Social Security System and Its Reform

Mukul G. Asher
interviewed by Dr. Byeong-Ho Gong


Mukul G Asher is an Associated Professor of Economics and Public Policy at the National University of Singapore, specializing in public finances of developing countries, social security arrangements in Asia and issues in economic cooperation among nations. He has acted as a consultant to the World Bank, Asian Development Bank, International Labour Organization, UN-ESCAP, Oxford Analytica, Institute of Southeast Asian Studies (Singapore), Malaysian Institute of Economic Research and Employee Provident Fund of Malaysia. He was a Member of the ASEAN-India Expert Group of Trade and Investment Relations set up in August 1994 and has served as an advisor on India to Singapore's Trade Development Board.


Gong: You have been engaged in researching on social security systems, especially in Asian countries. As I understand, most of the social security systems in Asia were imported from the developed countries in the West. Wasn't there any problem?

Asher: Since the affluence, urbanization, and aging were first experienced by the Western industrial countries, it is not surprising that some of the formal social security systems in Asia reflect Western thinking and practices. Even then mandatory national provident funds are much more common in Asia while they are no longer used in the Western countries; and in Asia, there is much greater reliance on informal family, and community support for the elderly than is the case in the Western countries.

The main consideration in designing social security systems should not be the origin of the ideas and practices but whether the reforms proposed are consistent with a given country's needs, capabilities, and characteristics. This requires careful planning and implementation.

There are two common problems of social security in Asia. The first concerns ensuring that the current largely unfunded social security arrangements can be financially sustained without adversely impacting on growth, equity, and international competitiveness of a country.

The second common problem is to devise policies and institutional and regulatory structures to enable the accumulated funds of the pension and provident funds to be invested in a prudent yet remunerative manner. This is indeed a complex and difficult task that requires sustained political commitment and concomitant reforms in such areas as financial and capital markets, corporate governance and disclosure practices and labor markets.

Gong: Could you briefly introduce the national pension in Singapore?

Asher: Singapore relies exclusively on state mandated and managed national provident fund, called the Central Provident Fund (Henceforth CPF) to provide retirement protection. At present (mid-June 1999), the contribution rate is 20 percent for the employee and 10 percent for the employer up to a wage ceiling which is about 2.4 times the average monthly earnings. Each employee accumulates the amount in his or her individual account, and barring a small sum, he can withdraw the rest at age 55 to finance retirement. Singapore thus emphasizes individual and family responsibility for financing retirement.

Gong: Does it have differences, compared to the pension system in Chile?

Asher: While both the countries rely on state mandated savings to provide retirement protection, there are significant differences in the philosophy, modes of operation, and investment returns on accumulated balances. Thus, in Chile, there is reliance on a multi-pillar system, including tax financed first pillar in the form of a minimum government guarantee to provide for adequate replacement rate. In contrast, Singapore relies exclusively on the state mandated and managed second pillar. Thus, it does not subscribe to the multi-pillar philosophy.

With regards to modes of operation, in Chile, an individual member has a choice of investment fund management companies whose sole business is to manage pension funds. An individual also has a choice of switching management companies among various approved pension fund companies. These companies are very tightly regulated and their investment policies and allocations are quite transparent. In contrast, in Singapore, the accumulated balances in the CPF are centrally managed in a non-transparent and non-accountable manner. Thus, the Singapore system of investment management and regulation runs a much higher degree of political risk than the Chilean system. In Chile, there is mandatory provision for purchase of annuity, while in Singapore, there is no such provision.

The Chilean system has provided much higher real rates of return on members balances than has been the case in Singapore. Thus for the 1981-98 period, the real annual rate of return was 11 percent in Chile. In contrast, in Singapore, for the 1987-97 period , the real annual rate of return on accumulated balances with the CPF was approximately zero.

Gong: Some experts are insisting that Korea should implement 'Three Pillar System.' However, they don't recognize the significance of privately managed 'Earnings-related Pension' which the World Bank are suggesting. In order to have the 'Three Pillar System' operate properly, privatization of 'Earnings-related Pension' should be included in the 'Three Pillar system.' What is your opinion on this? Also, could you explain more on the 'Three Pillar System' that World Bank are suggesting?

Asher: The World Bank has recommended three-pillar social security system. The first pillar is tax financed re-distributive pillar to ensure that the lifetime poor do receive certain minimum social safety net. The second pillar is the mandatory savings pillar under which the investment function is undertaken by the private sector while the administration and record keeping is ideally undertaken centrally to take advantage of economies of scale. The pre-funding of retirement involved in this pillar constitutes the main element of the three-pillar social security system of the World Bank. All the countries that have moved to the multi-pillar system such as Australia, UK, and Chile have put the second pillar as the main element of their social security system. It would be appropriate for Korea to consider introducing pre-funding inherent in the second pillar as an important element in its own social security reform. The third pillar is the tax advantaged voluntary savings pillar.

Gong: Would you summarize the social security reform in the world?

