A STUDY ON SOCIAL SECURITY IN GHANA BY PROF. KOFI KUMADO FACULTY OF LAW UNIVERSITY OF GHANA LEGON & DR. AUGUSTINE FRITZ GOCKEL DEPARTMENT OF ECONOMICS UNIVERSITY OF GHANA LEGON FEBRUARY 2003

A STUDY ON SOCIAL SECURITY 1 Introduction: The Need For Social Protection In a world of stupendous technological and medical advances, social protection has become mankind’s pre-occupation due to the vagaries of life. Well before the advent of Europeans, and the rise of modern states, traditional African societies realized the need to put in place measures to cater for the major contingencies through collective security and mutual help to one another. Until recently, the extended family was the institution in Ghanaian communities that provided social and economic support to various family members at the appropriate times of need. Traditionally, the family was the critical focus in the provision of support when members become old and are threatened by economic deprivation, disability, and social isolation. In appropriate cases, the community as a whole provided the social net for those aged without a family. In fact, it was legend that traditional extended family practices transcended socio-economic protection to offering psychological stability and moral upliftment. In contemporary times and especially with the decline of the extended family, there is a gradual shift away from primary reliance on the extended family towards dependence on more semi-formally institutionalized social security systems. The pressures created by the promotion of economic growth and social mobility together with the severe resource constraints confronting traditional systems, are all putting strains on the extended family as an effective cohesive unit that provides income security for the aged and the disabled, care for the sick and unemployed members of the family, the new born child and the mother, the orphan and even the complete stranger. The breakdown in traditional social protection schemes has been compounded by the new economic order in the 1980s based on market-led strategies to economic growth. The main thrusts of the new economic order that have affected the level of social protection include the following: • Minimalist state intervention in economic activity with policies of deregulation and divestiture of state-owned enterprises, • Stabilization and structural adjustment policies, • Liberalization of both domestic and foreign trade that entailed removal of price and

distribution controls by franchise, • Removal of subsidies under incomes reform to reward performance consistent with allocative efficiency policies and cost-effectiveness in the production and delivery of goods and services, especially those produced by the public sector, • Exit policies such as redeployment/retrenchment in the public sector, wage restraint, revision and transfer payments. 2 Macroeconomic Environment And The Engendered Social Protection Needs The implementation of the economic policy reforms outlined above have undoubtedly reversed the earlier years of economic decay. Although the economy struggled with chronic inflation over the 1996-2001 period, economic growth was positive through out. However, when population increase of 2.6% per annum is netted against the average 4.3% GDP growth rate, per capita GDP growth becomes marginal. Table 1 shows the trends in GDP growth, inflation and exchange rate between 1996 and 2001. Table 1 Trends In Real GDP Growth Rates, Inflation And Exchange Rate: 1996-2001 Year Projected Realized Inflation Actual Rate Cedi/Dollar Rate of Real GDP GDP Target% Inflation% Exch. Rate Depreciation 1996 Growth Rate5 Growth Ra4t.e6 20 46.6 1740.00 20.33 1997 5.5 4.2 15 27.9 2250.00 22.70 1998 5.6 4.6 9.5 15.7 2346.00 4.10 1999 5.5 4.6 9.5 13.8 3500.7 49.22 2000 5 3.7 12.5 40.5 6886.8 96.73 2001 4 4.2 25 21.3 7100 3.1 Source: Government Budget Statements And Statistical Service Monthly Newsletter, Various Issues. The inadequacy of incomes was compounded by intractable high rates of inflation and exchange rate depreciation so that those on fixed incomes, particularly pensioners, have become worse off. For example, although inflation was brought down from its highest ever level of 123% in 1983 to 10% by end of 1992, it nevertheless remained intractable. Indeed, inflation, coupled with the steady decline of the cedi/dollar rate, which reverberates through the economy, has ever since been the bane of workers. Thus, the sluggish per capita growth tends to cause Ghanaian workers to feel less well off as time goes by. This is worse for retired workers. Consequently, the 3

emergent economic order has brought with it new social demands. Although the economic reforms appeared to have yielded positive results as the declining trends observed in key macroeconomic variables prior to its inception have been arrested, indications are that poverty levels, whether by means test or needs test, are still a major problem. In Ghana Poverty Reduction Strategy 2002- 2004, it is reported that though overall poverty level has continued to fall since 1992 from 51.7% to 39.5% with extreme poverty falling from 36.5% to 26.8% by the end of 1999,“these figures mask incidents of growing and deepening poverty and evidence of intensification of vulnerability and exclusion among some groups and in some areas, especially in the north of the country and the Central Region”. This is moreso when we note that market principles are the over-riding thrust of Ghana’s growth strategy. Indeed, the actions of the social matrix and poverty studies tend to suggest that, paradoxically, while the economy is said to be growing, the high level of poverty is due to the social costs of adjustment and failure of the anticipated trickle-down effect of the realized economic growth. The reduction of human employment in the public sector and the privatized industries under the exit policy of SAP has meant that income security has become a major concern of Ghanaians. Moreover, there is no clear cut exposition of policy as to what Government proposes to do with the large amount of labour that has become unemployed and the large number of young people who are hopeful of employment and on the threshold from the point of view of age. The current Government’s declaration of the Golden Age Of Business with the private sector as the engine of growth to absorb the retrenched and the maturing labour lacks conviction as it is not backed by pragmatic programmes. The private sector is not growing to accommodate the exit policy of market-based programmes. Paradoxically, the informal sector, the most vulnerable sector with declining social protection capacity, is the only growing sector of the economy as the outlet for survival. Indications are that, with the economic reforms, especially under the exit policies in privatization of State-Owned Enterprises and retrenchment in the public sector, some breadwinners of the family lost their income-generating activities, and as a result, their income security. Moreover, subsidies, which hitherto could be regarded as social protection measures, are being progressively removed on all goods and services including utilities. Indeed, in some instances, taxes have been levied on commodities. The notable ones are the taxes on petroleum products. These, invariably, have led to very high costs of living for most Ghanaians. Consequently, the policies under the new economic order have increased the need for social protection both indirectly by their effects, and directly by the withdrawal of social protection 4

