Asher, Mukul G., 1997b, “Singapore’s Social Security System: How Well Does it Address the Central Issue in Social Security Reform?” paper presented at a Conference on Solving the Global Pension Crisis (draft: Washington DC, 1997)
Heller, Peter S., R. Hemming, P. Kohnert, and a staff team from the Fiscal Affairs Department, 1986, Aging and Social Expenditures in the Major Industrial Countries, 1980-2025, Occasional Paper No. 47 (Washington: International Monetary Fund).
James, Steven, 1997,“A Public versus a Private Canada Pension Plan: A Survey of the Economics,” Government of Canada, Department of Finance Working Paper.
U.S. Government, 1998,“The Pension Benefit Guaranty Corporation, About PBGC.” Available via Internet: http://www/pbgc.gov/about.hpt.
Paper presented at the APEC Regional Forum on Pension Fund Reforms, Cancun, Mexico, February 4-6, 1998.
The ADB report asserts that “defined contribution pension schemes have much to commend them .... Indeed, for an aging population (such as Asia faces), this is the only approach to pensions that ensures fiscal prudence. Thus, Asian countries would do well to consider defined contribution schemes.” (ADB, 1997), p. 182 (my italics). See also The World Bank (1994) for the most thorough argument for these schemes as the preeminent pillar of a multi-pillar social insurance system.
Only Korea and Taiwan, Province of China, have DB/PAYGO public pension systems with substantial (though not universal) coverage of the work force. Labor force coverage is far more limited for other countries with DB schemes (notably Philippines, China, and Hong Kong SAR, Indonesia, and Vietnam), such that the large majority of workers are outside the system. In China, although workers in state-owned enterprises (SOE) have been covered by DB-type schemes, with minimum pensions for members, the expected rationalization and privatization of the SOE sector raises questions as to how such benefit guarantees will be honored. Even so, no more than 28 percent of China’s working age population is covered by such schemes. Although Malaysia, Singapore, and more recently, Thailand and Hong Kong SAR have established provident funds, the evidence suggests that such funds are largely forced savings rather than pension schemes. Even in the most well-developed countries, the amounts that have been mobilized are not sufficient to provide an adequate retirement income (see Asher (1997a), p. 22). Only Indonesia has begun to introduce a defined contribution scheme for workers in “defined enterprises” (large enterprises and state-owned enterprises), but coverage under this scheme remains extremely limited. As to safety net schemes, only Hong Kong SAR appears to have a means-tested welfare scheme to raise incomes to a minimum level for eligible persons among the elderly, disabled, youth and indigent groups in the labor force.
To render such an approach viable, it is necessary that any resulting surpluses are used to retire public debt rather than leading to higher government consumption. See Hemming (1998).
This may include deferring the age for pension eligibility, reducing the generosity of indexation provisions (which may either be a discretionary feature of the system or of a prescribed character); changing the formula used for determining the initial pension benefit; restricting eligibility for disability or survivor’s benefits; and modifying the tax treatment of pension contributions or pension incomes.
See S. James (1998) for a presentation of the continuum of alternatives.
The transparency “ advantage” of DC/FF schemes is also open to question. While certainly the “political risk” factor raises questions over the transparency of the benefit that a DB scheme may in fact offer (even if individual account statements were to be annually distributed), the DC scheme cannot be said to fare better. How much credibility can be attached to rate of return projections for the next 10-15 years, let alone 25-35 years? Equally, since annuity values at retirement are predicated on the rate of return that will prevail during the 20-30 years after retirement, the level of uncertainty is even greater.
See Asher (1997a), p. 9, as concerns the establishment, by the Malaysian Government, of a M$60 billion (U.S. 20.1 billion) fund in late 1997 to reverse recent sharp declines in stock prices on the Kuala Lumpur Stock Exchange.
