The Selected Issues paper analyzes the health of the corporate sector, areas of vulnerability, and the effect of corporate restructuring in Japan. It reviews the deflation and associated economic problems of Japan, demonstrates the impact of fiscal policies on public debt, and estimates the expenditure and revenue adjustments needed to restrain the growth of debt. It also analyzes the issues in the Japanese labor market, structural developments, changes in the behavior of stakeholders, policy issues on employment insurance, and the social safety net.


The Selected Issues paper analyzes the health of the corporate sector, areas of vulnerability, and the effect of corporate restructuring in Japan. It reviews the deflation and associated economic problems of Japan, demonstrates the impact of fiscal policies on public debt, and estimates the expenditure and revenue adjustments needed to restrain the growth of debt. It also analyzes the issues in the Japanese labor market, structural developments, changes in the behavior of stakeholders, policy issues on employment insurance, and the social safety net.

VI. Employment Insurance and the Social Safety Net1

1. This chapter discusses issues associated with Japan’ s policies on employment insurance and the social safety net.2 The chapter is structured as follows. Section A outlines the benefits and financing of the employment insurance scheme. Section B compares social safety net expenditures in Japan with those in other G-7 countries, with a view to shedding light on whether Japan’ s scheme should be more generous. Section C examines active labor policies, which aim for example to reduce unemployment among young and old workers; and section D briefly discusses some policy options. One of the key implications of the analysis is that Japan’ s social safety net adequately protects the unemployed. Furthermore, streamlining some of its benefits and subsidies would be desirable. It is also suggested that the financing of the employment insurance scheme could be strengthened by increasing premiums if needed.

A. Benefits and Financing

2. The employment insurance account managed by the Japanese government has two subaccounts: unemployment benefits and the so-called “three services” (Box 1). The unemployment benefits account is used by the national government to pay benefits to workers, principally to job-seekers (benefits for job-seekers). In addition to the benefits for job-seekers, the government makes transfers to encourage the unemployed to look for jobs (benefits for employment promotion), to help workers improve their skills (benefits for training and education), and to support special categories of the unemployed such as old workers and those on child-care or home-care leave (benefits for continuing employment). The amount and duration of benefits are determined by objective criteria, and the government has no discretion to change them; for example, the basic allowance of benefits for job-seekers is determined by criteria such as the previous wage, years worked, age, and the reason for unemployment.

3. The three services account subsidizes employers, with the aim to prevent unemployment. The three services refer to programs that aim to stabilize the labor market, develop workers’ skills, and promote workers’ welfare. To stabilize the labor market, in particular, the government subsidizes firms to encourage them to retain workers. In addition, the three services account bears the cost of construction and maintenance of facilities for workers, including facilities for skill development and consultation services.3

4. The two subaccounts are financed by premium contributions and budget transfers. Premiums for unemployment benefits (1.4 percent of workers’ salaries, paid equally by both workers and firms) and for the three services (0.35 percent of workers’ salaries, paid solely by firms) provide most of the financing.4 The employment insurance premium is low in Japan compared with some other major countries; for example, the premium rate is around 6 percent in France, more than triple that in Japan.5 Budget transfers also contribute to funding, for instance, paying for 25 percent of the basic allowance of the benefit for job-seekers. According to the Ministry of Health, Labor, and Welfare (MHLW), among other advanced countries, only Germany has the same type of government contribution, although in Germany it funds only 15 percent of the benefit.

Employment Insurance System

(As of May 2003)

article image
Source: MHLW.

Benefit Amounts and Duration—Benefit amounts are set according to the job-seeker’ s age and previous wage. Benefits are equal to a certain proportion (50-80 percent) of the previous wage (the proportion is higher for lower wages) and subject to upper limits that depend on the age of the job-seeker: generally, the older the job-seekers, the higher the upper limit. Benefit duration is determined by the period the beneficiary contributed to the insurance system, age, the reason for unemployment, and the prospects of being rehired.

Daily Benefits

(50-80 percent of previous daily wages, 45–80 percent for those aged 60–64)

article image

Benefits Duration

(In days)

article image
Source: MHLW.

5. During the 1990s, rising unemployment and expanding benefits have led to deficits in these accounts and nearly depleted their accounts’ once-high reserves (Figures). High accumulated reserves and unpredicted higher unemployment in the 1990s led the government to introduce new subsidies and benefits. These new expenditures seem to have had limited effects on employment, but along with higher unemployment they caused a sharp fall in the systems’ reserves. Reserves of the unemployment benefits account declined from ¥4.8 trillion in FY1993 to only ¥0.2 trillion—less than one month’ s insurance payments—in FY2002. Also, reserves of the employment stability fund (part of the three services account) fell from a peak of ¥436 billion in FY1992 to ¥170 billion in FY2002.6 In response to the worsening of the balance in the former account, the premium rate was increased successively from 0.8 percent in the 1990s to 1.2 percent in 2001 and further to 1.4 percent in 2002, but these increases could not stop the decline in reserves.