measures such as subsidies, and price controls. It must be emphasized that whilst the traditional social protection schemes were fast breaking down, those in the formal sector who could provide the stop gap measures were increasingly threatened by the new economic order with retrenchment/downsizing or right sizing policies. In an editorial captioned STANDING UP FOR THE FAMILY, The Mirror of November 16, 2002 has this to say: In recent times a lot has been reported in both the print and electronic media about the increasing fragmentation of the Ghanaian family....Undeniably, the family in all societies and cultures plays a key role in the development of individuals who constitute the human resource base of the country. It provides the necessary environment where children can be nurtured in love, friendship and discipline. Admittedly, current economic pressures are making it imperative for both parents to work outside their homes from dawn to dusk, thereby leaving children to their own devices with virtually no adult supervision.... In days gone by parents, grand parents, aunts and other extended family members provided good childcare practices but today even grandmothers are trying to eke out a living. It must be emphasized that family and community arrangements that traditionally provided for old people and the poor are under threat as well. These traditional support networks appear to be weakening with the contemporary market- based approach to economic development world-wide so that formal retirement programs will become increasingly important. Any economic system that relies heavily on markets needs ways of providing for the people that cannot participate in the markets, especially labour market. In this case, Ghana has to evolve suitable social security schemes to protect the unmarketable people of the nation. But the question is what is already available by way of social protection? The challenge is to try to make SSNIT more responsive to its mandate and also to develop other formal systems of retirement income without accelerating the decline in informal support systems and without shifting more responsibility to government than it can handle. Some of the social protection needs engendered are those of unemployment contingencies, housing, health care and the disruption of the traditional collective social protection schemes that could take care of the aged, disabled, the sick, new born babies and their mothers, the orphan and the unemployed. Indeed, indications are that, even in the cases where some formal social insurance schemes exist, benefits provided are grossly inadequate. For instance, it is lamentable that the pension provided by SSNIT to a large number of beneficiaries is only ¢50,000 per month 5

per beneficiary. Consequently, the emergent economic order has aggravated the social protection needs. The social protection issues raised are not peculiar to Ghana, although they have been compounded by macroeconomic instability. World trends suggest that social instability is noticeable in all developing countries pursuing structural adjustment policies and all societies in transition generally. Indeed, the world trend has shown that there is considerable concern about social protection, especially pension and retirement issues, even in the developed countries. It is generally known that old age security systems are ni trouble around the world due, in large part, to declining birth rates and increased longevity. This means fewer potential contributors and greater numbers of beneficiaries will be dependent over longer periods on the benefits.. This problem is particularly acute for unfunded programmes such as Ghana’s CAP 30 and the defunct End of Service Benefits. As the population ages, formal pension programmes come under strain that can lead to increased costs, higher taxes and contribution rates. These pressures can limit private sector growth as the engine of growth in the market-led paradigm to development. Another way of dealing with the demographic pressure is to reduce benefits, an implicit recognition that pension systems’ ability to protect the old is declining. The trend seems to be accelerating everywhere. In industrialised countries, those over- 60 years of age already account for 20% of the population, compared with around 10% only fifty years ago. The 2000 population Census shows that Ghana’s percentage of over-60s is just around 5% but as in other countries, it is steadily growing. The trend in Ghana is therefore no exception. Table 2 shows payments and receipts associated with SSNIT’s pensioners and contributors. While the number of persons and the amount of payments has grown in each category over the years in the Table, there are some signs for concern. The number of pensioners is still relatively small but its relation to the number of contributors is changing: in 1996, there were 31 contributors for every pensioner; in 2000 that ratio has declined to 21 contributors for each pensioner. This is worrisome particularly since SSNIT may be reaching its limit of contribution in the formal sector. How many more workers are there in the formal sector without SSNIT coverage? Meanwhile, the number of pensioners steadily increases. There is much talk about the informal sector; SSNIT has managed to cover about 5000 persons but they are a hard sell. The pension payments reported in the Table are not purely retirement payments; they include invalidity payments as well, reflecting SSNIT’s multiple roles in the area of Social Security. 6

Table 2 Contributions And Payments To SSNIT Fund Thousands Of Persons At Year End Billions Of Cedis Received/Paid In The Year Pensioners Contributors Contributions Received** Pensions Paid* 1996 22 677 170 13 1997 27 722 211 22 1998 31 767 235 31 1999 36 808 373 44 2000 41 851 459 68 Source: SSNIT, Annual Reports * These payment totals include invalidity payments but do not include death and survivors payments ** The amounts listed represent cash payments received, not including contributions in arrears Furthermore, in low-income adjusting countries such as Ghana, where formal employment growth has been slower than the growth of the labour force, general retirement problems are magnified because of the limited means of the economy to provide for the aged. The 2000 Population Census and SSNIT data suggest that although Ghana’s population is increasing, more and more people are deprived of social protection. While the increasing trend in the labour force continued into the new millennium, formal sector employment and the contribution to SSNIT has declined persistently from 10.7% of the economically active population in 1997 to 9.3% in 2001. If the present mediocre performance of the economy continues with an annual average real GDP growth rate of about 4%, then contributions to SSNIT Fund are at a peril with disastrous consequences for social protection, both in the formal sector and also with traditional extended families. Table3 shows trends in the population, active labour force and contributors to SSNIT Fund. Table 3 Trends In Population, Economically Active Population And Contributors To SSNIT Year Population Economically SSNIT Active Percentage Of SSNIT’s 1997 17087037 7754295 726167 10 1998 17525042 7953067 767155 10.4 1999 17980693 8159847 788540 9.7 2000 18845265 8552200 790778 9.3 2001 19335242 8774557 794220 9.3 Source: SSNIT, 2002, Medium Term Strategic Plan: 2002-2006, Accra, May 2002 7

