Equitable and Sustainable Pensions

Chapter 12. Singapore’s Pension System: Challenges and Reform Options

Benedict Clements, Frank Eich, and Sanjeev Gupta
Published Date:
March 2014
  • ShareShare
Show Summary Details
Mukul G. Asher and Azad Singh Bali 


Singapore has evolved from a low-to-middle-income country to a high-income country in a relatively short period. Singapore’s economic success notwithstanding, several factors suggest that policymakers should focus more on promoting the equity and sustainability of the country’s current pension arrangements. Concern about congestion externalities—particularly in transport, housing, and recreational facilities—and competition for positional goods—such as cars, housing, and education—have also become an integral part of the political and social discourse in Singapore.

The chapter is organized as follows: The next section briefly discusses demographic and labor market trends and is followed by a section that discusses the main characteristics of Singapore’s pension system. The subsequent section provides an assessment of the equity and sustainability of the pension system, with a discussion of possible initiatives such as implementing social pensions.

Singapore’s Pension System

Current Structure

Singapore’s pension system relies overwhelmingly on a mandatory savings tier administered by the board of the Central Provident Fund (CPF) and supervised by the Ministry of Manpower. Set up in 1955, the system has evolved into a key socioeconomic institution that affects the welfare of Singapore households. Given the system’s complexity and multifaceted nature, a brief overview of its characteristics that are relevant for analyzing equity and sustainability is provided.1 These aspects are aggregate indicators, high preretirement withdrawals, administered rate of interest credited to members, and CPF LIFE, an annuity scheme at the payout phase. Civil service and military pension arrangements are also discussed.

Aggregate indicators

The CPF system was established in 1955, but only since 1968 have a variety of preretirement asset-accumulation schemes been introduced. The schemes were introduced in response to various ad hoc policy objectives, and have frequently been revised. The contribution rate structure has been altered accordingly.

The CPF system is open only to Singapore’s citizens and permanent residents. The contribution rate and the shares allocated to the three different accounts2 vary with age. The contribution rate—subject to a wage ceiling of S$4,5003 per month—ranges from 36 percent for members less than 50 years old, to 32.5–14.5 percent for members between 51 and 65 years old, and 5 percent for those older than 65. The proportion of the contributions allocated to a member’s different accounts also varies by age, with the share explicitly allocated for retirement purposes in the Special Account falling with age. These allocations and arrangements suggest that policymakers assign low priority to accumulation of cash balances for retirement.

Key aggregate indicators for the CPF are presented in Table 12.1. The active CPF members—those who have contributed in any of the past three months—include 85 percent of the resident labor force (that is, citizens and permanent residents). Thus, the CPF’s coverage has been comparable to that in other advanced economies and can be considered universal; however, there are a large number of inactive members.

Table 12.1Key Aggregate Indicators of the Central Provident Fund, 2012
Total MembersShare (percent)
Active members/total members52.0
Active members/total labor force54.6
Active members/resident labor force85.0
Aggregate BalancesS$ billionAs percent of 2011 GDP
Total member balances224.768.7
Ordinary Account90.427.6
Special Account51.915.8
Medisave Account58.617.9
Retirement Account23.87.3
Sources: CPF Board (2012); Government of Singapore (2012a).Note: Data are as of September 30, 2012; labor force data are for 2011. Details may not add up to the total due to rounding.

CPF member balances as of September 30, 2012, representing the gross savings of the 3.4 million members, were the equivalent of 69 percent of GDP. The largest share was in the Ordinary Account from which withdrawals can be made for house purchases and other expenses during a member’s working life. The balances to finance health care expenditure were 18 percent of GDP, while the remaining was for financing old age. This breakdown suggests that the CPF system has multiple purposes, only one of which is financing old age.

