6 Comments on “Medium-Term Strategies for Long-Term Goals”

Keimeir Kaizuka, and Anne Krueger
Published Date:
July 2006
  • ShareShare
Show Summary Details

Comments by Hiromitsu Ishi

This is a very interesting chapter that is likely to be of great help to us in designing reform policies. The discussion is well developed and the focus on the pension and health care systems is of immediate relevance for Japan. Overall, the chapter is very informative and insightful.

However, the United States and Japan have quite different social security systems with different levels of benefits and contributions. Until now, Japan has developed its own full-scale system of pension, health care, and nursing care services, while the U.S. has relied upon more self-insurance and private initiatives for welfare activities. The basic philosophy behind the welfare programs is fundamentally different between two countries. I wonder what lessons we could learn from the U.S. experience? For example, how does Dr. Aaron assess the nursing care system in Japan in the face of pressures from population aging that the U.S. doesn’t yet experience? Is the Japanese approach good policy or not?

Let me raise several additional questions for discussion:

First, many observers think that the 2004 pension reform has not succeeded in regaining the sustainability of the scheme and that in the future the pension system will have to be changed more fundamentally. Should a two-tier system consisting of basic and income-related pension benefits be maintained or should it be scrapped in a move toward a completely different system, like the Swedish one?

Second, I believe that a tax-assisted pension plan should be promoted in Japan as the third tier to supplement poor payment of first and second types of pension benefits mentioned above. Personal accounts and 401k plans in the U.S. may provide an example for Japan’s future reform.

Third, in Japan, a spirited debate has concerned whether the consumption tax (Japan’s VAT) is to be earmarked to finance pension and other social benefits or not. I don’t accept the earmarking idea, partly because there is no linkage between the consumption tax and welfare programs, and partly because the consumption tax is one of the major revenue sources.

Fourth, health care spending is projected to expand much faster than other programs in the future. Accordingly, it has been proposed to restrain its growth by linking it to a statistical indicator, say the growth of nominal GDP. Japan’s Ministry of Health, Labor, and Welfare has strongly rejected such adoption of rigid capping rules because health care spending is so closely tied to quality-of-life issues.

Fifth, in order to stimulate a still depressed level of personal consumption, it is sometimes recommended that the consumption tax rate should be raised gradually, say by 1 percent per year. In my view, this is absolutely unacceptable, mainly because tax administration and compliance cannot easily adapt to frequent institutional changes. Think of the economic costs of periodically changing cash registers, price tags, and so on.

Lastly, under the ongoing Trinity reform package, tax sources will be transferred from the central to the local governments to the scale of ¥3 trillion. This is a ridiculous policy in my opinion, although it is a good idea to reduce conditional grants. Substantial amounts of tax revenue would be transferred from the poorest level of government (the central government) to a relatively less worse off one (the local governments). In fact, both governments should independently try to increase tax revenues by using their own tax system.

Comments by Kenneth N. Kuttner

There is no doubt that Japan’s fiscal situation is precarious. The fiscal surpluses of the boom years of the 1980s turned to deficits as the economy stagnated in the 1990s, reaching 6.1 percent of GDP in 2004—the highest among OECD nations. Twelve years of deficits have already pushed the (net) debt-to-GDP ratio to nearly 80 percent of GDP, a figure that is sure to continue rising for the foreseeable future.1

Henry Aaron’s survey of Japan’s predicament poses three important questions: first, what explains the troubling reversal in Japan’s government finances?; second, what steps have been taken to remedy the situation?; and third, what more can be done to restore some semblance of fiscal stability? In addressing these questions, his chapter identifies three problem areas for Japan’s fiscal authorities: the government-sponsored pension (social security) system, the national health care system, and the delegation of spending power to the prefectural and local levels. Aaron offers a number of practical, incremental suggestions for remedying each of these structural problems, while also stressing that taking steps to stimulate economic growth would also be of great help in narrowing the fiscal imbalances.