3 The Research Issues And Research Method As at now, there are three main retirement schemes namely the recently resurrected End of Service Benefits, CAP 30 and the SSNIT Fund. As we demonstrate shortly, besides the SSNIT Fund, the other two are virtually unfunded schemes with implications for sustainability. The Government of Ghana does not have the means to take on pension obligations that are not funded for its citizens. Neither have long-term savings pools developed because of persistent macroeconomic instability. Within the framework of the new economic order, these weaknesses and increasing privatization contributed to the freezing of ESB in 1990. The macroeconomic conditions that caused this payment crisis continued, and in some cases have worsened. Other key economic indicators, like total debt compared to GDP, the fiscal deficit, and the current account deficit, make it clear that the government lacks the means to commit funds for pension purposes. A fiscal deficit in 2001 of 2.8 trillion Cedis increased total national debt to approximately 62 trillion Cedis which is about 160% of 2001’s GDP of 38 trillion Cedis, clearly a ratio that cannot be prudently sustained, let alone increased. Moreover, around 40% of the current revenue budget relies on funds from donors. Accordingly, efforts have to be made to provide effective social security to a larger number of people than presently exists. But there is a long way to go to make SSNIT a true national pension provider and an effective source of long-term capital for productive investment in the economy. It is against this background that this study, at the instance of the Trades Union Congress of Ghana with assistance provided by Friedrich Ebert Stiftung, seeks to provide insight into social security and pensions delivery in Ghana. In particular, TUC has expressed concern at the apparent mis-management of contribution to the Social Security and National Insurance Trust and has noted that this does not only make the benefits under the Scheme inadequate but more importantly, puts the sustainability of the SSNIT Scheme into question. Consequently, the purpose of this study is to assist TUC make an informed contribution to the national dialogue on pension reforms towards the review of the Social Security Law to establishing an effective framework for the development of social insurance long-term savings plans in Ghana. The main activities under review according to our Terms of Reference are: • Conduct a comparative assessment of various social security systems highlighting particularly best practices • Investigate the law and practice of social security in Ghana in relation to best practices 8

elsewhere so as to bring to the fore the following issues: -Ownership and control of SSNIT -Membership of SSNIT Board -Impact of the Oath of Secrecy sworn by workers representatives on SSNIT Board -Investment Standards of SSNIT • Determine whether there could be additional benefits under the SSNIT Scheme • Make recommendations for a revised system of social security in Ghana, which is supported by legislation, collective bargaining and trade union policy. Official information and figures as published and provided by SSNIT are used to do the analysis. This is supplemented by informal interviews held with members of the SSNIT Board, key officials of SSNIT and officials of the TUC. Secondary information on best practices is also examined with emphasis on the Chilean experience as the acclaimed model in town. 4 The Social Security and National Insurance Trust- SSNIT For purposes of completeness, we examine the major provisions of the legislation that were in force before the passage of PNDC Law 247 that has made the SSNIT Scheme a pension scheme. We think this is necessary if we are to find best practices over the years as a way of blending these with best practices across countries. Although several attempts were made at providing social security to the Ghanaian worker, the first encompassing scheme was to be found in the passage of NRCD 127 in November 1972 to rectify some of the shortcomings of the 1965 Parliamentary Act 279. NRCD 127 established a body corporate, the Social Security and National Insurance Trust(SSNIT) as an independent body to administer social security schemes in Ghana. The social security scheme that emerged with SSNIT was essentially a Provident Fund Scheme under which lump sum benefits were paid to beneficiaries. In terms of coverage, Act 279 of 1965 and as amended by NRCD 127 of 1972 provided for compulsory coverage for workers in establishments that employ at least five workers. An establishment with less than five employees had the option to join the scheme, but there was no compulsion. However, the following categories of workers, although they employed more than five persons, were exempt by law from joining the scheme; a). members of the Armed Forces, the Police Service and the Prison Service; b). foreigners in the diplomatic missions; and c). senior members of the universities and research institutions. 9

The Provident Fund Scheme operated by SSNIT had a superannuation benefit and five direct benefits to cater for the following contingencies: 1. sickness; 2. invalidity; 3. death/survivor's 4. emigration; and 5. unemployment. Although ILO Convention 102 of 1952 required that countries provide nine minimum standards of social security or contingencies, Ghana’s provident fund scheme provided only five, as listed above. Indeed, the ILO’s concept of social security has as its working definition, the protection which society provides its members, through a series of public measures, against the economic and social distress that otherwise would be caused by the stoppage or substantial reduction of earnings resulting from sickness, maternity, employment injury, unemployment, invalidity, old-age, and death; the provision of medical care; and the provision of subsidies for families with children. Thus, four other contingencies mentioned among the nine ILO Convention 102 as minimum standards of social protection are not provided for by the Provident Fund Social Security Scheme. These are: • medical care(of a curative or preventive nature) • employment injury benefit • family benefit, and • maternity benefit. It is important to emphasize a fundamental point here that apart from sickness and unemployment benefits, in the cases of all other contingencies that were provided under the defunct provident fund scheme, beneficiaries received a lump sum payment made up of their actual contributions plus 3% compound interest rate. 4.1 The Thrust of the Social Security Law, 1991: PNDC Law 247 In February 1991, the PNDC Government repealed NRCD 127, and replaced it with PNDC Law 247, captioned"Social Security Law, 1991". PNDC Law 247 was an attempt to redress some of 10