High preretirement withdrawals

Because the CPF has multiple purposes, substantial withdrawals are made for housing, retirement, health care, and other purposes, leading net contributions to be relatively modest. Withdrawals for 1997–2011 averaged 74.9 percent of contributions. During this period, the CPF’s contribution to gross national savings averaged 8.1 percent, including interest income earned on CPF balances. This is a relatively modest contribution because Singapore’s gross domestic saving rates have been high, averaging 49 percent during 1997–2011. This figure suggests that public and business sectors have contributed much more to Singapore’s high saving rates than have mandatory CPF savings.4

Administered interest credited to members

In any defined-contribution (DC) system, returns on member contributions and accumulated balances affect member savings for retirement. Since the CPF’s inception, members have been credited with an administered rate of interest, with a minimum government guarantee of 2.5 percent nominal interest on all accounts. On January 1, 2008, savings in the Special, Medisave, and Retirement Accounts (SMRA) were pegged to the 12-month average yield of the 10-year Singapore Government Security plus 1 percent. To help CPF members adjust to the floating SMRA rate, the government pledged to maintain a 4 percent minimum rate for government securities until December 31, 2013, after which the 2.5 percent floor rate legislated in the CPF Act will continue to apply for all CPF accounts including SMRA.

The asset side of the CPF balance sheet comprises nonmarketable government securities, on which interest is determined after the fact. How the proceeds from these securities, which form a major part of Singapore’s internal debt,5 are used by the government lacks transparency. IMF data indicate that during 2000–12, Singapore’s overall fiscal balance exhibited an annual surplus of 5 percent of GDP.6 This figure suggests that the government has no need to borrow from the CPF system to finance its expenditure. It is therefore a reasonable assumption that the optimum deployment of the CPF balances is not made by the CPF Board or the general government itself. The Singapore Government Investment Corporation (SGIC), Singapore’s prominent sovereign wealth fund, which invests globally, is presumed to be the agency ultimately deploying CPF funds. By statutory provision, SGIC is not required to reveal its financial performance and activities. Thus, CPF members are not provided information on the ultimate investments of their balances.

Figure 12.1 presents implicit real returns on CPF balances in comparison with growth in real wages and real GDP for the period 1987–2011. Because the interest credited by the CPF Board to its members’ accounts is used to estimate implicit returns, the estimation includes differing administered interest rates paid to various accounts. Mean wage data reported by the CPF in Ministry of Manpower statistics were used to calculate growth in real wages. Growth in real GDP and real wages exceeded the implicit real rate on CPF balances by nearly six and three times, respectively, in the 25-year period. Compound interest calculations suggest that a sum of money growing at 1.4 percent per year will double in nearly 50 years. In September 2012, the average CPF balance was S$66,000. Because CPF is the primary source of financing retirement expenditure, it appears that the current CPF member balances would be inadequate to provide a reasonable replacement rate during retirement, though the extent of the inadequacy will vary by member.

Figure 12.1Implicit Real Return on CPF Balances; Growth in Real GDP and Real Wages, 1987–2011


Source: Authors’ estimates using data from CPF (various years), Department of Statistics (various years), and Ministry of Manpower (various years).

Note: CPF = Central Provident Fund.

To the extent that the SGIC earns higher returns on CPF balances than is credited to members, there is an implicit tax on CPF wealth that is recurrent, highly regressive, and often quite large (Asher and Nandy, 2011).

The replacement rate reflects the ratio of retirement income to preretirement income, so the higher rate of wage growth, as compared with returns on balances, has adversely affected the replacement rate for CPF members. More detailed analysis of the replacement rates requires disaggregated data on the density of contributions and age-specific member balances, but these are not available. This lack of data appears to have significantly restricted the public policy debate on old-age financing issues.


CPF LIFE, introduced in 2009, is a deferred annuity scheme, with individuals bearing the cost of purchasing the annuity from their accumulated balances in their retirement accounts.7 When introduced, the scheme offered four plans: plus plan, balanced plan, basic plan, and the income plan, which varied by the amount of bequest (if any) that the participant would want to leave for his or her beneficiaries. The scheme has two parts: a deferred annuity component and a savings component. Once an individual decides to participate in the scheme, he or she is required to purchase an annuity (the premium is based on gender and age, and is paid from his or her Retirement Account), which starts either at the participant’s drawdown age or at age 80 or 90. The remaining savings (if any) in the participant’s Retirement Account is drawn down on a monthly basis until age 80 or 90, after which time the deferred annuity begins. Payouts are not indexed to prices.