In Singapore, there is often a difference between the Singapore dollar nominal return on reserves and the nominal interest rate actually paid on the CPF balances of contributors; in 1995, the difference was about 1.8 percentage points, representing an implicit tax on CPF member’s balances. See Asher (1997a), p. 16; Asher (1997b), p. 4.
Such asset price declines can be seen as one way in which the “circle can be squared” of how to accommodate the tensions between the working and retired generations after 2010-2015. The elderly will require, for consumption, a greater share of the real output of that period. How much of this output will, in fact, be absorbed by the elderly, and how much by the working population and their children, and how this transfer will be intermediated are key questions. Much higher payroll taxes—as under PAYGO systems—may be one mechanism. Higher real commodity prices, reflecting the greater purchasing power of those who may hold the lion’s share of the capital stock, is another. Equally plausible, may be the possibility that reduced asset prices may limit the magnitude of the transfer that will take place between the young and the elderly.
It is interesting to note that one explanation of the recent Asian crisis was the search by international investors for “attractive domestic investment opportunities,” given the weak growth of Japan and Europe. As Fischer (1998) notes, “large private capital flows to emerging markets…were driven, to an important degree by these phenomena and by an imprudent search for high yields by international investors without due regard to potential risks.” With inevitable limits on the pace of growth that the U.S. equity market can sustain over the medium to long term, one must wonder whether, on average, the large majority of pension funds will be able to “beat the market” and to sustain the extraordinary rates of return realized in recent years.
Even in Chile in 1997, asset prices slumped, with the valuation of the equity market dropping by 4 percent (The Economist, Jan. 10, 1998, p. 90).
Asher (1997b, p. 4) notes that in recent years, Singapore has “consciously directed a greater proportion of its resources (presumably including the CPF funds) into investments in East Asia).”
The Chilean Government guarantees a minimum pension for those with a qualifying period of contributions (20 years) if, by the time they reach retirement (65 years for men, 60 years for women), their accounts do not contain adequate resources to provide a predetermined minimum benefit (22 to 25 percent of the average wage). The Government also pays a pension for those who qualify under its social assistance programme, but the number of beneficiaries is restricted to 300,000 (see United Nations (1996), p. 196).
The issue of conjectural liability can also be extended beyond pensions to other areas of social insurance where contributions to a private insurance fund are mandatory, e.g., mandatory contributions to private medical insurance funds (as in Korea). What obligations does the government have to cover operating deficits of such insurance schemes in the face of unexpected economic pressures? For example, in Korea, it would not be surprising if the economic downturn in 1997-1998 resulted in a fall in such contributions funds. Presumably, outlays of such funds would be less sensitive to cyclical developments, such that these funds could begin to run deficits. Pressures on the government would also emerge from government obligations to cover the premia of the rising number of unemployed workers.
Even where governments provide indexed annuities, there may be risks associated with how governments measure the extent of inflation. The choice of inflation index may be distorted by inclusion of commodities subject to price controls.
Indeed, as S. James points out, the World Bank (1994) itself notes the difficulties of ensuring sufficient adequate public regulation of private annuity markets, particularly if they were to become mandatory. But with voluntary annuitization, “private annuity markets will suffer from adverse selection and the key characteristic of a pension—the efficient provision of longevity insurance—will be absent.” (p. 17).
In countries where there is to be a transition from a DB/PAYGO pillar, it is generally recognized that the transition generations (viz., typically the current working population) will pay a double burden of pension support, associated with the payment, through taxes and debt service, of the cost of supporting existing retirees as well as the cost of the contributions required to accumulate assets for their own retirement. Less recognized is that there may be a further burden that may still need to be borne, either by this transitional generation or subsequent working generations, to the extent that the value of the assets accumulated by the time of retirement under the new schemes proves insufficient to cover the costs of retirement (because of possible asset price deflation) such that benefits are less than envisaged or the government is required to finance additional benefits from general tax revenues.