the major defects of the defunct Provident Fund Scheme. In this respect, the main thrust of Law 247 is that the social security system in Ghana has been converted from the payment of lump sum benefits into a pension scheme under which periodic monthly payments are to be made to members until their death. Unlike its antecedent scheme 1 , the 1991 Social Security Scheme is founded on insurance principles by which there is an element of social solidarity with pooling of resources to meet certain contingencies. It also involves inter-generational transfer of resources. 4.2 Coverage Under PNDC Law 247 Under Law 247, coverage is more encompassing than obtained under the NRCD 127. Law 247 provides that the scheme is open to all classes of employees, both in the formal and informal sectors of the economy. Unlike provisions in NRCD 127, which exempted enterprises with less than 5 employees from coverage, the scheme under Law 247 covers even the"self-employed persons, who opt to join the scheme". Workers and employers who were members of the defunct Provident Fund Scheme became automatic members of the 1991 Social Security Scheme. Similarly, beneficiaries below 60 years who received lump sum amounts under the Provident Fund Scheme were also given the option to pay back these amounts plus some interest to qualify for benefits under the new pension scheme. The scheme does not, however, cover"officers and men of the Armed Forces and other officers as are expressly exempted by Law". As we will see later, these categories are provided for under the 1946 Pensions Ordinance, popularly known as CAP 30. Others have equally been exempted as per the Police Service Pension Law 1985 (PNDCL 126); the Legal Service Public Officer Pensions Amendment Law 1986(PNDCL 165) and Prisons Service Pension Law, 1987(PNDCL 168). To these categories of public sector workers, pension is non-contributory with defined benefits under CAP 30. The Social Security Scheme under PNDC Law 247 is to be self-financing and self-sustaining through the contributions of members. The rates of contribution remained as was under the defunct scheme, i.e., employee and employer contributing 5% and 12.5% respectively of the employee's basic monthly salary. In the case of a self-employed person, such a member must contribute 17.5% of his or her monthly earnings. Whereas the employer was enjoined by law to 1 Literally, as it existed, the Provident Fund Scheme was nothing but a compulsory savings account with a 3% compound rate of interest that did not give real rates of returns to the funds saved. By this, payments made out by the Fund were essentially from the beneficiaries own savings This and other factors engendered dissatisfaction among the rank and file of unionized labour to call for the eventual conversion of the Scheme into a pension scheme in 1991. 11

remit within fourteen days at the end of each month the contributions of workers, flexibility is provided for the self-employed to determine the period or periods of payment. This is particularly so in the case of self-employed persons with seasonal incomes. A penalty sum equal to 3% per month of the"contribution payable" is levied when contributions are not paid within the prescribed or conceded period. Whilst the defunct Provident Fund Scheme made provisions for 5 contingencies, the benefits to be provided under PNDC Law 247 are only three. These are: 1. superannuation pension; 2. death/survivors benefits; and 3. invalidity benefits. Essentially, despite the fact that market mechanism as growth milieu poses more threats to social security, Law 247 has fallen short not only of ILO Convention 102 on minimum standards of social protection, but more fundamentally, it has reduced the number of benefits provided for under the defunct Provident Fund Scheme from 5 to 3. The important question, which we shall return to in due course, is whether a self-financing social security scheme, with inter-generational contracts such as is being operated by SSNIT, can cater for other contingencies including housing without derailing the scheme! In the interim, we examine the qualifying conditions of the benefits stipulated under old age, death/survivors and invalidity. OLD AGE To qualify for old age full pension payment, a worker should have contributed to the scheme for at least 240 months, i.e. 20 years, and should have attained either the voluntary age of 55 or compulsory retiring age of 60. However, any person who has worked as an underground mine worker or in steel works or in such hazardous employment for an aggregate period of 240 months or more qualifies for full pension at age 55. The minimum pension payable is 50% of the average of the 3 best years salary for a minimum contribution period of 240 months. For any additional month served after the 240 months, a worker earns a pension right of 0.125%, i.e. 1.5% for every 12 months in addition to the 50% start off. Thus, a worker can theoretically earn up to 80% pension when he shall have worked and contributed to the scheme for 40 years. All other persons who opt to retire at age 55 or before 60 years are entitled to a reduced pension. That is, voluntary retirement is possible from age 55, with a reduced pension computed as a 12

percentage of the full pension, depending on the age of the beneficiary as follows: age 55, 60%; 56, 67.5%; 57, 75.0%; 58, 82.5%; 59, 90.0%. Thus, a reduced pension is calculated as 60% of a full pension plus 7.75% for any additional year served until the age 59 when the reduced pension would have worked out to 90% of the full pension. While the pension is paid monthly, one may opt to collect an advance payment, 25% of his 12 years pension in a lump sum and subsequently be on a reduced pension. However, if a contributor were not able to work up to 240 months before he retires either voluntarily or compulsorily, he would be entitled to receive his actual contribution plus interest at half the prevailing Government Treasury Bill Rate. In this latter case, the scheme reverts to a compulsory saving mechanism without the capacity to earn the full rate of return as persons who by themselves have invested in treasury bills or other high yielding assets. DEATH/ SURVIVORS With regard to survivors benefit, if a contributor dies while still a member, his dependants qualify for a lump sum of the earned pension. When a member contributes to the Fund for 240 months before dying, a lump sum equal to the value of his pension for 12 years shall be paid to his survivors. If a member dies without having contributed to the fund for 240 months, the payment to his survivors will be his proportional pension for a period of twelve years. Where a member who has retired dies before he is 72, his survivors will be paid in lump sum the unexpired pension up to age 72. INVALIDITY PENSION To qualify for invalidity pension, a member shall have contributed to the Fund for 12 months within the last 36 months before becoming invalid. In addition, the member should have been certified permanently invalid and incapable of gainful employment by a medical board including a medical practitioner appointed by SSNIT. 5 The National Pension Scheme- CAP 30 Although the SSNIT scheme is the basic National pension scheme for all Ghanaians, it is only so in name as there is a parallel pension scheme that applies to some public sector employees. Known as CAP 30, it pre-dates any existing social security scheme and derives its name from Chapter 30 of the Pension Ordinance of 1946. This Pension Ordinance served as the“ pension right” for pensionable officers in the Civil Service and the Armed Forces until the promulgation and coming into force of the Pension and Social Security Amendment Decree 1975(SMCD 8), and the subsequently PNDC Law 247. CAP 30 has presumably been closed to new entrants from 13