Effective January 2013, the CPF Board reformed the CPF LIFE scheme to offer two plans: the standard plan and the basic plan. The scheme will be compulsory for those born after 1958 and who have S$40,000 in their Retirement Accounts at age 55 or at least S$60,000 at their drawdown age. Members can voluntarily join the CPF LIFE scheme at any age between 55 and 80. As before the reform, the plans vary by the amount of bequest that members wish to leave. The annuity is to be bought at age 55, but in the standard plan, payouts will not begin until the drawdown age of 65. It should be noted that CPF LIFE does not increase resources available to individuals for retirement; however, it gives greater control to the CPF Board over the stock of savings of CPF members, because CPF LIFE is organized and administered by the CPF Board. It also provides greater security to those purchasing CPF LIFE products because of the government’s implicit backing.

Civil service and armed forces pension arrangements

Singapore initiated measures to establish a common system for pensions for civil servants and private sector employees. As a result, an overwhelming proportion of civil servants, political representatives, and appointees are members of the CPF.8

Armed forces personnel are governed by a separate DC scheme called the Saver Fund, which was established in 1998. The value of the accumulated pension benefits at the time of the fund’s introduction were estimated and transferred into members’ accounts.

The Saver Fund receives contributions from the Consolidated Revenue Account of the government, mandatory contributions from personnel, and income earned from its investments. The contribution rate for the first six years of service is 13 percent, after which it is increased to 15 percent. Unlike the CPF, the Saver Fund gives members limited options for investing their accumulated balances.

Future Challenges

Singapore’s pension system faces a number of challenges, which are partly related to the country’s past and future growth strategies.

First, one of the consequences of Singapore’s growth strategy has been the rising share of the country’s noncitizen population, a ratio that increased from 14 percent in 1990 to 26 percent in 2000, then to 38 percent in 2012.9 The growth rate of citizens has been a fraction of the growth rate of noncitizens. Such a trend is difficult to sustain in any country in the long term.

Second, the population of Singapore is expected to age very rapidly in the next two decades as a result of below-replacement-rate fertility rates since 1975.10 The United Nations (UN) projects that the population older than 65 will grow from about 0.46 million in 2010 to 1.40 million in 2030, an increase of 207 percent in just two decades (UN, 2010). Life expectancy at age 65, which was 18.3 years for men and 21.8 years for women in 2011, is also expected to rise (Government of Singapore, 2012d). It is projected that a substantial proportion will live until age 85, and the population in their nineties will also increase in the future. These demographic trends are likely to drive the need for age-related pension (and health care) expenditure.

The official response to the rapid aging of Singapore’s population is to raise productivity through business restructuring and workforce retraining, encouraging higher labor force participation, and continuing to encourage foreign labor—albeit in a calibrated manner (Government of Singapore, 2012b). This response suggests that policymakers recognize that the single-minded pursuit of growth is no longer desirable. Growth needs to be calibrated to address social and political acceptability, particularly its impact on congestion externalities and positional goods. The trade-off between the rate and composition of growth on the one hand, and social needs and political acceptability on the other hand is a major policy dilemma for Singapore. This dilemma is especially vexing because policymakers appear unwilling to undertake substantive fiscal measures and family-friendly initiatives that could positively affect the current low fertility rates.

Third, Singapore has relied primarily on a single-tiered retirement financing system involving mandatory savings administered by the CPF. With increasing longevity and continuing increases in the old-age dependency ratio, relying on savings from income during working years to finance retirement—which, in some cases, may exceed the proportion of life spent in the labor force—has become increasingly untenable for a significant amount of the population.

