Issues in pension system reform in India   Ajay Shah Indira Gandhi Institute for Development Research Bombay, India August 1, 2000 Abstract ¡ Email ajayshah@igidr.ac.in, and URL http://www.igidr.ac.in/ ajayshah ¢ on the web. I am grateful to Robert Palacios, Gautam Bharadwaj, Renuka Sane and Susan Thomas for ideas and assistance. 1 Contents 1 Background 3 2 Existing programs in the pension sector 4 2.1 The civil servants pension . . . . . . . . . . . . . . . . . . . . . . 4 2.2 Programs run by the Employee Provident Fund Organisation (EPFO) 5 3 Problems 7 3.1 The civil servants pension . . . . . . . . . . . . . . . . . . . . . . 7 3.2 Programs of the EPFO . . . . . . . . . . . . . . . . . . . . . . . 7 3.2.1 Governance and administration . . . . . . . . . . . . . . 7 3.2.2 Poor rates of return . . . . . . . . . . . . . . . . . . . . . 8 3.2.3 Concerns about EPS being underfunded . . . . . . . . . . 8 3.2.4 Poor accumulation . . . . . . . . . . . . . . . . . . . . . 8 3.2.5 Premature withdrawal . . . . . . . . . . . . . . . . . . . 8 3.2.6 Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . 9 4 Issues in reforms 9 4.1 Universal, mandatory system . . . . . . . . . . . . . . . . . . . . 9 4.2 Individual accounts . . . . . . . . . . . . . . . . . . . . . . . . . 10 4.2.1 Political risk . . . . . . . . . . . . . . . . . . . . . . . . 10 4.2.2 Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . 10 4.2.3 Administrative costs . . . . . . . . . . . . . . . . . . . . 11 4.3 Role for defined benefits . . . . . . . . . . . . . . . . . . . . . . 11 4.4 Premature withdrawals . . . . . . . . . . . . . . . . . . . . . . . 12 4.5 Mandatory annuitisation . . . . . . . . . . . . . . . . . . . . . . 12 5 Reform proposals 12 5.1 A new voluntary, individual account pension system . . . . . . . . 13 5.2 Reforms to EPF . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 5.3 Reforms to EPS . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 5.4 Reforms to the Civil servants pension . . . . . . . . . . . . . . . 14 2 1 Background The problems of income security in old age in India are placed in a somewhat unusual setting, when compared with the experiences of the countries which em- barked on pension reforms in the recent twenty years. These initial conditions and constraints are important in understanding the present state of the pension system and its problems. They also have significant implications for the evalua- tion of alternative reform strategies. Hence, in this section, we take stock of this backdrop. Demographic transition India is the last major country in the world to experience the demographic transition; a sharp decline in TFRs only commenced in the late 1980s. In 2016, only 8.9% of the population is expected to be above 60; this fraction is expected to rise to 13.3% in 2026 (Visaria 1998). Policymaking and administrative capabilities of the State The governance capacity of the State is highly limited. In numerous areas, the State is translating significant inputs in terms of financial resources into poor outcomes in terms of the produc- tion of public goods. The process of policy formulation is often captured by special interest groups, or adversely influenced by corruption. From a more practical perspective of the pension system, there are significant hur- dles in the face of a nationwide pension system. For example, there is no universal citizen identification number. Fiscal problems of the State From the late 1980s onwards, the fiscal deficit has wors- ened, to a point where the aggregate deficit of the centre and the states is now around 10% of GDP. This puts severe constraints in the face of transfers from the State to the elderly. Financial sector development The equity market has made major gains in recent years (Shah & Sivakumar 2000). There is time-series evidence over 21 years suggesting that the equity premium has proved to be around 8 percentage points. There is a modern stock market index (Shah & Thomas 1998) with four index funds. Trading in index derivatives commenced in India in June 2000 and will start in Singapore in August 2000. The market capitalisation of the equity and bond markets are around Rs.10 trillion each. The stock of assets of the pension sector today is around Rs.1 trillion; hence there is no short–run constraint in terms of the ability of the equity or bond mar- kets to absorb pension investments. Hence, equity investments, especially through index funds (Shah & Fernandes 1999), are a viable strategy for the pension sector. Unorganised labour market In the 1991 census, 53% of the labour force were self– employed, and 31% were casual labour or contract workers. Only 15% of the labour force had stable, salaried jobs. 3 Table 1 The civil servants pension program (1997) Employees Pensioners Number Expense Number