تامین اجتماعی و مدیریت صندوق سرمایه گذاری تراست

In this paper we investigate why and to what extent the government should have a social security trust fund, and how it should manage the fund in the face of demographic shocks, based on a simple overlapping-generations model. We show that, given an aging population, a trust fund in some form could achieve the (modified) golden rule or to offset the negative income effect of a PAYGO system. Besides, in a closed economy where factor-prices effects dominate, using the trust fund as a buffer for demographic shocks could lead to a widening of intergenerational inequality. We also the discuss policy implications of our analysis on the social security reform debate in Japan, including the fixed tax method and the use of the trust fund in the face of a rapidly aging population.

The Japanese government has a huge social security trust fund, which amounted to nearly 150 trillion yen for the Employees' Pension Insurance program (EPI; Kosei Nenkin Hoken) at the end of 2002. This is equivalent to about five years of EPI benefits, compared to about three years of benefits in the United States and only one to two months of benefits in the United Kingdom, Germany, and France. In a pure pay-as-you-go (PAYGO) system, the social security taxes paid by young workers and benefits paid to retirees are balanced at any time. In reality, however, tax revenue can both exceed and fall below benefit payments, and the gap between the two determines the path of the trust fund. The very high level of the Japanese trust fund implies that social security taxes have generally exceeded benefit payments. This presumably reflects Japan's relatively young population structure thus far, as well as the government's strategy of building up the trust fund before facing a rapidly aging population. Given pessimistic prospects for birthrates and wage income growth, however, the current level of the trust fund appears not to be sustainable. In fact, the trust fund has started to fall short of official projections in recent years, mostly due to a reduction in social security tax revenue, reflecting lackluster economic conditions. It is unclear, however, what the optimal level of the trust fund is and how the government should manage it. European governments do not require a huge trust fund as they use the fund simply to absorb unexpected, temporary shocks to social security finances. Alternatively, the government seems to be able to use the trust fund more strategically. For instance, it is often argued that the government should redistribute income between generations through the fund to hold down social security taxes on smaller cohorts. Also, as suggested by Feldstein's (1974) argument, the government may want to accelerate capital accumulation by increasing payroll taxes and adding them to the trust fund. In fact, how to manage the trust fund in the long run is one of main issues in the social security reform debate in Japan. The Ministry of Health, Labor, and Welfare (MHLW), which is in charge of Japanese social security programs, has repeatedly claimed that it should keep the trust fund at a high level to hold down future increases in social security taxes given a rapidly aging population. This affects the current generations because, to sustain the trust fund at a certain level, they have to give up an opportunity to pay lower taxes or receive higher benefits. Indeed, the MHLW plans to continue raising the tax rate over the next decades, albeit with the trust fund at hand at a high level. More and more politicians and economists now insist that the government should start to reduce the trust fund to mitigate demographic pressures, which are expected to mount rapidly in the future. The outlook for the trust fund is also one of the key issues in the social security debate in the United States. The US Social Security Trust Fund now amounts to over one trillion dollars, but the latest 2003 Social Security Trustee's Report released the official projection that the OASDI trust fund will be exhausted sometime around 2042. In the United States the baby boom generation will begin to retire around the year 2010, at which time the large trust fund will start to be run down. If the fund is exhausted, benefits would have to be cut and/or taxes raised, or the government would have to completely change the social security programs. Aaron and Reishauer (1998) insist on shifting the existing trust fund from (non-marketable) government bonds to private securities to raise the rate of return of the trust fund, thereby extending the period before its balance is exhausted. Federal Reserve chairman Greenspan (1999) and many others criticize this, however, arguing that such a shift in the portfolio of the fund would do nothing for national savings and social welfare. This paper investigates why and to what extent the government needs the trust fund in social security policies, and how it should manage the fund in the face of demographic shocks, based on a simple overlapping-generations model.1 As pointed out by Burbidge (1983), Myles (1995), Abel, 2001 and Abel, 2003 and others, the trust fund, if appropriately managed, can help a PAYGO system to enhance capital stock and social welfare in the long run. Following these studies, we first discuss what factors can affect the optimal level of the trust fund, showing that it is difficult to manage a PAYGO system without the trust fund given an aging population. We demonstrate that social security can potentially affect national savings and capital accumulation, especially if it has a trust fund, within the general equilibrium framework of overlapping-generations models. We also discuss how these results change, if we assume a small open economy in which the impact on capital accumulation is neglected. Second, we examine how social security policy should tackle demographic shocks. A PAYGO system is often criticized for making a small cohort worse off, and using the trust fund is considered to be a “buffer,” which is expected to absorb the adverse impact of a declining population. This seems to make sense, if we ignore the impact on capital accumulation and on factor prices; in other words, if we live in a small open economy. In a closed economy, however, a small cohort can enjoy positive factor-price effects due to a higher capital–labor ratio, while a large cohort will have to face a lower level of lifetime income and utility. Indeed, Smith (1982) shows that a PAYGO system can raise social welfare by enabling intergenerational transfers of income. Also, Bohn (2001) emphasizes that even within the framework of defined benefits a small cohort may be better off due to factor-price effects. Thus, from an inter-temporal insurance perspective, it is debatable whether the government should use the trust fund to absorb demographic shocks. Moreover, if the government uses the fund to hold down the tax burden on a certain generation, subsequent generations would receive lower interest on the trust fund. Thus, management of the trust fund over time is related to distribution issues. The remainder of this paper is arranged as follows. Section 2 presents the theoretical framework on which our analysis is based and discusses the optimal structure of the social security system, which has both a PAYGO structure and the trust fund. Section 3 examines how the social security policy should respond to demographic shocks by comparing use of the trust fund and adjustment of the tax rate. Section 4 discusses the policy implications of these analyses on the management of the trust fund over the next decades in Japan. Finally, Section 5 summarizes the main results and presents limitations of our analysis and future research topics.