1 January 1972. However, the Decree provided an option to persons who, on January 1, 1972 became members of the Social Security Fund and who immediately before that date had a pensionable office in the Public Service to either remain under the national pension of Cap 30 or join the SSNIT Provident Fund as it was before conversion to a pension scheme under PNDC Law 247. Indeed, the 1975 Decree provided that should persons fail to exercise the option within the time allowed under the enactment, they would be deemed to have opted to remain under the Social Security Scheme. Once the option was exercised, it could not be revoked. While this provision meant that fewer people would remain on Cap 30 as the years go by, later events indicated that more people reverted to CAP 30 and this scheme runs alongside the SSNIT Scheme. This was made possible when the Pensions and Social Security(Amendment) Decree, 1978(SMCD 178) was passed. In addition, the Police Service and Prison Service have reverted to the Cap 30 Scheme with the promulgation of the Police Service(Pension) Law, 1985(PNDCL 126) and the Prison Service(Pensions) Law, 1987(PNDCL 168). Both enactments require that all amounts paid by members as contributions toward the SSNIT Social Security Fund should be refunded to the contributors concerned. Although CAP 30 Pension Scheme was established as non-contributory with defined benefits, subsequent amendments have made it contributory to civil servants and teachers who are still covered by it. Such members contribute 5% of their pre-tax salary, which is nevertheless not saved but recycled into the Consolidated Fund. However, it is still a non-contributory plan for the armed forces, the police, and the prisons services. These employees take home all of their earnings; no deductions for pension coverage. Other features of CAP 30 that are superior to those of SSNIT and are very contentious are: • 10 years for full retirement vs. 20 at SSNIT; • 70% of final salary compared to 50% of average of three highest years’ salary at SSNIT • CAP 30 pension payments are indexed annually to current salary scales. In the SSNIT scheme, there is the theoretical possibility that the pension right can be increased to 80% since any additional year served after 20 years earns a pension right of 1.5%. At this rate, a person must work for 40 years or 480 months to earn an 80% pension right. That is, such a beneficiary must have started contributing to the Fund at age 20, until he retires compulsorily at age 60. Appendix One shows the main features of CAP 30 and the SSNIT Scheme. Clearly, retirement benefits under CAP 30 are undoubtedly better than those under the SSNIT scheme, which is why those who can keep themselves in the plan do so, and others outside it are 14

fighting to get on it. The problem is not only that there is great dissatisfaction among those workers who do not enjoy the superior coverage of CAP 30; it is also that the largely unfunded nature of the plan is a drain on general revenue. This is more so when it is noted that even for those who make the 5% contribution, the funds are not available as long term saving for further productive purpose that would promote employment to increasing union membership. On a cautionary note, although the conditions under the CAP 30 Pension Scheme are more relaxed than those of SSNIT, it must be emphasized that a worker who is dismissed from the Public Service loses all pension rights. On the other hand, contributions to SSNIT are portable, unlike the restrictive character of CAP 30. The general scenario is that, social security as it exists today in Ghana, had been developed on a piece meal basis for different target groups. It lacks cohesion or overall design. Not surprisingly, those who are not under CAP 30 are fighting to get on it or trying to make SSNIT conditions identifiable with the largely advantageous benefits of CAP 30. As a matter of policy, it is advisable that all these different schemes should be rationalized and substituted by a comprehensive system of social security, which provides protection against social risks to all the people according to their status. There is no justification for two teachers to have gone through the immiserizing process of life with one having to enjoy better after service conditions. Injustice prevails when equals are treated unequally. What is prudent is that the disadvantaged person should be brought up but not for the advantaged person to be brought down: it is not acceptable in Pareto optimal relations or labour relations that conditions of service could be made worse. 6 SSNIT Ownership Status and Policy Formulation In an attempt to establish the ownership status of SSNIT, we look at the persons who provide the funds and who are eventually held liable or who would suffer losses in the event of collapse of the SSNIT Fund. Ultimately, the true owners of the SSNIT Fund are workers of all categories on whose behalf money is paid as under Section 20 of PNDC Law 247 2 . For these workers, Section 19 of PNDC Law 247 provides that,“ The Board shall cause to be maintained for each member an account to which shall be credited all contributions.” In fact, in antecedent legislation to PNDC Law 247, the Social Security Regulations, 1973(L.I. 818), Regulations 20 and 21 2 In reality, the funds belong to the employers who contribute 12½% and the workers who contribute 5% 15

explicitly provide that; The Trust shall maintain in respect of each member a record of contributions paid by or on behalf of the member plus interest and less any benefits and other authorized deductions” and that“The Trust at the end of each year, shall furnish to the employer of each member and the employer shall transmit to the member an annual statement showing the accumulation in the Fund at the credit of the member. Consequently, there is both financial and legal basis for asserting that the SSNIT Fund is a private fund with Government implicitly assuming custodial responsibility. All expenditures incurred in“the direction of the day-to-day business of the Trust and of its administration” are charged to members’ contributions as provided in Section 18 of PNDC Law 247 Emphatically, the workers whose money is held in the SSNIT Fund are responsible for the total costs of administering the scheme. Government resources or public funds are not used to administer the Fund; whatever money Government pays to the Fund is in its capacity as employer of specified employees just as any employer such as Ashanti Gold fields, Volta Aluminium Company, Volta River Authority, Lever Brothers or indeed any of the 20-odd companies listed on the Ghana Stock Exchange as well as all unlisted enterprises and the informal sector self-employed. The Law even indemnifies the members of the Board under Section 33 of Law 247, which states that; No suit or other legal proceedings shall be against any member of the Board, or any officer or employee of the Trust in respect of anything, which is done in good faith in pursuit of the objectives of this Law. That is, the losses made by SSNIT are borne by the contributors to the SSNIT Fund. Ownership of the SSNIT Fund, therefore, resides solely with these contributors who carry the risk and entrust their funds into the hands of the management of SSNIT. That the Minister of Finance should approve all expenses is only a means to ensure the checks and balances implicit in Government’s custodial role and to exact accountability from the Board so that contributors’ funds are protected! It is unfortunate that this custodial role has been exalted into ownership of the funds. However, discussions with workers, employers, and employees of SSNIT have tended to suggest that the ownership status of the SSNIT FUND is not clear. In one vein, SSNIT management 16