Financing of Unemployment Benefits

Citation: IMF Staff Country Reports 2003, 282; 10.5089/9781451820560.002.A006


Financing of the Three Service

Citation: IMF Staff Country Reports 2003, 282; 10.5089/9781451820560.002.A006

6. Against this background, in 2002 the MHLW proposed to further increase the insurance premium for unemployment benefits from 1.4 percent to 1.6 percent starting in 2003. However, the government postponed the suggested increase until April 2005 owing to concerns about its possible adverse effects on private business, limiting the options to address any possible future deterioration in the financing of the employment insurance account. Although the government could increase the premium from 1.4 to 1.6 percent without consulting the Diet if the situation is urgent, the government must obtain the Diet’ s approval if it wants to increase the rate above 1.6 percent.

7. To contain demand for benefits from the employment insurance account and thus take pressure off its financing, the government recently created a new fund to encourage the unemployed to find jobs quickly. The new fund, which amounts to ¥250 billion and will exist until March 2005, pays special benefits to job-seekers if they find a job within a specified period. The scheme aims to reduce moral hazard and encourage job-seeking. Although the unemployment benefits account already has similar benefits, the new fund provides more preferential treatment. The government expects that job-seekers will apply for the new benefits rather than the ordinary unemployment benefits (benefits for employment promotion), and that accordingly, the insurance account can stay afloat for the next five years. However, future unemployment trends and the effects of the new fund on the employment insurance account are uncertain.

B. Assessment of Unemployment Benefits

8. This section considers whether Japanese employment insurance—along with other related social services—provides an adequate social safety net for the unemployed, and whether more generous payments should be made if unemployment increases further. There are three reasons to believe that Japan should maintain the current level of unemployment benefits, and that expanding benefits is not warranted at this moment.

9. First, expenditure on employment benefits in Japan is comparable to that in other G-7 countries relative to GDP, but generous per unemployed person. According to the OECD (2002), Japan spent 0.55 percent of GDP in FY2000 on unemployment compensation, less than in France, Germany, and Canada, the same as in the United Kingdom and Italy, and higher than that in the United States. Per unemployed person, however, compensation in Japan is among the highest in G-7 countries (Table 1 and Figures).7,8 In addition, the total expenditure of insurance payments would automatically respond to possible higher unemployment in the future, as the individual benefits are predetermined and government has no discretion to change them without Diet authorization. Moreover, other elements of Japan’ s social welfare system such as social assistance and public health insurance provide additional cushions for the unemployed (Box 2).


Unemployment Compensation per GDP

Citation: IMF Staff Country Reports 2003, 282; 10.5089/9781451820560.002.A006


Average Unemployment Compensation per Unemployed, 2000

Citation: IMF Staff Country Reports 2003, 282; 10.5089/9781451820560.002.A006

10. Second, excessively generous unemployment benefits would run the risk of creating moral hazard and increasing structurally-high unemployment in the medium to long term. A number of studies indicate that many advanced countries have suffered from this problem.9 Moreover, there are reasons to believe that moral hazard is a from problem also in Japan. In particular, many middle-age job-seekers earned wages that were higher than their productivity when they were employed, and thus their unemployment benefits are often higher than the wages they would receive if rehired. Therefore, they have small incentives to find jobs before the benefits payment period expires. According to surveys, quite a number of job-seekers find jobs within a month after the payment period lapses.10

Table 1.

Annual Public Expenditures in Labor Market Programs

(In percent of GDP)

article image
Source: OECD Employment Outlook 2002.
Table 2.

Minimum Unemployment Contribution Periods and Entitlement Duration

(Workers aged 40, not the first claim)

article image
Source: OECD Employment Outlook 2002.