Expense (million) Rs. bln % of GDP (million) Rs. bln % of GDP Central Government 4.648 340.9 2.25% 3.603 111.4 0.73% Railways 1.564 107.4 0.71% 1.060 35.1 0.23% Post and Telegraph 0.852 57.4 0.38% 0.235 7.5 0.05% Defense 1.036 87.8 0.58% 1.844 49.5 0.33% State Government 7.600 396.1 2.61% 3.672 86.1 0.57% Total 12.247 737.0 4.86% 7.275 197.5 1.30% Small pension system Old age security in India, today, is dominated by private transfers. The fraction of the elderly, today, receiving pension is below 10%. Market for annuities There is one public-sector monopoly in the life insurance area. As of 1998, it had sold 670,000 annuities. There are no variable annuity products, or inflation-indexed annuities. Hence, we may say that the market for annuities is a fledgling one. Starting in 2000 AD, private entry into life insurance will take place. Context of mass poverty Depending on how poverty is defined, between 20 to 30 per- cent of India’s population is in a state of poverty, and has zero savings. This limits the extent to which workers can make contributions into the pension system. If, as experienced in India in the decade of the 1990s, a real GDP growth of 6.5% per year is used for extrapolation into the future, then poverty amongst the elderly may become the central issue in the analysis of poverty in India within fifteen years. 2 Existing programs in the pension sector There are two important components of the pension system today: (a) an un- funded, defined-benefit pension program for government employees, and (b) a pair of defined–benefit and defined–contribution programs which are mandatory for establishments above twenty workers. We summarise the basic facts about these programs in this section. 2.1 The civil servants pension The civil servants pension is an unfunded, defined benefit program (Asher 2000). The benefit rate is roughly 50% of the terminal wage. 4 Table 2 Growth of payments under civil servants pension system Component Growth rate (%) Central pension (94-95 – 00-01) 27.9 Defence 28.1 States (93-94 – 99-00) 25.6 Bihar 38.4 Maharashtra 35.5 Rajasthan 34.0 Uttar Pradesh 26.8 Tamil Nadu 23.0 Andhra Pradesh 22.0 Kerala 22.0 Source: CMIE (2000) . The size of the civil servants pension system is summarised in Table 1. The growth of payments on account of this system is summarised in Table 2. The growth of the pension outgo for all states, taken together, masks significant inter– state heterogeneity. For the major states (defined as states with an outgo in 1999- 2000 of above Rs.15 billion), we see values ranging from 38.4% growth for Bihar to 22% growth for Kerala. 2.2 Programs run by the Employee Provident Fund Organisa- tion (EPFO) There are two major programs which are run by the EPFO: they are the Employee Provident Fund, based on legislation enacted in 1952, and the Employee Pension Scheme (EPS), based on legislation enacted in 1995. They are mandatory for establishments with more than twenty employees. Most of India’s labour force is in establishments with less than twenty employees. The fraction of the labour force in these programs has steadily risen from 1% of the labour force in 1953 to 5% in 1995. The contribution rates for the two programs are summarised in Table 3. The EPF is a DC program; the participant receives a lump-sum upon retire- ment. The EPS is a DB program. It yields a benefit rate of 50% of the terminal wage, for individuals who have contributed for above ten years. The EPFO outsources fund management to a single fund manager. The rules under which this fund manager operates under are highly restrictive. Table 4 shows the required asset allocation. In addition, there are other constraints (e.g. 5 Table 3 Sources of contributions into EPF and EPS All contributions are expressed as percent of “basic wage”. Over recent decades, the labour market has steadily reduced the size of basic wage expressed as a fraction of the total wage. Hence, these values overstate the true contribution rates as percent of the overall wage bill. The values shown pertain to the rules that apply for 172 industries. There are five industries where slighly different rules apply. Source EPF EPS Individual 12.00 Employer 4.77 8.33 Government 1.16 Total 16.77 9.49 Table 4 Regulation of investments for EPF (i) Central government bonds 25% (ii) (a) State government bonds, (b) Government guaranteed bonds 15% (iii) (a) Securities of Public Financial Institutions, Public Sector 40% Enterprises, Banks and Infrastructure Development Finance Co., (b) CDs issued by public sector banks (iv) Above three categories as determined by Board of Trustees 20% (v) Private sector bonds with investment grade credit rating from 10% two credit rating agencies (from (iv) above) 6 bonds can be purchased but not sold). There is no competition in fund manage- ment, and participants are presented with no asset–management choices. The EPFO itself is governed by a 30-man board of trustees. Of these, 15 are representatives of organisations of employers, and 15 are representatives of trade unions. 