Fourth, policymakers in Singapore have expected that a longer working life will significantly contribute to retirement income security even while primarily relying on the mandatory savings tier, but the age-specific labor force participation rates (LFPR) for males and females in Singapore do not lend strong support to such an expectation. In 2011, Singapore’s total LFPR, at 66.1 percent, was higher than the corresponding figure for Japan (60 percent) and Korea (61 percent), but marginally lower than that of Canada (67 percent). However, for older age groups, between 60 and 64, Singapore’s LFPR at 51 percent compares unfavorably with those of Japan (55 percent) and Korea (56 percent). A similar pattern is also observed in the LFPR for the age 65–69 group. Nevertheless, any effort to encourage longer working lives merits consideration.

Reform Options

The discussion so far of Singapore’s pension system strongly suggests that the absence of risk-pooling and risk-sharing mechanisms has important implications for equity, particularly gender equity, and sustainability. This section assesses Singapore’s pension system focusing on specific measures that could improve its equity and sustainability. It suggests that introducing social pensions, which are noncontributory and financed from the budget, could help improve equity and sustainability while being fiscally affordable.

Equity Assessment

Equity may be interpreted in different ways but generally involves ensuring that different groups in the society, and men and women in particular, do not have vastly different levels of pension provision and characteristics. The three main areas for which opportunities to improve the equity of Singapore’s pension system exist are tax treatment, payout-phase arrangements, and treatment of foreign workers.

Tax treatment

CPF contributions (subject to a wage ceiling), income, and withdrawals are free of tax. Because CPF contributions by employees are tax exempt, the rate of subsidy varies with the marginal income tax rate. Individual income tax rates in 2009–10 ranged from 3.5 percent to 20.0 percent, and the total number of income tax payers was 33.8 percent of the labor force. Because the labor force includes many high-income-earning expatriates, the share of income tax payers among citizens and permanent residents is likely to be lower. Therefore, the implicit tax subsidy is regressive, with the vast majority of CPF members not benefiting from the income tax deduction. The wage ceiling on which CPF contributions are paid has been lowered, and so have income tax rates. Therefore, the extent of the implicit subsidy is likely to have been lowered as well.

The regressivity is compounded by the implicit tax on CPF wealth, which falls disproportionately on the bottom half of the income group. According to the SGIC’s Annual Report of 2008–09, for the 20-year period ending March 31, 2009, the annualized rate of return on its portfolio was 4.4 percent in nominal terms and 2.6 percent in real terms (SGIC, 2009). Assuming the CPF balances are managed by the SGIC, the implicit tax on the CPF wealth is the difference between what the SGIC has announced as its returns and what is credited to CPF members’ accounts. Applying the difference between 2.6 percent and 1.4 percent (Figure 12.1) obtained on CPF balances provides a crude estimate of the implicit tax of CPF wealth of S$2.63 billion (S$2.63 = (2.6–1.4) × S$219.3 billion).11 The tax is both large and regressive because households on relatively lower incomes are likely to have a larger proportion of their wealth in the form of CPF balances. In estimating Singapore’s household tax burden, this implicit tax should be included. This tax also affects intergenerational fairness, because the current generation bears the burden of reduced consumption for the potential benefit of future generations.

This implicit tax suggests an opportunity is available to improve equity by reducing or eliminating the implicit tax on CPF wealth. This tax could be eliminated by, for example, transferring progressively the CPF balances currently being managed by Singapore’s sovereign wealth funds to an autonomous CPF Board with the mandate to manage the CPF funds in the interests of CPF members in a transparent and accountable manner. The precise method of addressing the implicit tax on CPF wealth merits much more rigorous analysis than can be provided here.

Payout-phase arrangements

In 2009, CPF LIFE was introduced for the payout phase of the CPF. Its design raises at least two important equity issues.