itself has professed that the SSNIT FUND is privately owned as it is funded by contributors who are private persons. And yet, the operational practices of SSNIT management have shown that the FUND belongs more to Government than to the contributors. A clear practical demonstration of management confusion is that SSNIT management has registered several of their vehicles as government property with GV REGISTRATION. Thus if government vehicles are commandeered for national assignments, SSNIT vehicles and other property would also be commandeered. Such was the case during national elections, when SSNIT vehicles and other property were run down at cost to contributors. In fact, that Government has taken SSNIT as part of Government property was shown in Government’s divestiture of the then Social Security Bank registered in 1977 under the Companies Code as a private banking institution wholly owned by the Social Security and National Insurance Trust. From hindsight, one of the over-riding factors that led workers to request for change from the Provident Fund Scheme to the PNDC Law 247 is political interference in the management of the defunct fund. Under the defunct scheme, investment policy of the fund was determined by government when it was required that funds should only be invested in government stocks at a 6 per cent rate of return. Consequently, the funds provided a cheap source of income for government in financing its budget deficit, to the disadvantage of the true owners of the Fund. This conception(or is it mis-conception?) has been carried over to recent management practices of the Fund that SSNIT FUND is part of the public purse and can be used as such. It would be necessary for Government to clarify its position on this so that workers confidence can be restored in SSNIT as a credible pension scheme. In terms of policy formulation, the SSNIT Board is to be in control of General Policy. Specifically, Section 8(1) of PNDCL 247 states that; “The Board shall, subject to the provisions of this Law have general control of the funds and investments of the Scheme and the management of the Trust on matters of policy”. The composition of the Board is as follows: -a chairman and three persons appointed by Government -a representative of the Ministry of Finance -a representative of the Bank of Ghana -a representative of the Ministry responsible for labour -a person appointed by the Director-General of SSNIT 17

-two representatives nominated by the TUC -a representative nominated by the Civil Servants Association -a representative nominated by GNAT -two representatives nominated by the Employers’ association of Ghana. In all, the Board of Directors, which is tripartite in nature and comprises representatives of workers, employers and Government, is made of 14 members. Although the composition of the Board has taken into account the tripartite character to cater for all vested interests, the weight is unduly asymmetric in favour of the Government. Representations from the government number up to EIGHT, out of a total membership of 14. That is, the BUREAUCRACY has at least straightforward 57% strength on the Board so that in case where policy issues are to be decided on the basis of simple majority, Government policy designs will always carry the day. In practice, therefore, SSNIT FUND is a privately funded scheme run predominantly by Government. This dominant position of Government on SSNIT Board is what has generated the erroneous impression that SSNIT is part of the Public Sector and must therefore be privatized. By Section 8(5) of Law 247, the Director General as the Chief Executive of the Trust is charged with the direction of the day-to-day business of the Trust and of its administration, subject to the general control of the Board on matters of policy. The Director General is assisted by three deputies who head three functional areas: Operations, Finance and Administration; these are further subdivided into departments that take into consideration similarities of functions. Essentially, however, final responsibility lies with the Director General as the Chief Executive of the Trust. Now, what is intriguing about the office of the Director-General is its implied political character. The Director-General is in a unique position as a Government representative on a Board that is dominated by Bureaucratic representation. Since he is appointed by the Government, he owes his position and loyalty to government interests. Consequently, he would represent Bureaucratic interests more than the interests of the Trust, especially in so far as the private owners of the FUND are under-represented on the Board and cannot make any appreciable impact on policy if they should decide to disagree with policies initiated by Bureaucratic representation. In fact, discussions with former and current members of the Board revealed that government had tended to use SSNIT Funds for all purposes that compromised the viability of the Scheme. While some members of the Board felt bad about such flagrant Government abuse and misuse of SSNIT 18

Funds, they felt impotent to raise any queries to call Government to order. At worst, members made excuses and absented themselves from Board meetings so as not to be party to such decisions. If Government is to appoint the Chief Executive of the Trust, which is a privately owned pension scheme without government subvention or financial support of any kind, then it stands to reason that some checks and balances must be instituted to prevent temptation by Government to use the FUND to promote bureaucratic interests. If the Bureaucracy is to appoint the Director General, then its representation on the Board should be reduced to a minority position. In this direction, it may be necessary to take a second look at the composition of the Board and to prune it down to ELEVEN. This will mean reducing representation from the Government. Ministry of Finance and Bank of Ghana’s role appears to be overlapping. In any case, Bank of Ghana is the overseer of the financial sector and has sufficient supervisory authority without active representation on the Board of the Trust. Bank of Ghana is not represented on most other financial institutions, some of which even have more risky and volatile liabilities and assets structure than SSNIT. Therefore, representation by the Bank of Ghana can be eliminated. On the other hand, the Bank is less political than the Ministry. So the Ministry may be the one to go. Similarly, the FOUR persons appointed by Government, including the Chairman, can be reduced to a Chairman and one other person. Thus, the total membership reduces to ELEVEN. It must be emphasized that this will entail efficiency gains in management and reduce temptation by Government to use the SSNIT’s resources as public funds. We do not only call for a reduction in Government representation on the SSNIT Board. We would also advocate for a more fundamental change in terms of calibre of Board membership. The Fund requires experts who can make informed decisions on investments before they are undertaken. This is particularly important for representations on the Board made by the various contributors to the SSNIT Fund. Their interest is paramount and it is they who must ensure that the funds are managed efficiently to give maximum yield. It is therefore recommended that representatives from GNAT, TUC and Civil Servants Association should be people who have the requisite expertise to ensure efficient management of SSNIT Funds. It should be possible for these interest groups to engage experts to represent them on the Board if qualified personnel could not be found amongst their members. This will mean that apart from coming from a particular constituency, additional eligibility criteria, which would be sensitive to SSNIT’s business, should be introduced into qualification for Board Membership. 19