Aaron’s diagnosis is insightful and sensible, and there is nothing particularly controversial in any of his proposals. However, in discussing the range of solutions to the public pension problem, it might be argued that Aaron adheres too strictly to the pay-as-you-go principle; in addition, he offers little guidance as to how to choose between benefit reductions, tax increases, and alternative policy measures that could help close the funding gap. The purpose of these remarks is to highlight some of these issues, and to suggest ways in which Aaron’s prescription could be refined.

One of the nice features of Aaron’s chapter is the compact and intelligible way in which it summarizes the funding problem faced by Japan’s public pension system. The focus of this discussion is on the pay-as-you-go (PAYG) budget constraint, t = pxr, where r is the tax imposed on the working age population, p is the ratio of pensioners to working age population, and r is the replacement rate for pension benefits. As Aaron documents, Japan’s pension funding problem results from both a steep increase in p and a relatively high level of r. The sharp rise in p is due, in turn, to a perfect storm of adverse demographic factors: especially a remarkable increase in life expectancy, and a significant drop in fertility. Together, these trends imply a doubling of the old-age dependency ratio from roughly 30 percent to over 60 percent by 2050.2

Additional contributing factors include a decline in labor force participation among older male workers, shown in Figure 6.1, and the lack of any meaningful immigration. The only factor tending to reduce p is the rise in female labor force participation, shown in Figure 6.2, but this has been quite modest, and not nearly sufficient to offset the other factors that have increased p. As a result, tax revenues paid into the social security system are already insufficient to meet current payments to retirees—and the further deterioration of demographic trends will only exacerbate the imbalance.

Figure 6.1Labor force participation by men (%)

Source: International Labor Organization.

Figure 6.2Labor force participation by women (%)

Source: International Labor Organization.

Faced with this unpleasant reality, there are really only three things a policy maker can do to satisfy the PAYG constraint: increase t, reduce r, or reduce p. Japanese authorities have already taken a number of significant steps to do just that: in 1986, reducing the accrual rate; in 1994, raising from 60 to 65 the eligibility age for flat-rate benefits; in 2000, further reducing the accrual rate, shifting from wage to price indexation, and raising from 60 to 65 the eligibility age for the wage-related portion of retirement benefits.3 A further pension reform package was enacted in 2004, under which the replacement rate is scheduled to fall to 50.2 percent, from 59.3 percent, and the introduction of a “macroeconomic indexation” scheme for benefits. In addition, the payroll tax rate is to increase gradually from 13.58 to 18.3 percent in 2017. According to calculations performed by IMF staff, these measures are expected to stabilize the system’s operating deficit for the next 25 years or so, assuming a gradual drawdown in the system’s accumulated assets and continuing transfers from the government’s general account.4 Thus, even after the 2004 reforms, Japan’s pension system will fall short of satisfying the PAYG constraint—although the funding gap is relatively modest.

Yet the PAYG criterion is not the only way to assess the sustainability of a pension system. Aaron introduces, and then promptly dismisses, the idea that a pension system should be “self-sustaining,” in the sense that the present value of the benefits received by a given cohort should equal the present value of taxes paid into the system; in other words, that the pension burden B, defined as the difference between these two present values, should equal zero.

Aaron gives two reasons for downplaying this consideration. One is that a truly fair assessment of the burden imposed by one generation on another should include all taxes paid and benefits received, not just those specific to the pension system—a fair point, although calculating such an “all-in” burden would surely be a heroic task. His second point is that, under a PAYG system, the excess benefits received by “early” participants in the system will inevitably impose burdens on future generations; and consequently, the objective of intergenerational equity is a futile one.

Aaron’s chapter would benefit from a slightly more precise statement of this second point: in fact, so long as the population is constant (or growing), and the level of benefits is constant (or growing more slowly than income), the benefits received by the first generation to participate in a PAYG system need never be paid for by future generations. That is, B could be positive for the first generation, and zero for all subsequent generations. The problem arises when, as in the case of Japan, a “large” generation is followed by a “small” one. In this case, the PAYG constraint requires that the “small” generation pay more in taxes than it receives in benefits. A PAYG system therefore gives rise to intergenerational inequality when population size (or benefit levels) fluctuate.

