ANNEX I: History of Japan’s Main Public Pension Systems
1. The first pension system with wide coverage, the Employee Pension Insurance (EPI) system, was introduced in 1942. It was mandatory for all private sector employees and initially the contribution rates and benefits in percent of salary were relatively low. Benefits were increased significantly in 1965 and again in 1973. In 1961, the pension system coverage was broadened to include all people of working age. A flat rate contribution, flat rate benefit system—the National Pension—was introduced for everybody who was not covered under the EPI. The generosity of benefits increased over time for this system as well. Indexation of benefits to gross wages was introduced for both the EPI and the NP in 1973. Since 1986, the two systems were integrated financially—the beneficiaries of the EPI started receiving a flat portion benefit (under the National Pension) in addition to the wage-related component. Eligibility to receive the flat-rate NP benefit was also extended to nonworking spouses of EPI participants (although they do not make contributions to the system).
2. As a result of the increasing generosity of benefits, public pension income became the main source of income for many retirees (replacing in importance the traditional lump-sum retirement benefits given by employers). The replacement rate for a “typical” pension recipient (married with nonworking wife) was above 60 percent of salary in the 1980s.
3. In the 1980s, population growth rates slowed down and economic growth rates also slowed starting in the early 1990s. The last three reforms focused on restoring the long-term financial stability of the pension system through a combination of raising contribution rates and reducing the generosity of benefits:
The 1986 pension reform reduced the accrual rate for the earnings-related component from 1 percent per year to 0.75 percent per year. The number of contribution years required to receive a full NP benefit increased from 25 to 40;
The 1994 pension legislation increased the eligibility age for the flat-rate portion of the EPI benefits from 60 to 65, to be gradually phased in between 2001 and 2013 for men and 2006 and 2018 for women. In addition, benefits indexation was changed from gross to net wages;
The 1999/2000 pension reform introduced a number of significant changes. The accrual rate for the wage-related portion of the EPI was reduced from 0.75 percent to 0.71 percent. The indexation of both the flat-rate and the earnings-related benefit increases were changed to CPI from wages. A new earnings test was introduced that reduced the pension benefits of those still employed in their late 60s. The age at which people start receiving the wage-related portion of the EPI benefit was increased to 65, to be phased in between 2013 and 2025 for men and between 2018 and 2030 for women.
However, these reforms failed to restore financial sustainability of the system, since with each subsequent update of the population projections, it became clear that the demographic shock would be much worse than previously expected.
Alesina, A., Perotti, R. and Tavares, J., 1998, “The Political Economy of Fiscal Adjustments,” Brookings Papers on Economic Activity.
Chand, S. and Jaeger, A., 1999, “Reform Options for Pay-As-You-Go Public Pension Systems,” World Bank, Social Protection Discussion Paper Series, no. 9927, December.
Dekle, R., 2003, “Financing Consumption in an Aging Japan: The Role of Foreign Capital Inflows and Immigration,” mimeo, USC, Los Angeles, November.
Faruqee, H. and Mühleisen, M., 2001, “Population Aging in Japan: Demographic Shock and Fiscal Sustainability”, IMF Working Paper 01/40.
National Institute of Population and Social Security Research, 2002, “Population Projections for Japan: 2001–2100,” Tokyo, Japan, January.
Takayama, N., 1998, “The Morning After in Japan: Its Declining Population, Too Generous Pensions and a Weakened Economy,” Tokyo: Maruzen Co. Ltd.
Takayama, N., 2003, “The Japanese Public Pension System: What Went Wrong and What Reform Measures We Have,” Working Paper, Hitotsubashi University, October. http://www.ier.hit-u.ac.jp/~takayama
Prepared by Dora Iakova (ext. 35365). Robert Gillingham constructed the long-term pension finances forecasting model for this paper.
Since the EPI contribution is a fixed percent of salary independent of family status, the system effectively results in higher benefits for those with dependent spouses and those with low salaries.
The reduction will be achieved through a so-called “macroeconomic indexation.” A macroeconomic index will be defined every year, reflecting the decline in the number of pension system contributors and the increase of life expectancy. MHLW estimates that the average annual change of that index will be about 0.9 percent. The initial level of benefits for new retirees will be a function of their average lifetime wages adjusted for the index, and benefits for existing retirees will increase by CPI inflation minus the change in the macroeconomic index (if that is negative, benefits will remain flat). The indexation will be discontinued in 2025—by that time the replacement rate is expected to be down to 50.2 percent. On a technical note, any temporary increases in the CPI due to upward adjustment in the consumption tax rate ought to be excluded from the index for pension benefit adjustment, although the current law does not specify this.
In fact, the 2002 projection is already proving to be too optimistic—the actual birth rate in 2003 was 1.29, while the projected rate was 1.32.
In addition to rising pension costs, the aging of the population will lead in a sharp rise in age-related medical and elderly care expenditure as well. The projected increase in medical and elderly care related expenditure over the next 20 years alone is 4 percent of GDP. The portion of medical costs financed by general taxes revenue will rise significantly as the population ages, since the government finances a large share of the medical care provided to the elderly. Estimates by the MHLW suggest that public subsidies as a share of medical expenditure will rise from about 27 percent in 2002 to 36 percent by year 2025.
Currently more than 37 percent of “type 1” insured do not make the required NP contributions. Although participation is compulsory, enforcement is rather weak. In addition, part-time workers frequently limit their working hours to stay below the EPI participation threshold (currently 30 hours/week).
Takayama (1998) finds that the average disposable per-capita income of retirees after redistribution is higher than the average income of middle-aged workers. This is at least partly due to unequal tax treatment of retirees and active workers. Pension recipients enjoy large pension tax exemptions, in addition to age-related deductions. The tax-free income for pension recipients is almost twice as large as the tax-free income for employees. The 2004 tax reform has partially addressed this problem: the age-related deduction has been abolished and the pension deduction has been reduced.
See Chand and Jaeger (1999) for an overview of different reform options and discussion of the benefits and drawbacks of each option.
The present reform will use an indexation mechanism similar to that used in NDC systems to gradually reduce benefits, but it will stop once the desired benefits floor has been reached.
Proponents of pension privatization point to the fact that historically, market returns have been on the average much above the implicit return of PAYG systems. However, the risks of a private pension system and the administrative costs are also much higher. Nonetheless, private retirement savings are a useful complement to a less generous public pension system.