3 Problems In this section, we summarise the difficulties faced with the civil servants pension and the programs of the EPFO. 3.1 The civil servants pension The pension outgo on account of the civil servants pension is significant (1–2% of GDP) when compared with the fiscal deficit. Over the period for which data is seen in Table 2 (1993-94 to 2000-01), inflation has averaged 5%. Hence, the growth rates of the pension outgo, of above 20% per annum in real terms, is extremely large. The pension outgo of the centre and states is expected to be larger than the wage bill within three to four years. The information systems and administration of the civil servants pension leave a lot to be desired. There is no central database about pension recipients. Basic facts such as the age distribution of pensioneers are not known. Concerns have been expressed about pension payments being made to many who are deceased. These difficulties with governance and administration are important in under- standing the problems of the civil servants pension. We see a fragment of this in Table 2. Three of the four states with above–average growth in the pension outgo are drawn from the bimaru set of poorly governed North Indian states. The three states with below–average growth are the relatively well governed South Indian states. It is useful to observe that the bimaru states have lower levels and rates of change of life expectancy when compared with the South Indian states. This should normally result in slower growth of pension payments. 3.2 Programs of the EPFO 3.2.1 Governance and administration The governance of the EPFO is seriously flawed in terms of skills and incentives. This hurts policy formulation of the EPFO. The programs of the EPFO offer poor customer service. Workers get no state- ments and face no choices. There are often delays of over a year in paying out 7 benefits. 3.2.2 Poor rates of return The portfolio of the EPFO is essentially formed of government securities; only 5.5% of the assets are held in something slightly non-government (bonds issued by public sector firms). Hence, the long-run rate of return is close to the growth rate of wages, and no risk premia accrue to the pension assets. 3.2.3 Concerns about EPS being underfunded An actuarial evaluation of EPS was done in 1996. Under its assumptions, there was no serious funding gap for EPS. However, a thorough analysis of EPS is not available, atleast in the public domain. Concerns have been frequently expressed about EPS being underfunded. This problem should be seen in a dynamic context. Given the governance problems of the EPFO, there is a risk that at a future date, there could be politically motivated decisions to raise benefits or drop contributions. 3.2.4 Poor accumulation Viewed as a blackbox, the EPF program does a poor job of accumulating retire- ment wealth. The average assets of the worker who exits from the system are around Rs.25,000 or $600. This appears surprising given the fairly high contribu- tion rates. 3.2.5 Premature withdrawal The main problem with the EPF appears to be premature withdrawals. Only 17.7% of the disbursements in 1995-96 were on account on retirement, illness or death. The remaining 82.3% were premature withdrawals. The mean value of premature withdrawals per participant works out to $400 per year, which is quite large when compared to the mean lumpsum paid out upon retirement of $600. Of these, the major source of withdrawals seemed to take place upon resigna- tion. There are two factors at work here. Sometimes, a participant moves into a job at an establishment which is not covered by the EPFO. Alternatively, a partic- ipant moves from one covered establishment to another, but the EPFO is unable to link up the records for the two. In this case, the EPFO pays out the accumu- lated assets of the previous job to the participant, and the accumulation of pension assets commences afresh at the new job. The tax treatment of the EPF involves tax-exemption of both contributions and withdrawals from EPF. In combination with the low rates of return of the 8 EPFO, these factors give strong incentives for individuals to engage in tax arbi- trage through withdrawals. 3.2.6 Coverage The mandatory programs of the EPFO only cover roughly 5% of the labour force. This suggests that the bulk of India’s problem in the area of old age income secu- rity is not being addressed by the programs of the EPFO. 4 Issues in reforms With this background, we can evaluate some of the issues in pension system de- sign in India. 