Similar issues also arise in the context of national medical insurance reform initiatives, whereby mandatory contributions are legally required for the funding of private medical insurance funds (as in Korea (see Heller, 1997)).
Some have raised the question of why one might wish to distinguish between future outlays on pension and medical insurance from other types of government expenditures for which there is a similar type of commitment, e.g., education outlays. Formally, one should include such expenditures as well. However, the issue principally arises because one can expect that the impact of aging populations will be to increase pension and medical care expenditures significantly, with increasing deficits at current tax or contribution rates. In contrast, for education, one might expect indeed to see some decline in the tax rates required to finance this sector, given trends in the share of the youth population. Consistency would thus equally require a reduction in the magnitude of public debt in the future (or correspondingly a reduction in tax rates). As shown in Heller (1987) and more recent OECD studies, however, the magnitudes of potential savings do not appear large. Thus, for empirical rather than analytical reasons, the issue has not acquired much policy significance.
The tax treatment of pensions can equally create distortions on the size of taxes and expenditures. Where pension contributions are tax deductible and taxation is deferred until the pension income is realized at the time of retirement (both the original contribution and the rate of return earned over the period of accumulation before retirement), it is equivalent to a tax being received by the government which is then relent to the individual, which would of course lead to both higher government revenues and expenditures at the time of the pension contribution. In fact, such a treatment of pension contributions is effectively equivalent to a tax expenditure and thus excluded from revenue and expenditure statistics.
In the United States, the budget is inclusive of social security operations. In Japan, the general government balance is typically portrayed inclusive and exclusive of social security operations.
Even such measures are often then subject to more complex analysis—taking account of the weighted impact of different types of fiscal instruments. More broadly, if one is a Ricardian, one may seek to assess the extent to which private sector agents seek to offset the impact of current fiscal operations that engender government debts that will need to be paid off by higher taxes in the future.
Such schemes also typically incorporate other types of redistribution associated with the treatment of benefits for nonworking spouses, working spouses, children, and, in some cases, even for unmarried daughters. Specific occupational groups in hazardous activities may also receive additional benefits.
There have been many empirical studies seeking to estimate the actual redistributional consequences of such DB/PAYGO schemes (see the World Bank (1994) for a useful survey of the results, which largely derive from a few industrial countries—the US, UK, Netherlands, and Sweden). Such studies have not only focussed on the extent of intragenerational redistribution across income groups but also on the extent to which different generations receive different rates of return on their lifetime contributions. Intragenerational studies have revealed that taking account of differences in the length of the overall working life, the pattern of wage and salary income over the course of a working life, and in the life expectancy prospects of workers of different income groups, the intentions built into the legal formulae to achieve redistribution may not in fact occur. Since the poor start work earlier, typically have flatter age-earnings profiles, and die earlier, their contributions tend to be higher, and their benefits are both lower relative to lifetime income and lesser in duration during retirement. What is striking to a reader of these studies is the difficulty of making any conclusive assessment of the redistributional consequences of a public pension scheme, given the multiple perspectives that one can have on how to evaluate any redistribution that may occur.
See Asher (1997a). In Malaysia, “contributions (subject to deduction ceiling), interest paid on balances, any capital gains arising from withdrawals before retirement for housing and other assets, and retirement withdrawals are all exempt from income taxes and estate duties. As the income tax rates are nominally progressive, …the tax saving is proportionately greater for those earning higher wages.” (P. 7). A similar approach prevails in Singapore (p. 14).
In the case of the UK, where there is provision for some contracting out to the private sector of funded DB plans, the National Insurance Fund, which is funded by contributions by workers, nevertheless provides for an indexed flat rate (minimal replacement rate) basic pension, paid in full to those who have paid NICs for 90 percent of their working life. For those who have contracted out of the State Earnings Related Pension Scheme (SERPS), there may also be an entitlement to additional SERPS payments to the extent that their entitlement under the guaranteed minimum pension of the private schemes is less than would be the guaranteed minimum under the SERPS.