
      1. Introduction
    3. page125
      2. Data Construction
    4. page127
      3. Concept and Formulation of Labor Demand
    5. page134
      4. New Formulation of Labor Demand by Age and by Gender
    6. page146
      5. Estimated Results
    7. page153
      6. Phase Transition of Labor Demand
    8. page154
      7. Conclusion
  • 1-6-1 Nagata-cho, Chiyoda-ku, Tokyo 100-8914, Japan.
    Tel: +81-3-5253-2111