7 Investment Portfolio of SSNIT Like all funded pension schemes elsewhere, large amounts of surplus funds are accumulated that need to be invested. With the SSNIT Scheme under Law 247, investment policy is expected to be much more professionally designed and implemented as the Scheme was conceived to be selfsustaining. Two things are required to achieve the professional ethos envisaged:(a) all restrictions on investments should be removed and,(b) responsibility for investments should be invested in the SSNIT Board, largely free of direct Government interference. Theoretically, investment returns are to be above a minimum acceptable level in aggregate over the long term. Cognisant of the long-term nature of the liabilities, investments in long-term projects are to be undertaken as long as short-term requirements are met. Generally, SSNIT’s investment policy is guided by seven general principles, which are: 1. safety of investments; 2. yield or rate of return; 3. liquidity; 4. maintenance of the Fund's monetary value; 5. diversification of portfolios; 6. spread of investments by duration; 7. harmonization with national objectives 3 . With these guiding principles, the investment portfolio of SSNIT Funds comprised investments in fixed and non-fixed income investments made up as follows: • Short-term government instruments • Government bonds • Corporate loans • Student Loans • Equity • Property Table 5 shows the relative share of SSNIT investment portfolio between 1997 and 2000. Generally, short-term investments make up the least share in the total, hovering around 5%. By the nature of SSNIT’s liability structure, this appears to be alright: SSNIT is matching maturity 3 See SSNIT,(May, 1992), op.cit, p.13. 20

profiles. Next is followed by government bonds, which although of longer-term character, are however on a declining trajectory. From 9.43% of total investments in 1997, this declined consistently to 7.16% in 2000. Together, short-term investments and bonds are the most risk free assets and tend to give SSNIT the highest rate of return. Investments in corporate loans, which were as high as about 24% of SSNIT’s total investment in 1997, fell to about 11% in 2000. Of significance is SSNIT’s equity holdings, both in listed and unlisted companies on the Ghana Stock Exchange. The equity portfolio has increased significantly to more than half a trillion Cedis, and from about 13% of total investments in 1997, SSNIT’s equity shares now stand almost 30% of the portfolio. In fact, SSNIT is the single largest holder of shares on the Ghana Stock Exchange and there is no stock on the Ghana Stock Exchange that SSNIT is not a major shareholder. SSNIT is the largest institutional investor in Ghana and is as such badly exposed in the capital market. Table 5 SSNIT Investment Portfolio: Relative Structure(Percentage of Total) Investment Type December 1997 December 1998 December 1999 December 2000 Short-Term 5.29 5.39 8.23 5.8 Bonds 9.43 8.98 7.32 7.16 Corporate Loans* 23.91 13.89 14.5 10.95 Student Loans 8.42 11.32 12.97 15.24 Equity 12.91 22.55 19.6 27.49 Property** 40 37.42 37.38 33.35 Total Investment 100 100 100 100 Source: Calculated from SSNIT Annual Reports, Various Issues *Calculated from amounts that are net of provisions. Figures may not add up to 100 because of rounding. ** Property includes completed projects and work in progress Poor returns have resulted from bad experience with corporate loans, unsuccessful investments in unlisted securities and property investments with inadequate cash flow. Property investment has declined as a percentage of the total over the years but it is still the single largest investment category at about one-third of the total. This is in spite of the fact that the property market is very undeveloped in Ghana. It is therefore surprising that over a third of SSNIT’s investment portfolio is in real estate and landed property. Indeed, in 1997, SSNIT’s property investment was as high as 40% of its total investment. This has declined marginally in 1998 and 1999 until it stood at 21

33.35% in 2000. In the same vein, SSNIT appears to be winding down its corporate loan portfolio, which were as high as about 24% of total investments in 1997 but declined to 11% by 2000. The basic problem with this portfolio is that, its quality is doubtful. Of further concern is the contentious student loans category of SSNIT’s investments. This grows each year both absolutely and as a percentage of the total. From about 8% in 1997, student loans currently represent about 15% of SSNIT’s total portfolio. Although the Students Loans scheme is a laudable idea as a social protection measure for students who would otherwise have been denied access to tertiary education for want of funds, the scheme as it operates raises some concerns. The first major concern is the use of SSNIT resources for that purpose against the background of increasing number of students in recent times. Indications are that the student numbers would increase as not only the polytechnics but also all private sector tertiary educational institutions admits more students. Invariably, these students are under the impression that the SSNIT is part of Government Bureaucracy and the funds are public funds that could be used for meritorious public services. Consequently, students demand, rather erroneously as a matter of right, increasing loan amounts from SSNIT. Any delays in the payments of the loans are met with student protests largely directed at SSNIT. Unfortunately however, the interest payments on the loans are not timely. Indications are that not only are the loans not market-determined but interest payments due are also not paid by Government. Accordingly, SSNIT treats such interest payments as accruals. Worst still is the loan repayment rates by the students themselves. SSNIT is not making any conscious attempts to compel repayment. This is not surprising because the loans are secured to the benefits of guarantors of the loans. When students fail to repay the loan, which is commonplace, pensioners lose their entitlements. Thus for those contributors who guarantee student loans, the SSNIT scheme ceases to be the social protection scheme that was envisioned in its conception to be as specified in Section 3(b) of PNDC Law 247 namely“to provide social protection for the working population for various contingencies such as old age, invalidity, and such other contingencies as may be specified by law”. The financing of student education is necessary not only for ultimate benefits of education but because better education itself increases the level of productivity, employment and consequently 22