First, the premiums charged for the annuity are based on age and gender, with women paying a higher effective premium than men because women, as a group, live longer. Thus, the CPF LIFE premium is structured along private, not social, insurance methods. This structure is particularly disadvantageous to women, who, as a group, have lower CPF balances than do men but need income support for a longer period.12

Second, the annuity benefit is specified in nominal terms, which implies that the real value of the annuity will decline in line with Singapore’s future inflation rate. This will not only reduce the replacement rate as the retirement period progresses, but will also not permit individual retirees to benefit from Singapore’s future economic growth.

The above analysis suggests that preretirement income inequalities are not only carried forward in retirement but are also accentuated. Therefore, the policy focus in Singapore needs to shift from targeting absolute levels of income to relative levels of income. There is also considerable merit to introducing a social insurance method for providing the annuity at the payout phase and to increasing the transparency of the actuarial and other assumptions relating to CPF LIFE.

Treatment of foreign workers

The stock and flows of foreign workers are subject to government regulation and administrative measures. The structure of the levy and its design are complex, with variances across sectors, skill levels, and nationality, and along the dependency ratio ceiling. Disaggregated data on the number of foreign workers are unavailable, but at end-2010, there were 201,000 foreign domestic workers. The budget does not provide separate revenue data for levies on foreign workers, because it combines the information with the Airport Passenger Service Charge under the “Other Taxes” category. In 2011, the revised revenue estimate from Other Taxes was S$3.15 billion (equivalent to 6.9 percent of total tax revenue; Government of Singapore, 2012a), suggesting that the revenue from the levy on foreign workers is nontrivial.

A substantial proportion of the Other Taxes collected can reasonably be assumed to be levies on foreign workers. These workers, therefore, contribute significantly to fiscal revenues in Singapore. However, they are not members of the CPF, are ineligible for social and community benefit schemes, and do not receive health care subsidies and benefits like Singaporean residents do.

To help enhance equitable treatment of foreign workers, a better balance between the revenue derived from them and the budgetary expenditure primarily benefiting them merits serious consideration.

Sustainability Assessment

It is important to distinguish between financial and economic sustainability of a pension system. Financial sustainability refers to the matching of liabilities and assets. If projected liabilities are greater than assets, then the pension plan must either reduce benefits, increase contributions, increase income earned on accumulated contributions, be subsidized by the government, reduce its administrative costs, or undertake a combination of these measures to ensure that the scheme is financially viable or sustainable. By contrast, economic sustainability is the capacity of the economy to finance projected liabilities without sacrificing economic growth or other priorities. In this context, the most important macroeconomic variable is the long-term trend of economic growth (Barr and Diamond, 2008).

Sustainability in a DC-based pension system is intricately linked to adequacy of pension benefits. In a DC system—in which contribution obligations are defined but the benefits depend on the accrued balances of contributions, interest earned, extent of preretirement withdrawals, and outcomes of conversion of accumulated balances into a retirement income stream—individual members bear the investment and macroeconomic risks, such as unemployment, relatively stagnant wages, or higher-than-anticipated inflation.13

The sustainability of Singapore’s pension system may be viewed from a narrow perspective of the CPF system’s fiscal sustainability and from a broader perspective of constructing a pension system that provides adequate real (that is, inflation-adjusted) income throughout old age, thus mitigating longevity, inflation, and survivors’ risks. From the narrow financial perspective, the CPF system is sustainable.

The CPF system is obligated to return only the accumulated balances, with an explicit government guarantee of 2.5 percent nominal interest on CPF balances and an implicit guarantee that the balances and CPF LIFE obligations will be met by the government.14 These promises are credible given Singapore’s strong public finances, as indicated by its long-term structural fiscal surpluses (IMF, 2012) and the political importance of the CPF system.

The fact that the CPF system is fiscally sustainable is an important achievement, but is not sufficient to ensure sustainability in the broader sense. The primary reasons are an oversimplified organizing principle (mandatory savings using a DC method) and a consequent absence of social risk pooling, requiring individuals to bear macroeconomic and other risks for which they are not necessarily equipped. This oversimplification is accentuated by assigning a predominant role to mortgage finance, by requiring savings to pay for health care expenditures under the CPF system, and by employing an administered rate of return that leads to an implicit tax on CPF wealth.