4.1 Universal, mandatory system Should India adopt some kind of universal, mandatory pension system? Whether or not some such system is desirable, there are acute constraints in the face of such a strategy: £ Such a system is likely to involve a central role for the State in policy formulation, on issues such as defining rules about contributions, benefits, asset management, etc. There are significant political risks that flow from placing such demands upon the State in India. £ The administrative capacity to collect contributions from the 350 million work- ers, spread all over 3.3 million square kilometres, does not exist. For instance, at present, there is no concept of a unique citizen’s identity number. The EPFO is fairly successful in enforcing the rule which requires that each establishment with more than 20 workers should be connected up with the EPFO; yet it only covers 5% of the labour force. £ The administrative capacity to correctly pay out benefits does not exist. There is a significant risk of fraud in payouts: paying out benefits to dead people, paying benefits multiple times to some individuals, etc. If we rule out a universal, mandatory system, the alternatives that remain are a combination of a mandatory pension system for employees of large establish- ments, and a voluntary individual account system. 9 4.2 Individual accounts We now evaluate an individual account system without defined benefits, which places significant choices in the hands of participants on questions such as risk exposure or choice of fund manager. 4.2.1 Political risk One important appeal of individual accounts is the sense in which individuals interpret their account balances as personal wealth. This reduces the free rider problem, and encourages individuals to take interest in questions of governance and policy formulation in the context of the pension system. As long as individ- uals do face choices on the questions of risk exposure and the choice of the fund manager, individual accounts appear to require reduced inputs in terms of sound governance. India’s experience with the EPF program suggests that simply having defined contributions with an individual account does not suffice in obtaining sound gov- ernance. India’s experience is a positive one in the sense that the long–term real return on EPF has not been negative, as compared with some other countries where extreme failures of governance of pension assets have generated sharply negative returns (Iglesias & Palacios 1999). However, the structure chosen in India’s EPF is clearly inadequate: individuals are powerless, fund management practices are clearly not in the interests of participants, and the long-run historical real rate of return has proved to be below 3%. 4.2.2 Knowledge The most important concern with individual accounts, and the idea of placing critical asset management choices in the hands of individual, is the question of fi- nancial literacy. Individuals may face two kinds of choices: between asset classes, and between fund managers. £ Choosing asset classes. There is a real risk that individuals may choose to ex- pose themselves too little to risk factors and thus obtain poor consumption in old age.1 This issue has a peculiar twist in an environment with mass poverty: given sufficiently small contributions, high investments in equity offer the highest proba- bilities of avoiding poverty in old age (Thomas 1998). £ Choosing between fund managers. The same question of knowledge in the hands of individuals is also important when it comes to making choices between fund 1 Several recent papers have looked at this question in the context of industrialised countries (Whitehouse 1999, Lillard & Willis 2000). These studies find that individuals do seem to often lack the knowledge that is required in consciously exposing themselves to risk factors. 10 managers. Researchers equipped with econometric models have found it difficult to choose a fund manager who fares well, out of sample. Investors are often seen as not being sensitive to expenses and fees on the part of fund managers. In countries such as Chile and Argentina, there have been concerns about excessively high sales expenses by pension fund managers. In this situation, there may be a case for the pension system to be designed in a way which helps curtail fees and expenses. Proposals in this direction include an auction to choose fund managers who offer the lowest consolidated fees and expenses, and mandatory indexation. 4.2.3 Administrative costs The third major issue in an individual account system, particularly when contri- butions or account balances are small, is the question of administrative overhead and transactions costs (Whitehouse 2000, Murthi, Orszag & Orszag 1999). These questions are particularly important in the Indian setting, where the average con- tribution and the average account balance would be amongst the smallest in the world. James, Smalhout & Vittas (1999) suggest that the most important vehicles for keeping costs low are: (a) constrained portfolio choice, using passive manage- ment, and (b) reduced sales expenses. 