In dismissing the self-sustainability criterion, Aaron seems to be arguing that tax and benefit levels should be adjusted in such a way as to satisfy the PAYG constraint—even if that means shortchanging retirees, current workers, or both; his proposal for equal parts tax increases and benefit reductions essentially asks the old and the young to make comparable sacrifices.

While this proposal has a certain intuitive fairness to recommend it, Aaron offers little guidance on the economic criteria for choosing between tax increases and benefit reductions; both will alter the economic incentives facing workers, firms and prospective retirees, and thus labor supply and demand. Consequently, these changes will affect GDP, and the retiree dependency ratio, p; and these, in turn, will have second-round effects on the system’s finances. And, of course, adjustments to distortionary tax (and benefit) rates will also have welfare implications on a microeconomic level. Taking these effects into account may indicate that the adjustment take place through a reduction in r rather than an increase in t, on the grounds that a further increase in payroll taxes would reduce employment and GDP in the long run—a conjecture that turns out to be consistent with the findings of Faruqee and Mühleisen (2003) based on simulations of the IMF’s multimod model. Such a proposal may prove politically unpalatable, of course; but the economic consequences of the alternatives should nonetheless be made explicit.

Another important issue is whether the PAYG constraint should be relaxed in favor of an intertemporal approach that takes the self-sustainability criterion into account, at least in some limited way. After all, the steep tax increases and benefit reductions required to maintain a PAYG system can, as noted above, impose significant economic costs: taxes (particularly those on employment) are distortionary; and, as observed by Takayama (2005), participants tend to drop out of the pension system when the present value of future benefits is perceived to be small, relative to the taxes paid into the system. Thus the 2004 pension reform, not to mention the additional measures advocated by Aaron, may cause behavioral changes that partially offset any favorable effects on the system’s solvency.

Taking these effects into account, the optimal response to a one-time contraction in population is surely to spread the burden (thinly) over future generations by borrowing, rather than impose the entire burden on the current young generation by adhering to the PAYG principle. This point is acknowledged in Aaron’s recommendation to finance the unfunded liability represented by the system’s “legacy debt.” Given the findings of Takayama (2005) that young workers’ future contributions already exceed the benefits they can expect to receive, Aaron’s advocacy of additional tax increases and benefit reductions is puzzling: based on Takayama’s work, Aaron should be arguing for a reduction in social security taxes, rather than impose the legacy debt burden disproportionately on the current young workers.

A third issue deserving of further investigation is the extent to which the retiree dependency ratio p can be reduced by promoting employment among the elderly. Aaron notes that employment among elderly men is already quite high—but, as shown in Figure 6.1, participation rates have dropped over the past 30 years, despite significant increases in life expectancy. Surely, one factor contributing to the decline in older workers among the elderly is the social security system itself: labor supply would be reduced both by the income effect of the benefits, and, as pointed out by Seike (2003), through the means testing applied to certain benefits. Thus a high replacement rate r has compounded the system’s problems by leading to an increase in p. This suggests that any further reforms of the system should be engineered with an eye to reducing p. Eliminating the negative labor supply incentives created by means testing is one possibility; another is raising the benefit eligibility age from its current 65 years to 67 years, as it is for social security benefits in the U.S.

Overall, Aaron provides a useful and balanced overview of the fiscal challenges currently facing policy makers in Japan. The chapter also does a great service in pointing out that there has already been a great deal of progress on some fronts: in particular, a number of significant steps have already been taken to restore the pension system to a sound financial footing. Much remains to be done, of course, but Aaron points the way to a number of incremental measures that would help narrow Japan’s fiscal imbalances.


Debt and deficit figures are from the Organization for Economic Cooperation and Development (2005, annex tables 27 and 31).

Demographic statistics are from the Statistics Bureau (2005).

A useful summary of these policy changes appears in International Monetary Fund (2004, p. 61).


    Other Resources Citing This Publication