Social Assistance and Health Insurance

Low income households can receive social assistance from the government. This assistance encompasses eight types of aid, including key aid for living expenses, medical fees, and housing fees. The government decides a minimum level of living expenses and provides an allowance that increases the household income to that minimum. The minimum level of living expense is about 70 percent of the consumption of the average household and varies depending on the number and age of people in the households and where they live. For example, a typical household of three people in Tokyo would receive assistance to reach a total income of ¥162,000 (equivalent to about US$1,400) per month, including the assistance and the other incomes. Since there is no maximum allowance, households without any other income can receive the same amount as the minimum living expenses (for example, ¥162,000). Social assistance is not taxable and has no duration limit. In 2001, the government spent over ¥2 trillion (0.4 percent of GDP) on assistance to low income households.1,2

In addition, the unemployed have access to the public health insurance system. Unemployed people are insured by the National Health Insurance (NHI) system, which is part of the public health insurance system and run by local governments. The NHI charges insurance premiums, but the fee depends on the household’ s ability to pay; thus, low income households can be exempted. For most health problems, those insured have access to all necessary medical treatment at any medical facility, with a 30 percent co-payment. The health insurance system is considered to provide proper medical care including to people in need, such as aged and low-income people; some consider that this has helped to keep the life expectancy of Japanese citizens above the level in any other country.

1 According to the expenditure assessment survey by the government in 2003, local governments raised concerns about moral hazard problems related to generous benefits. In this connection, a government council on social welfare under the MHLW recommended a further review of social assistance.2 Rising unemployment has increased the number of recipients of social assistance. In 2001, 1.1 million people (0.9 percent of population) received assistance, up from 0.9 million in 1995.

11. Third, job-seekers seem to be more concerned about finding jobs than about receiving higher unemployment benefits. People in Japan emphasize maintaining or increasing employment rather than improving unemployment benefits, presumably because (as often reported) most people regard unemployment as a social stigma. In a survey by RENGO, Japan’ s largest labor union confederation, even among job-seekers, more people asked the government to take measures to promote employment than to upgrade unemployment benefits.

C. Active Labor Policy and Policy Efficiency

12. Recent increases in unemployment have intensified the debate about active labor policies, such as promoting youth and elderly employment. In particular, promoting youth employment is under heated discussion, as noted in the previous chapter. However, the effectiveness of such measures in the past has been questioned. As is the case with passive labor policy such as unemployment benefits, expenditure per unemployed person on active labor measures is lower than in some European countries, but on the other hand, Japan spends more on active policy measures than do the United Kingdom, United States, or Canada (Figures).11


Public Expenditure on Active Labour Measures

(Per unemployed)

Citation: IMF Staff Country Reports 2003, 282; 10.5089/9781451820560.002.A006


Public Expenditure on Active Labour Measures

(Per unemployed)

Citation: IMF Staff Country Reports 2003, 282; 10.5089/9781451820560.002.A006

13. As in other countries, the experience in Japan implies that spending on active policies has not always been very effective in bolstering employment, particularly during recessions.12 In response to rising unemployment, the Japanese government has introduced a number of measures in recent years, especially under the three services account, to promote employment (Box 3). These measures included expanding existing subsidies and establishing new ones to generate and maintain employment in local areas and in small- and medium-sized enterprises, creating employment in business sectors that are growing, deregulating private job placement services, and promoting employment of old job-seekers. Despite these measures, unemployment has risen amid weakness in the economy.

14. Against this background, in 2000, the Diet asked the government to examine the effectiveness of subsidies under the three services account and to streamline and restructure them. In response to this request and related recommendations from a government council in 2002, the MHLW reduced the number of these subsidies from 46 to 35 by 2003 by abolishing and combining some subsidies. In addition, the council recommended a continuous review of these subsidies thereafter.

Recent Policy Initiatives (By March 2002)

Since the late 1990s, the government launched a number of policy initiatives to create and maintain employment. Below is a summary of these initiatives:

article image
Source: MHLW.

Budgeted amounts were for 15 months.

15. The government also began to cut back some measures amid tight fiscal constraints. In the past, ample reserves (reflecting then-favorable employment conditions and limited flexibility in using these funds) loosened the budget constraint in the three services account.13 Now, however, with the unemployment rate at near record highs and a pressing need to curtail expenditure, measures under both subaccounts are being selectively reduced. For example:

  • Training and education benefit. The benefit was created under the unemployment benefits account in 1998 and subsidized both employed and unemployed workers who complete eligible private courses to develop their work skills. The subsidy was generous: 80 percent of expenditure was covered, up to a limit of ¥300,000 (equivalent to about US$2,500). Over 20,000 education and training courses were approved as eligible, even though some of them were just for beginners.14 During the first three years, 700,000 people received benefits, and the benefit payments amounted to ¥80 billion (equivalent to about US$700 million). There was criticism, however, that some of the courses were not effective in enhancing workers’ skills and that the high subsidy rate attracted students who were not sufficiently motivated to take advantage of the courses. The MHLW recently decreased the subsidy rate and its upper limit, and removed some courses from the list of those that qualify for the subsidy.15

  • Employee welfare services. As a part of the three services, a government affiliated body has built and run facilities such as hotels and gymnasiums to improve the welfare of workers, receiving over ¥1.5 trillion (0.3 percent of GDP) from the employment insurance account to build these facilities and additional subsidies to operate them. The government has decided to dispose of these facilities by transferring them to local governments or closing them by FY2005. However, progress in disposal has been slow and some facilities have been sold at very low prices.