4.3 Role for defined benefits There are two aspects where an element of a DB system is attractive; both pertain to the extent to which exposure to risk factors is adopted. The first issue is the problem of the investment risk that has to be borne by participants in an individual account system (Alier & Vittas 1999). A DB system may offer a mechanism for risk sharing, thus reducing the risk borne by an individual. The second issue is about decision making in fund management. A DB system is likely to place decisions in the hands of finance professionals, who would be less likely to avoid risk factors when compared to the decision-making of many individuals. These are important arguments; however they pressupose an environment in which a DB system is run with sound policy formulation and governance. This capacity for governance is suspect in India. The two DB systems that exist in India today are unsuccessful in both aspects: they either use no assets (the civil servants pension) or investment in government bonds (EPS). Further, it is possible to envi- sion political pressures in favour of generous benefits and low contributions, and policy formulation which imposes liabilities upon the State. For these reasons, it seems desirable to avoid DB systems in thinking about the pension system in India. 11 There is another dimension in which simplistic DB systems (such as India’s EPS) can yield unsatisfactory outcomes: this is on the question of regressive trans- fers. The value of an annuity varies strongly with wealth and education, so that the NPV of a stream of benefits is the highest for the wealthy and well educated (Lillard, Brien & Panis 1993). This is particularly the case in India, where life ex- pectancy for the rich is comparable to that seen in industrialised countries, while overall life expectancy at birth is just 65 years. Hence, many DB designs can be- come a mechanism for transfers from the poor to the rich, which is, in general, an unsatisfactory outcome. 4.4 Premature withdrawals The illiquidity of pension assets is a constraint, in the eyes of participants in a pension system, in an environment where credit is constrained. For many individ- uals who are potential participants in a pension system in India, credit markets are inaccessible when exposed to consumption shocks. If the pension system offers no possibility of premature withdrawals, then it becomes relatively unattractive (Walliser 1999). In India, we have one extreme example in the EPF experience. EPF combines poor fund management with a peculiar tax treatment (contributions, asset returns and premature withdrawals are all tax–free). This yielded an outcome with a high rate of withdrawals. This is a polar case which should clearly be avoided. Yet, the design of a pension system for India should seek to avoid complete illiquidity of pension assets. 4.5 Mandatory annuitisation In many countries, pension regulations require a certain degree of mandatory an- nuitisation. This is partly motivated by the existence of an extensive safety net. In- dividuals who run through their pension wealth “too quickly” can count on falling back upon poverty alleviation programs which are run by the State. This generates private incentives to annuitise too little. In India, the poverty alleviation program which is run by the State pays a subsidy of $2 per month. Hence, this is not a concern which should motivate mandatory annuitisation. 5 Reform proposals In 1998, the Ministry of Social Justice and Empowerment created the Dave Com- mittee to take a fresh look at India’s pension system. The proposals of the Dave 12 Committee (Dave 2000) are summarised in this section. These proposals are presently being evaluated by the political process. 5.1 A new voluntary, individual account pension system A new individual account pension system is proposed. It would have the following features: 1. Participation and contributions would be voluntary. It is targeted at the large mass of individuals outside the organised sector in India. 2. Each individual would have a single account number. The system would offer “points of presence” through which this account could be operated, all over India. The report goes to great lengths to sketch the information technology and distri- bution system which would be used to achieve this. The report envisages a single centralised depository where records would be centralised. Instructions of indi- viduals would all be funnelled to the depository, which would issue consolidated instructions to fund managers. This is expected to yield major gains in reduced administrative costs. 