Asher: It is difficult to quickly summarize the social security reform in the world, but let me try. There is no ready-made blueprint for social security reform. Each country needs to undertake reforms in the light of its own needs, capacities, and socio-economic characteristics. Nevertheless, certain general patterns are emerging among those undertaking reforms. The basic outline of the reformed pension systems, particularly in Latin America (Chile was the first to introduce this model in 1981, subsequently, Peru, in 1993; Argentina and Columbia in 1994;Uruguay in 1996; Mexico and Bolivia in 117; El Salvador in 1998; and Venezuela in 1999 have adopted some variation of this) and Eastern Europe, but also in the United Kingdom, Australia in 1992; and New Zealand, is to make the tax financed first pillar quite limited in terms of targeted population (this is done through age and means testing), and in terms of replacement rate provided (25 percent of average earnings is quite frequently used ,for example, in Chile, Mexico, and Australia); increase reliance on the Defined Contribution(DC), funded second tier incorporating individual accounts with contributions from both the employees and the employers, with the government's role confined as regulator and as an enforcer of prescribed guarantees to ensure minimum benefit; and creating conducive fiscal and regulatory environment for voluntary saving for retirement.

In the OECD countries, there is recognition that the Pay-as-you-go (PAYG) method of financing social security is fiscally unsustainable. In general Western Europe has opted to tighten administrative rules for pension benefits to cope with the problem, but this is clearly inadequate. Netherlands for example has considerably tightened the requirements for disability pensions.

Some OECD countries, notably Britain have been moving towards greater reliance on Defined Contribution (DC) funded method for more than a decade. The objective is to create a social security system where the tax financed re-distributive component provides only a basic pension. Sweden has also made (during the 1990s) many innovative changes, such as introduction of an individual account with some pre-funding elements, in its current PAYG system.

Gong: It is estimated that Korea has about 180 trillion of implicit pension debt. Would you elaborate more on implicit pension debt to the Korean public?

Asher: The notion of implicit pension debt is analytically straightforward though its numerical estimates are sensitive to the various assumptions made. The implicit pension debt arises because under the pay as you go (PAYG) system, an individual accumulates pension rights along with the social security contributions. While the contributions are used up to pay for current pension benefits, the accumulated pension liabilities of the existing contributors remain. Implicit debt calculations involve putting dollar values to the accumulated pension liabilities at a given point in time.

I understand that Korea has a system under which the current rate of contributions are not going to be enough to pay for the promised benefits.

If the current system is to continue, then either contribution rates will have to go up, benefits would have to be reduced in some way, either directly or indirectly through increasing the age at which benefits can be claimed etc. There are no other options under the present system.

Some countries, for example, Canada, have however decided to invest part of the current excess of receipts over pension benefits into indexed funds, and use higher returns to partially fund future benefits. There is considerable debate in the United Sates about the merit of this approach. For it to work, capital markets must be deep and well supervised, and political economy should permit a hands-off approach to investment management.

If a transition to a fully funded, individual account system is made, and this is a very complex process, then the existing pension rights earned by the existing contributors should be recognized and explicitly paid for. Chile did it through the recognition bonds issued by the government. This increases public debt, and may impact on the interest rates.189 trillion won is indeed a large amount. What does it come to as percent of current GDP?

The second financing cost arises because the present pensioners who already contributed in the past in the expectation of pensions, must continue to receive benefits, but under a funded system, there will be no more contributions coming in. Does your 189 trillion won calculation include this? So the government expenditure will increase and must be financed either through higher taxes, higher borrowing, or reduction in other government expenditure. The macroeconomic effects would depend on the mix chosen, with higher borrowing putting upward pressure on the interest rates. It is because of the above, that strong fiscal position is an advantage in undertaking pension reform.

Gong: We are making efforts to work out countermeasures against increases in medical cost. I have heard that the 'Medisave Accounts System' is very helpful in reducing the medical cost. So, I'd like you to explain further?

Asher: The "Medisave Accounts System" essentially consists of requiring individuals to mandatory put aside a certain sum every month for financing medical expenses. This system is therefore somewhat similar to self-insurance. However, unlike the self -insurance which is useful for small and predictable medical expenses, the Medical Accounts System is more appropriate for large and unforeseen medical expenses such as for hospitalization or for acute illnesses.

The impact of such a system on medical costs depends on which of the two scenarios sketched below obtains in a given situation.

Under the first scenario, if such a system increases demand for medical services but leaves the supply unchanged, then the impact will largely be on the prices of medical care rather than on its quantity. In this case, the impact on medical costs will be reverse of what is intended.

Under the second scenario, if the Medical Accounts System makes individuals less inclined to make excessive use of medical services, and if simultaneously the government takes measures to monitor costs of medical care and to increase its supply, then the intended result of reducing at least a growth in medical costs is more likely. What is important is that along with the introduction of the Medical Accounts System , costs and supply side regulations should also be put in place.

Gong: Nowadays, it seems that labor is very freely moving among nations. I personally think the social security system in Singapore is very compatible with this. And the Adam Smith Institute suggests 'Fortune Accounts' as an alternative idea to the national pension system. If you have an ideal model of social security system, could you please explain it?

Asher: Singapore's social security system is on the individual account base. The fortune account system is not dependent on a single employer. Thus, there is portability of pensions. In that sense, it is consistent with labour mobility.

The Fortune Account idea of the Adam Smith Institute is designed to ensure that each person does have an identifiable account with balances at an early age so that it can take advantage of the magic of compound interest to finance certain merit goods including retirement. This account also incorporates the idea of pre-funding. The government thus, will be responsible only for the initial funding of the Fortune Accounts and for helping the lifetime poor. This would minimize the recourse to the tax transfer process.

In the globalized era, any ideal social security system would involve, significant amount of pre-funding, risk diversification across assets and across countries, and prudential regulation. These elements would need to be combined with the measures to help the lifetime poor secure an adequate replacement rate.

How this ideal should be approached is of course dependent on country specific conditions, and cannot be specified a priori.