16. The creation of the training and education benefit in 1998 represented a gradual shift of the government spending from firms to workers.16 The government often subsidizes firms to provide courses to increase the skills of their workers belonging to the firms. Under this arrangement, firms usually decide which workers are allowed to attend these courses, and the topics are limited to those relevant to the firms and specific sectors. This would not help workers find new jobs in different industrial sectors, even though labor movement from unproductive sectors to productive sectors is desirable. If, alternatively, the subsidies were provided directly to workers, they would have the opportunity to choose among a wider set of courses, including some that may be relevant to finding a job in a different firm, or even in a different sector.

17. The experience so far suggests that the assessment of active labor policy measures’ effectiveness can be improved. These measures are now under the government comprehensive policy assessment system, introduced in FY2002, but the quality of assessment of these measures has been unsatisfactory. For example, the assessment of training and education benefits described the use of the funds but did not include an evaluation of their effect on employment. In addition, the assessment did not include useful recommendations for reforming existing policy and formulating new policies. For example, the decision to downgrade the subsidy in the training and education benefit was not based on the previous policy assessment, as this assessment did not evaluate the level of the subsidy. Better-designed assessments could provide guidance on where expenditures might be targeted and whether, for example, improving the counseling function, and making greater use of private expertise including in lectures at public institutions to build workers’ job skills, could be beneficial compared with the cost.

D. Policy Appraisal

18. The government has taken reform steps including financial measures and streamlining subsidies, but more needs to be done given current tight fiscal constraints. In particular, more focused and effective spending would be useful. In this connection, six main conclusions may be drawn.

19. First, the current level of the core unemployment benefit (the basic allowance of the benefits for job-seekers) seems appropriate and should be maintained. Compensation per unemployed person is high by G-7 standards, so an increase in benefits per person would seem difficult to justify. However, if the number of unemployed persons increased, it would be reasonable to allow total employment insurance expenditure to increase, to avoid adverse effects on the economy.

20. Second, it would be useful to further review non-core benefits disbursed under the unemployment benefits account. Spending could be more effective if it were focused on benefits for job-seekers. Among non-core benefits, those for training and education, which are still generous after the reform, may warrant further review, in order to increase effects on employment and to reduce possible moral hazard.

21. Third, it would be useful to further streamline expenditures under the three services account, based on strict policy assessments. More decisive steps to dispose of facilities to improve employee welfare, at fair prices, would be useful. Also, education and training at public institution to build workers’ job skills can be improved through more active use of private sector expertise. In addition, further efforts to assess the impact of such policies, and use these assessments to direct further reforms, would be warranted.

22. Fourth, if necessary the unemployment insurance premium could be increased to 1.6 percent to bolster the financing of the system. If unemployment increases further, it would be useful to request Diet authorization to increase the rate beyond 1.6 percent, given that the current premium is relatively low.

23. Fifth, more should be spent on the workers rather than on the firms. The suggested shift in spending from firms to workers could increase labor mobility and help to decrease mismatches in the labor market.

24. Sixth, the government could consider consolidating the unemployment benefits account with the three services account to increase flexibility in using these funds. As explained, money in one account cannot be used for expenditures under the other account in principle, even when the former has ample funds and the latter faces a shortage. However, given the tight budget constraint, money saved by streamlining subsidies in the three services account should be available for the financing of the benefits for job-seekers. The law governing employment insurance, which was amended in 2003, enables the government to transfer the funds in the three services account into the unemployment benefit account. This change was a welcome step, but even under the amended law, the government has to repay the transfer at some point in the future.


  • Blanchard, Olivier, 2000, “The Economics of Unemployment: Shocks, Institutions, and Interactions,” Lionel Robbins Lectures 1-3 (unpublished; London: London School of Economics). Available via the Internet:

    • Search Google Scholar
    • Export Citation
  • Blanchard, Olivier, and Thomas Philippon, 2003, “The Decline of Rents, and the Rise and Fall of European Unemployment,” (unpublished) Available via the Internet:

    • Search Google Scholar
    • Export Citation
  • Higuchi, Yoshio, 2001, “Koyou to Shitsugyo no Keizaigaku (The Economics of Employment and Unemployment),” (in Japanese).