3. There would be exactly six fund managers and three fund management styles, giv- ing 18 investment alternatives in all. Individuals would be free to allocate their pension assets between these ¨¦¤ § ¥ alternatives. The report takes special efforts to reduce the information processing and knowledge required in making performance comparisons between these ©¤ § ¥ choices. 4. The six fund managers would be selected using an auction in which the bidders who quote the lowest consolidated fees and expenses would be selected for a period of five years. The fund managers would not need to interact with participants for contributions or withdrawals; they would have a single consolidated transaction with the depository every day. This would reduce their overheads. 5. The three fund management styles range from zero equity exposure to high equity exposure. All equity investment is to be done using index funds only. 6. There is a provision of a “micro–credit” facility for small loans, using pension assets as collateral. 7. Upon retirement, annuitisation is mandatory to a low extent (which yield consump- tion which is twice of the poverty threshold). Beyond this, the individual would receive a lump-sum. 5.2 Reforms to EPF The report proposes the following changes to EPF: 13 1. The provision for premature withdrawal should be withdrawn. 2. Participants should have the choice of moving their EPF balances into the new individual account system described above. 5.3 Reforms to EPS The report proposes the following changes to EPS: 1. The contribution of 1.16% of the basic wage which comes from the government should cease. 2. The assets of EPS should be professionally managed, using the investment guide- lines proposed for the new individual account system. 3. An actuarial valuation of EPS should be done every year. The report should be publicly released. The benefits or contributions into EPS should be adjusted every year in response to this report. 4. EPS today is based on outsourced fund management. This outsourcing should extend to the purchase of annuities also. 5.4 Reforms to the Civil servants pension The report proposes the following changes to the civil servants pension system: 1. Better information gathering, and 2. A shift to a fully funded system within a decade. 14 References Alier, M. & Vittas, D. (1999), Personal pension plans and stock market volatility, in ‘New ideas about old age security’, World Bank. Asher, M. G. (2000), Reforming civil service pensions in selected Asian countries, Technical report, National University of Singapore. CMIE (2000), Public Finance, Centre for Monitoring Indian Economy, Bombay, India. Dave, S. A. (2000), Project OASIS Report, Committee report, Ministry for Social Justice, Government of India. Iglesias, A. & Palacios, R. J. (1999), Managing public pension reserves: Evidence from the international experience, in ‘New ideas about old age security’, The World Bank. James, E., Smalhout, J. & Vittas, D. (1999), Administrative costs and the orga- nization of individual account systems: A comparative perspective, in ‘New ideas about old age security’, World Bank. Lillard, L. A., Brien, M. J. & Panis, C. W. A. (1993), The value of annuities at age 65: Race, marital status, wealth, health and mortality, Technical report, Rand Corporation. Lillard, L. A. & Willis, R. J. (2000), Cognition and wealth: The importance of probabilistic thinking, Technical report, University of Michigan. Murthi, M., Orszag, J. M. & Orszag, P. R. (1999), Administrative costs under a decentralized approach to individual accounts: Lessons from the United Kingdom, in ‘New ideas about old age security’. Shah, A. & Fernandes, K. (1999), The relevance of index funds for pension in- vestment in equities, in ‘New ideas about old age security’, World Bank. Shah, A. & Sivakumar, S. (2000), ‘Changing liquidity in the Indian equity mar- ket’, Emerging Markets Quarterly 4(2), 62–71. Shah, A. & Thomas, S. (1998), Market microstructure considerations in index construction, in ‘CBOT Research Symposium Proceedings’, Chicago Board of Trade, pp. 173–193. Thomas, S. (1998), Alternative investment strategies on multi–decade horizons, Technical report, IGIDR. 15 Visaria, P. (1998), Demographics of an ageing India, Technical report, Institute of Economic Growth. Walliser, J. (1999), Regulation of withdrawals in individual account systems, in ‘New ideas about old age security’, World Bank. Whitehouse, E. (1999), ‘Pension reform, financial literacy and public information : A case study of the united kingdom’. Whitehouse, E. (2000), Administrative charges for funded pensions: An interna- tional comparison and assessment, Technical Report Social Protection Dis- cussion Paper No. 0016, World Bank. 16