We have investigated why and to what extent the government needs the trust fund in social security policies, and how it should manage the fund in the face of demographic shocks, based on a simple overlapping-generations model. We have two major results. First, given an aging population, a government can achieve the (modified) golden rule or offset the negative income effect of a PAYGO system by having a trust fund. The optimal level of the trust fund is positively related to the rate of social security tax and negatively related to the rate of population growth. A pure PAYGO system and a fully funded system generally fail to optimize social welfare. Second, in a closed economy, using the trust fund as a buffer for demographic shocks could lead to wider intergenerational inequality by adding to the utility of a small cohort, which enjoys positive factor-price effects, at the expense of other generations. In addition, the trust fund cannot fully absorb demographic shocks, because the interest on it will change and affect the utility of subsequent generations. We have also discussed the policy implications of these results for the social security reform debate in Japan, focusing on an introduction of the fixed tax method and the use of the trust fund. The fixed tax method, which basically aims to contain aggregate benefits within tax revenue and the interest on the trust fund, has a larger negative effect on intergenerational equity in a closed economy than in a small open economy. However, this method can to some extent help avoid net pension liabilities from exploding given an aging population. The use of the trust fund in response to accelerated aging is to some extent a distribution issue, but it depends on where we live. A reduction of the fund is more likely to be justified in a closed economy than in a small open economy, if future generations can enjoy positive factor price effects due to a smaller population size. Our analysis has several limitations and reservations. Most of all, the factor-price effects might be overstated in this paper. Our two-period OLG model seems to be biased towards emphasizing those effects, since it assumes that the young has only labor income and the old has only capital income, and it neglects the labor market where different cohorts compete with each other. Actually, some Auerbach–Kotlikoff-style simulation studies on the transition path to an aging Japan, such as Kato (1998) and Uemura (2001), find mixed effects of a cohort size on the utility assuming a closed economy. A more careful analysis with a multi-period OLG models are needed in order to correctly assess the magnitude of the factor-price effects. The actual relevance of those effects should be judged by empirical studies,9 although it is difficult to get time-series data of sufficient length and stationarity to allow credible statistical inferences. Secondly, policy implications tend to depend heavily on the “initial state.” If the social security is not a “matured” PAYGO system, discussions in terms of intergenerational equity becomes more complicated. The government may have promised the current elderly (and even the old working generations) to give them more benefits than their own tax payments, and postponed burdens will at least partly offset the factor-price effects which future generations are expected to enjoy. If that is the case, reducing the trust fund may adversely affect intergenerational equity. In addition, there remain many Japan-specific, institutional issues yet to be addressed. Most of all, as correctly pointed out by Tamaki (2003), interdependency and transparency of fund management, as well as ownership problems, are serious issues, especially if the government continues to invest the trust fund in private stocks as well as government bonds. In addition, although we implicitly assume that the trust fund and private capital are perfectly substitutable, it is more realistic to assume that the trust fund is used less efficiently than private capital. Actually, most of the trust fund has been invested in or lent to inefficient government institutions via the Fiscal Investment and Loan Program, as analyzed by Doi and Hoshi (2003). These issues should be explicitly taken into account to make discussions about trust fund management more relevant and realistic.