  • International Monetary Fund, 2003, “Unemployment and Labor Market Institutions: Why Reforms Pay Off.” World Economic Outlook, April 2003, pp 129 –150, IMF.

    • Search Google Scholar
    • Export Citation
  • Japan Institute of Labor, 2002, “The Labor Situation in Japan 2002/2003.”

  • Japanese Trade Union Confederation (RENGO), 2002, “RENGO White Paper, 2003.”

  • (Former) Ministry of Health and Welfare, 1999, “Kousei Hakusho, Heisei 11 nen ban (White Paper on Health and Welfare, 1999),” (in Japanese)

    • Search Google Scholar
    • Export Citation
  • Ministry of Health, Labor, and Welfare (MHLW), 2002, “Kousei Roudou Hakusho (Heisei 14 nen ban) (White Paper on Welfare and Labor 2002),” (in Japanese).

    • Search Google Scholar
    • Export Citation
  • Organization for Economic Co-operation and Development (OECD), 1993, “Active Labour Market Policies: Assessing Macroeconomic and Microeconomic Effects, “Employment Outlook.

    • Search Google Scholar
    • Export Citation
  • Organization for Economic Co-operation and Development (OECD), 2002, “OECD Employment Outlook July 2002.”

  • Roudou Seisaku Shingikai Shokugyou Antei Bunkakai Koyou Hoken Bukai (Employment Insurance Team, Labor Stabilization Sub-committee, Labor Policy Council), 2002, Koyou Hoken Seido no Minaoshi ni Tsuite (Review on the Employment Insurance System),” (in Japanese).

    • Search Google Scholar
    • Export Citation
  • Shakai Hoshou Shingikai (Social Welfare Council), 2003, “Kongo no Shakai Hoshou Kaikaku no Houkousei ni Kansuru Iken (Opinions on Social Welfare Reform in the Future),” (in Japanese).

    • Search Google Scholar
    • Export Citation

Prepared by Takuo Komori (ext. 37613).


For related analysis, see the Selected Issues chapter entitled “Structural Changes in Japan’ s Labor Market.”


Formerly, the three services account also bore the cost of construction and maintenance of lodging facilities.


The financing of the two subaccounts is different, presumably because the recipients are different: premiums paid solely by firms (in the three services account) are spent on firms. However, for example, in the United States, premiums, which are solely paid by firms, are used for benefits for unemployed workers. In many other countries, part of the subsidy are given to workers rather than the firms, and premiums paid by firms are higher.


The premium rate was 6.2 percent in France and 5.4 percent in Sweden in 1998 (White Paper on Health and Welfare (1999) by (former) Ministry of Health and Welfare). In these countries, most of the premium is paid by firms, which is in contrast to Japan where the premium is paid almost equally by firms and employees.


The fund is annexed to the account, pooling both the account surplus and the budget contribution.


Unemployment compensation per unemployed person was calculated, in order to adjust for the difference in unemployment rates and GDP, using total unemployment as the denominator. Denominators include unemployed people who do not receive compensation, including (i) job-seekers such as new graduates who were not eligible to receive benefits because they had not paid premiums, (ii) those whose benefits had expired, and (iii) those who simply did not apply for benefits. In Japan, the duration of benefits is shorter than in other advanced countries, so the proportion of those people with expired benefits is likely to be higher (Table 2).


Difference in ages of job-seekers and their previous wages across countries may have some effects. If all countries provide identical benefits to workers, a country that has more middle-aged job-seekers whose previous wages were high would have higher average compensation than other countries. Japan might be such a country. However, it seems unlikely that Japan has so many middle aged job seekers that it would change the thrust of the result.


For example, see IMF “Unemployment and Labor Market Institutions: Why Reforms Pay Off (2003). This study demonstrates that high unemployment is largely structural in nature, and provides evidence on the linkages between structural employment and the institutional features of labor markets, such as generous unemployment insurance.


Employment Insurance Team, Labor Stabilization Sub-committee, Labor Policy Council (2002).


Most of the subsidies under the three services account are for active labor policy. In addition, some expenditures under the general account are also for active labor policy.


OECD (1993) discusses the advanced countries’ experience including those in other countries.


As discussed, only firms are recipients of these subsidies in principle. Moreover, the money in the three services account cannot be transferred to the unemployment benefits account even when the latter suffers a shortage of money, and vice versa.


Eligible courses include those regarding English conversation and computer skills.


The MHLW decreased the subsidy rate from 80 percent to 40 percent and its upper limit from ¥300,000 to ¥200,000. The change came into effect in May 2003.

Japan: Selected Issues
Author: International Monetary Fund