The CPF Board does not publish the cash balances of its members and other relevant information, so it is not feasible to estimate the average replacement rate—that is, the ratio of preretirement income to retirement income in real terms throughout retirement. It also does not publish replacement rates for members. Nevertheless, there are indications that the inflation-adjusted replacement rate will be inadequate for most CPF members. First, even for a significant proportion of active members, the requirements for setting aside the stipulated Minimum Sum for basic needs are not being met. In 2011, only 45 percent of those 33,644 active members who turned 55 in 2011 had sufficient cash balances to meet the Minimum Sum requirement of S$112,000 (in 2003 prices). Thus, the majority of members did not even meet the stipulated premium for the basic income requirement, let alone being able to maintain their preretirement living standards. The Minimum Sum is set to increase to S$120,000 (in 2003 prices) by July 1, 2015, about twice Singapore’s per capita income.

Second, retirement provision is focused on meeting basic minimum needs in old age. However, it is not absolute but relative poverty that is an increasing concern in affluent societies, including Singapore. Singapore’s pension system does not focus on relative poverty, as evidenced by the absence of social insurance principles, the absence of a budget-financed inflation-adjusted basic pension, and the insistence on levying health insurance and CPF LIFE premiums according to age and gender. The gender discrepancy particularly reduces the adequacy of retirement income for women, who, as a group, live longer and participate in the labor force in much lower numbers.

Reform Options and Their Projected Fiscal Costs

Based on the arguments above, a proposal to introduce social pensions with benefits related to average wage income or to per capita GDP, and financed through the budget, merits consideration to improve the equity and sustainability of Singapore’s pension system. A social pension could help improve equity by providing noncontributory benefits to those who have low lifetime incomes.15 In Singapore, a social pension would particularly benefit women, who live longer but on average obtain lower income from labor markets and capital.

Social pensions could also improve sustainability in the broader sense by addressing the adequacy of Singapore’s pension system. Because they are paid until death, social pensions better address longevity risks. Because social pension benefits are related to average wages or to per capita income, they also enable beneficiaries to share in future economic growth—which the current mandatory savings-based pension system does not allow.

The remainder of this section presents the estimated fiscal costs of a hypothetical social pension scheme that can potentially address the equity issues discussed above. The assumptions made in this estimation are as follows: GDP is assumed to grow in nominal terms between 5 and 6 percent per year. This growth rate allows for inflation of 3 percent a year, and real growth of between 2 and 3 percent. These parameters are consistent with the Singapore government’s macroeconomic goals (Government of Singapore, 2013). Similarly, nominal wages are assumed to grow between 4 and 5 percent, lagging nominal GDP growth, but permitting real wage growth. No means testing using income or assets to qualify for the social pensions are used in the estimation. Social pension benefits are restricted to citizens age 65 and older and are provided on an individual basis. Ideally, because there are economies of scale with respect to family size, the benefit levels for households with more than one elderly person should be adjusted to reflect such economies of scale. However, this adjustment has not been made because of the lack of relevant data on elderly household composition.

Furthermore, the estimation does not have any behavioral foundations; it does not take into account the administrative or compliance costs of a social pension scheme. Although it takes into account the share of elderly citizens (age 65 and older) in the total population, it does not take into account the increasing life expectancy of the beneficiaries. In 2011, the average life expectancy at age 65 was 20 years (Government of Singapore, 2012d), and is expected to increase with advances in medical technology, diagnostics, and treatment. Thus, the simulation presents only the static fiscal costs required to finance the social pension.

The total population ages 65 and older is estimated from United Nations (2010) using the medium variant projections. In 2010, 9.0 percent of the population was 65 or older. This proportion is expected to increase to 23.3 percent by 2030. This projection assumes that the citizen share in total population will remain at 60 percent until 2030.16

Table 12.2 presents the estimated fiscal costs of a hypothetical social pension scheme in Singapore, assuming different benefit levels under two widely used parameters (wages and per capita GDP). Because the benefit levels are proportionate to wages or to per capita GDP, the benefits proposed enable the individual to share in the growth of the economy, which is important for improving the equity of Singapore’s pension system.

Table 12.2Estimated Fiscal Cost of Hypothetical Social Pensions in Singapore(Share of GDP)
Benefit Level
YearVariant A 20% of median annual wage incomeVariant B 30% of median annual wage incomeVariant C 15% of per capita GDP (current prices)Variant D 20% of per capita GDP (current prices)










Sources: Authors’ estimates based on data from United Nations (2010); Government of Singapore (2012c, 2012d).Note: Figures in parentheses refer to the range for the assumed 4–5 percent nominal wage growth, and the 5–6 percent nominal GDP growth. Because the benefit is indexed to per capita GDP and these estimates are shares of GDP, there is no range for the estimates for Variants C and D.

The benefit level used (20 percent and 30 percent of median annual wages in Variants A and B, and 15 and 20 percent of per capita GDP in Variants C and D) is based on international experience. For example, Australia’s social pension benefit in 2013 was S$24,799 (equivalent to 32.7 percent of per capita GDP) per year per individual and S$37,382 per year (equivalent to 49.3 percent of per capita GDP) for a two-member household.17 These benefits are subject to income and asset criteria, and are not provided on a universal basis. Additionally, individuals who may not meet the criteria receive partial pensions (Government of Australia, 2013). More conservative benefit levels are used here because access to disaggregated household composition and income data were not available to implement a means test.

The projections indicate that with the rising share of the population age 65 and older, and with increasing wages and incomes, the fiscal cost will increase rapidly between 2012 and 2030. Thus, assuming a benefit level of 30 percent of the median annual wage, the fiscal cost will rise from 1.16 percent of GDP in 2012 to 2.60 percent by 2030. The range of costs projected in the brackets for Variants A and B is due to the different assumptions of the parameters. The range widens over time because of the compounding effects of the two parameters.

It is important to reiterate that these estimates are based on many assumptions, and more rigorous simulation exercises are needed to actuarially project fiscal costs of social pensions; however, the relevant data for such exercises are not available. The exact design of the social pension will have significant implications for the fiscal outlay required to finance it. Eligibility, benefit levels, and the extent to which partial pensions are to be paid are important policy considerations that require rigorous empirical scrutiny.

How are these proposed fiscal costs to be financed? The government has resisted pressures to increase social expenditure on health care and pensions substantially by arguing that it would have to increase taxes to finance them, which would hurt the country’s growth potential (Government of Singapore, 2012a). As mentioned, Singapore has enjoyed structural fiscal surpluses that averaged 6 percent a year during the past decade (Figure 12.2). This suggests that pension reforms designed to enhance adequacy and equity are not constrained by fiscal capacity or fiscal space.

Figure 12.2Singapore: Overall Fiscal Balance, 2000–13

(Percent of GDP)

Sources: Estimated from IMF Article IV Staff Consultation Reports for Singapore, 2005, 2008, and 2011.

Note: Figures for 2012 and 2013 are estimates.


This chapter assesses Singapore’s pension system from the perspective of equity and sustainability, including adequacy of risk mitigation and coverage of the population. Singapore’s pension system has been organized around two main premises: that mandatory savings during the working years, micromanaged by the state, will be sufficient to provide adequate retirement financing, and that public policy should focus on the absolute rather than the relative level of income.

A pension system organized around these two premises faces increasing limitations in equity and sustainability (adequacy), requiring a fundamental rethinking. Parametric changes such as adjusting the contribution rate or the wage ceiling, or the design of the payout phase of the pension program, will not significantly improve equity or sustainability. This chapter argues that social risk-pooling arrangements through social insurance and through a proposed noncontributory budget-financed social pension, which is based on relative rather than absolute income levels, have become necessary and merit consideration.

The chapter suggests several specific measures, including changing payout-phase arrangements, ending the administered interest rate on CPF balances, and introducing social pensions to enhance the equity and sustainability of Singapore’s pension system.

The analysis suggests that structural budget surpluses will help finance pension reforms along the lines discussed above without serious reductions in expenditure on other government priorities, such as human resource development, health, and infrastructure. The main constraint on pension reforms is not fiscal, economic, institutional, or capacity related—it is the need to assign a higher priority to reforming existing pensions.


For a more detailed discussion of the various components of the CPF system, see Asher and Nandy (2011). The official website of the Central Provident Fund is http://www.cpf.gov.sg.

Contributions are channeled into three accounts: Ordinary, Medisave, and Special. Broadly, these are to be used for housing and investments, health care expenditures, and old-age and retirement investments, respectively.

In mid-2013, US$1 was approximately S$1.27.

Data on the contribution to savings (compulsory and voluntary) from households, businesses, and government are not available.

Singapore’s domestic debt was S$321.2 billion in 2010 (105.8 percent of GDP).

Calculated using data presented in the IMF Article IV Consultation Reports for 2011 and 2005 (IMF, 2006, 2012).

The CPF LIFE scheme applies only to the Minimum Sum. At age 55, when members can withdraw funds from their CPF accounts, they are required to retain a minimum balance (US$139,000 in 2012). If at age 55 a member does not have the minimum balance, the member’s real property (regardless of whether purchased with CPF Ordinary Account funds) is automatically pledged to meet the difference. If the member’s accumulated balances are insufficient for the Minimum Sum, then there is no requirement that additional funds be deposited to equal the Minimum Sum. The Minimum Sum is based on an absolute concept of poverty.

There are, however, some differences in the design details, including benefit levels, between civil servants and private sector employees (Asher and Bali, 2011).

In 2012, of the total population of 5.31 million, citizens constituted 3.29 million, permanent residents 0.53 million, and foreign skilled and semiskilled professionals on various work visas and their dependents 1.43 million (Government of Singapore, 2012d).

In recent years, the total fertility rate, which measures the average number of children that would be born to a woman during her reproductive years, has fluctuated between 1.1 and 1.4, well below the replacement rate of 2.1 (Government of Singapore, 2012d).

Members benefit from the government’s guaranteed interest rate on CPF balances. This benefit has not been taken into account in estimating the implicit tax, and thus the net burden of the tax is likely to be lower.

The mean wage for women, as a group, was 77 percent that of men in 2011 (Government of Singapore, 2012c).

In a DC system, the impact of rapid population aging is felt when net contributions to the system decline and eventually become negative, and when asset prices change in response to the elderly selling physical and financial assets to finance consumption.

Individual members benefit from government guarantees of individual retirement accounts, because such guarantees have a market value (Lachance and Mitchell, 2003). Because CPF members do benefit from government guarantees, this benefit to members must be included in a fuller analysis of the implicit tax. However, this more rigorous analysis is not attempted here.

To the extent that a positive relationship exists between income and longevity, social pensions that are paid until an individual’s death could benefit higher-income groups for a longer period. However, this inequity can be partially addressed by income and asset testing of beneficiaries.

A government report published in January 2013 projected the share of citizens in Singapore’s population to be between 58.3 percent and 62.0 percent in 2020, and between 52.1 percent and 58.4 percent by 2030 (Government of Singapore, 2013). Because the social pension is targeted only at citizens, if their share in the population declines, then the fiscal cost of the scheme will be lower than projected in this chapter. Thus, if the share of citizens in 2030 is assumed to be 55 percent, the fiscal cost of social pensions—assuming a benefit equivalent to 30 percent median annual wage income—will be 2.18 percent with a range of 1.80 to 2.59 percent of GDP in 2030.

See Government of Australia (2013) for more information. The conversion from Australian dollars to Singapore dollars assumes an exchange rate of 1.3.

    Other Resources Citing This Publication