higher levels of social protection. However, the financing of education is the fundamental responsibility of Government or parents or the consumers of the fruits of education. Private resources such as SSNIT funds should not be used for such purposes as if they were public funds. Government must discipline itself to make the necessary distinction between SSNIT contributions as private funds and taxes or government revenue. The contributors’ disenchantment with SSNIT can mainly be repaired when people know that government would keep its hands off SSNIT funds and would not use it indiscriminately. Especially, if SSNIT is going to be the national vehicle of social protection that it is envisaged to be and to rope in the self-employed and other informal sector individuals, SSNIT must not be seen to be playing politics with members contributions. If Government requires additional revenue for other purposes, some of which may have social insurance connotations, then transparency demands that this must be raised through taxation of if it must borrow from SSNIT, it must be on market terms. It is suggested that the SSNIT legislation be reviewed and purged of all inconsistencies such as the law tagging the Students Loan Scheme to SSNIT. SSNIT should be able to devise a product on its own initiative to meet students’ needs if the SSNIT Board does a feasibility studies and is accordingly interested in such an investment. Forcing SSNIT to finance students’ loans without proper appraisal of such investments tends to compromise the viability of SSNIT and jeopardises the future of the scheme. 8 Structure Of Investment Incomes SSNIT has a large pool of funds that it can manage efficiently to provide an effective first tier social protection for a greater number of Ghanaians. The investment portfolio of SSNIT is approaching two trillion Cedis but has ongoing problems with return, liquidity and asset quality, particularly with non-listed equity, property and corporate loans. As we noted earlier, investments in property and equity take 33.35% and 27.49% of SNNIT’s total investment. However, rent incomes to SSNIT from property investments averaged about 2.27% between 1997 and 1999. It increased marginally to 3.63% of total investment income. Similarly, dividends accounted for 4.92% of total investment incomes in 1997. It then increased to 6.11% in 1998 and then to 9.54% in 1999. It was only in 2000 that dividend incomes as a percentage of total investment income was greater than 10% although its investment share was about 28%. Table 6 shows the investment incomes from the various assets held by SSNIT between 1997 and 23

2000. At a strategy plan launching in June2002, SSNIT reported that in all, it recorded a negative 7.42% average return on its total investments over the past five years(emphasis ours). As the report itself puts it, After allowing for inflation the Trust’s investment portfolio yielded-25.9%, 11.1% and-3.1% in 1996, 19997, and 1998 respectively. It recovered in 1999 posting a real return of 4.4% and declined to-1.4% at the close of 2000 in line with the unfavourable macroeconomic environment that prevailed during the year (SSNIT, Medium Term Strategic Plan 2002- 20006, page 5). To say the least, this is unacceptable and must not be countenanced under any circumstances. Accordingly, it is recommended that SSNIT should restructure its assets portfolio so as to maximise returns and protect the quality of its investments to ensure sustainability of the scheme as a social insurance fund. Table 6 Investment Income: Relative Shares in Percentages Investment 1997 1998 1999 2000 Govt Registered Bonds 5.62 0.25 0.12 4.47 Term Deposits/Treasury Bills 11.54 15.06 17.28 15.66 Student Loans 12.16 15.29 15.69 28.53 Corporate Loans 38.32 60.78 51.28 25.74 Rent 2.20 2.52 2.10 3.63 Dividends 4.92 6.11 9.54 11.51 Profit On Disposal of shares 24.84 0.00 1.36 0.69 Miscellaneous 0.39 0.00 2.63 9.77 Total 100.00 100.00 100.00 100.00 SSNIT, Annual Report, Various Issues. Calculated from data reported in the annual accounts. 9 General and Administrative Costs of SSNIT One other major concern about the management of the SSNIT fund is the relatively high administrative costs incurred. Indeed, not only are the costs of administering the fund high but they are on the ascendancy, reaching as high as 30.1% in 1998. The experience of SSNIT, especially since 1997, is really worrisome. Between 1997 and 2000, SSNIT was using, on the average, about 26% of total contributions to administer the Fund. More disturbing is the fact that such general and administrative costs exceeded the amounts paid out in benefits. For example, in 2000, the administrative cost was ¢123172m compared to total benefits payment of ¢95462m. These administrative expenses of SSNIT are undoubtedly a major area of concern, especially if 24

SSNIT is to be made efficient so that resources can be stretched for improved benefits to members. Table 7 shows the trends in general and administrative expenses of SSNIT between 1991 and 2000 Table7 Year Trends in Contributions and Administrative Expenses: 1991- 2000 Administrative Contributions Benefit Paid Administrative Administrative Expenses ¢m Received ¢m Pensioners ¢m Expenses As% Expenses As% 1991 4998.9 20657 786.6 24.2 1992 6992.4 32361 1448.8 21.6 1993 9933.1 86100 3321.3 11.5 1994 17758.9 83352 8123.9 21.3 1995 20299.5 115685 13544.8 17.5 1996 30500 170000 20800 17.9 1997 49479 211371 33384 23.4 1998 70846 235146 45369 30.1 1999 86788 372782 62347 23.3 2000 123172 458838 95462 26.8 Source: SSNIT, Annual Reports, Various Issues Percentages calculated using SSNIT data on contributions and expenses. general 635.5 482.6 299.1 218.6 149.9 146.6 148.2 156.2 139.2 129 and administrative Although the experiences of other developing countries suggest that many funds are in trouble because of high costs of administering the fund, Ghana’s experience as described in Table 7 is legend compared to what Mesa-Lago(1989, p.254) 4 reports on as follows: “The percentage of administrative expenditure over total expenditure in the system is very high for Latin America, fluctuating from 7 percent in Costa Rica, Chile and Uruguay to 10 percent in Peru and 18 per cent in Mexico, percentages far above those in the developed countries. The majority of administrative expenditures relate to the remuneration of personnel, which is excessive in practically all of the countries” SSNIT data suggest that serious administrative inefficiencies are driving up costs and absorbing an inordinate amount of incoming revenue. Launching its Medium-Term Strategic Plan for 2002-2006, SSNIT Management reported that the total staff strength of the Trust as at 31 st December 2001 was 2650. This was made up of 2151 permanent staff and 499 contract staff. Out of this, management and senior staff made up 49.7%, while junior staff was 50.3%. Thus it was 4 C. Mesa- