«D I R E C T I O N S I N D E V E LO P M E N T Human Development Public Disclosure Authorized The Elderly and Old Age Support in Rural China Challenges ...»
Engen and Gale 2000). However, given progressive tax scales in most OECD countries, subsidies tend to benefit better-off workers disproportionately in absolute terms. In any event, using tax exemptions to subsidize pension contributions would likely prove less effective in developing countries, where the poor are less likely to be subject to taxes in the first place and where distributional concerns on the use of public subsidies may be more pressing. In China, a tax exemption would be meaningless for the large majority of rural workers because their incomes are below the minimum personal income tax threshold. Because the majority of rural workers do not pay income tax, the direct match approach is preferred. That option is precisely the one chosen by the Chinese authorities in the national pilot, and it seems appropriate.
Experience with MDCs in developing countries is limited. Legislation introducing MDC schemes has been passed in the Dominican Republic, Indonesia, and Vietnam, but they have not yet been implemented. A noteworthy example comes from the Indian states of Rajasthan and Madhya Pradesh, both of which have MDC pension schemes for certain categories of informal sector workers. The schemes provide a one-to-one match on contributions from the state governments. Funds are invested by contracted asset managers with no guaranteed rate of return for contributors. The state of Andhra Pradesh started a similar scheme in 2009 that targeted women in self-help groups, with a subsidy of about US$10 per year per person, which is expected to produce a pension just above the poverty line. The oldest such MDC scheme in India is in West Bengal, which has provided a one-to-one match for roughly 60 categories of informal sector workers since the early 2000s. The Indian government is considering an MDC approach for informal workers on a nationwide basis under its new pension scheme (see box 6.1).
As is the case for all defined contribution schemes, the design of MDCs needs to address two policy issues: (a) how to handle the death or disability of contributors, and (b) how to manage financial risk for contributors (particularly those with low financial literacy). With respect to the former, pure defined contribution schemes do not, by construction, provide insurance against premature death or disability during the 122 The Elderly and Old Age Support in Rural China
Rajasthan’s Vishwakarma MDC Pension Scheme In 2007, the government of Rajasthan, India, introduced an MDC scheme for 20 categories of low-income workers, almost all of whom are in the informal sector. To be eligible, workers must (a) be residents of Rajasthan, (b) be between 18 and 50 years of age, and (c) not be covered by any other provident fund arrangement supported by the government or an employer. If workers contribute 100 Indian rupees (just over US$2) per month for at least 10 months each year, the state government will finance a 1,000-rupee annual match on their contribution. An annual interest rate is paid on the combined accumulation; the interest rate is announced each year and is equivalent to the rate of return on the formal sector provident fund (about 8 percent per year in 2008). This computation has been designed to provide a benefit at retirement just above the poverty line. At current interest rates, a 30-year-old worker who contributes for 30 years would receive a monthly pension of about 2,000 rupees at retirement at age 60. The scheme does not allow early withdrawals of accumulations. Account holders are provided an annual account statement. Web-based individual accounts are opened for each worker in the scheme, and a computer-generated passbook is provided within 30 days, which contains a scanned photograph and the enrollment form of the account holder. The individual receives a unique identifier at the point of registration, using an application called the Social Security Solution (sCube), which allows portability of the account and permits the worker to make contributions at any location in the state. Using a variety of options, sCube allows online and offline data entry and requires only a half day of training to operate. Raising the awareness of members is strongly emphasized before and following enrollment, using short films, comics, interactive pension calculators, and other approaches.
Source: Invest India Micro Pension Services Pvt., Ltd., http://www.iimp.in.
accumulation phase. As a result, group life and disability insurance policies are common for contributors in the scheme. Such policies would cover the difference between the individual’s account accumulation at the point of death or disability and the benefits paid under the policy in the event of death or disability. The premium is calculated using actuarial statistics for the group. Managing financial risk is handled differently in different countries, ranging from imposing conservative portfolio rules (which is the case in China now) to guaranteeing rates of return to transferring all risk to contributors (which is the case in the Indian schemes Issues for the Evolution of the Rural Pension System 123 previously described). Given the likely degree of risk aversion on the part of rural workers in China and the country’s underdeveloped regulatory framework, offering a guaranteed rate of return, as suggested above, seems to be the preferred option for the medium term.
Could the Basic Rural Pension Benefit Evolve into a Social Pension?
A key policy question for the future development of the rural pension system is whether the basic benefit should in time evolve into a “social pension” for which eligibility is not linked to contributions to an individual account, as happens now. International experience reviewed later in this section suggests that full pension coverage is very difficult to achieve at the level of the individual in rural and informal sector schemes based on approaches that require contributions, even with incentives to contribute.
The objective of a social pension would be to ensure basic subsistence for those elderly who are not covered by existing pension provisions or who are unable to generate adequate retirement income from their contributions during their working age. An inability to generate adequate retirement income could occur for a variety of reasons, including sickness, disability, or time out of the workforce (for child rearing, study, and so on). The rationale for a social pension is twofold: (a) to provide a low level of support for the growing number of elderly with minimal resources to prevent them from falling into poverty in their old age, and (b) to reduce incentives for overly high precautionary saving before retirement.
A social pension could follow uniform design parameters nationwide, although benefits would reflect local characteristics. Provincial authorities would be held accountable for observing the national framework. The benefit amount could be set as a percentage of the regional average wage, with the objective of exceeding the per capita benefit under the existing dibao scheme. A “pension test” could also reduce the benefit by a proportion of the benefits received from other pension arrangements, say for those 65 to 74 years of age. The pension test would need to be designed to reward those who contributed to other pension schemes while targeting those with the least income from all sources, including other pension benefits. Given the need to eventually increase the retirement age to 65 for men and women, setting 65 as the minimum age for both men and women to qualify for a social pension seems appropriate. Doing so is 124 The Elderly and Old Age Support in Rural China important from a labor supply incentive viewpoint, although it would need to maintain consistency with urban pension policy. Such a benefit would be noncontributory and financed by national, provincial, municipal, and local resources.
A social pension approach would be broadly consistent with the design of the basic benefit provision under the rural pension pilot. As noted above, any social pension should be subjected to a pension test, which would adjust the amount paid to those 65 to 74 years of age by a proportion of benefits they receive either from the individual account or from an urban workers’ pension. The adjustment factor could be set as high as 50 percent initially (to provide a strong incentive to contribute to the individual account), but it should be adjusted on the basis of actual experience. A stylized example is provided in figure 6.3, using a benefit level linked to rural individual minimum consumption and applying a 50 percent adjustment factor.
As with the basic benefit under the national pilot, the value of a social pension relative to the dibao threshold in different localities is an important design parameter. In this regard, the practice of most current rural pension pilot schemes seems appropriate (that is, setting the social Figure 6.3 Stylized Example of Rural Pension Benefit Levels and Composition under Proposed Scheme
1,400 1,200 1,000 1, 1, 1, 1, 1, 1,
Source: World Bank forthcoming.
Note: Assumes social pension benefit of 788 yuan per year minus the Ravallion-Chen rural consumption poverty level increased by the CPI from 2003 to 2009. In this example, social pension benefit is reduced by 50 percent of alternative pension income.
Issues for the Evolution of the Rural Pension System 125 pension at a level above the dibao per capita threshold). This practice is important for incentive reasons, but—for fiscal reasons—the level should not be substantially higher. If a social pension were set too high, the incentive to contribute to the individual account would be weaker; if it is set too low, its poverty alleviation objective may be undermined. Under the rural pension pilot, the central government has set the minimum flat benefit at RMB 55 monthly while allowing provision of additional funds from subnational sources. This minimum flat benefit compares with a national average rural dibao threshold of RMB 82 in 2008, with a range from a minimum of RMB 26 to RMB 267 (or 11 to 41 percent of the national rural average wage). The dibao threshold rates suggest that closer attention needs to be paid to the relative level of the social pension floor to align protections and incentives appropriately. Finally, pensions from current pilot rural pension schemes and any possible future social pension should not affect the receipt of the allowance for following family planning policies. However, some Chinese researchers have proposed integrating the programs by using the family planning subsidy as an additional subsidy toward individual pension contributions (Mi and Yang 2008; C. Yang 2007; Y. Yang 2007).
Indicative cost estimates suggest that a social pension set around the urban income poverty line would be about 0.11 percent of 2010 GDP, rising to 0.31 percent of GDP by 2040 (see figure 6.4).10 A social pension set around the rural poverty consumption line would be about 0.13 percent of 2010 GDP, declining to about 0.12 percent of GDP by 2040.
These estimates would vary substantially with the benefit level, the growth in the benefit level, the level of urbanization, and the observed benefit reductions arising from the pension test. The upward trend in cost projections reflects population aging and urbanization. By comparison, if the social pension were set at about 28 percent of the urban average wage (the OECD average), the cost would be 0.75 percent of GDP in 2010 and would rise to 2.10 percent of GDP by 2040.11 Although the higher figure is probably more than China may wish to allocate to such a benefit, the experience of several OECD countries noted below suggests that a benefit level somewhere between the lower bound and the OECD average could be affordable.
Countries with all levels of income have widespread experience with noncontributory social pensions. The experience with reducing old-age poverty is generally positive, although issues of fiscal sustainability have arisen in lower-income settings. However, social pensions are proving to be an increasingly popular method for bridging the old-age coverage gap in pension systems.12 126 The Elderly and Old Age Support in Rural China Figure 6.4 Indicative Cost Projections for Urban and Rural Social Pensions 0.45 0.40
0.30 0.25 0.20 0.15 0.10 0.05
Source: World Bank staff estimates.
Note: Assumes (a) median variant urbanization 2010–40; (b) total fertility rate of 1.8; and (c) the age distribution for the population over age 60 is the same as for the total population. The rural reduction from the pension test and transition introduction is 5 percent in 2011, rising by 3 percent each year to a maximum of 50 percent. The urban reduction from the pension test is 40 percent in 2010, rising by 1 percent each year to a maximum of 60 percent.
The benefit level is assumed to grow at the rate of GDP.
OECD countries have different approaches to social pensions (see table 6.1).13 Although about half the countries rely on a single approach to social pensions, the remainder rely on combinations of three basic
• The basic pension, often called a “demogrant,” is a flat benefit for the elderly awarded independently of income. Countries with basic pensions usually have some qualifying provisions, such as residency or contribution tests.
• A resource-tested social pension is provided either through a separate program for the elderly or as part of a general social assistance scheme.
Eligibility is subject to some form of means testing, which can include income or income plus assets.
• The minimum pension is similar to a resource-tested pension in that it targets older people with lower incomes. The key difference is that only income from pension schemes is considered when calculating entitlement to a minimum pension. Consequently, someone with substantial Issues for the Evolution of the Rural Pension System 127
income from nonpension sources can still qualify. In countries where a contribution history is required to qualify for social pensions, periods of unemployment and disability are often counted toward the qualifying period. As a result, in practice, most people qualify.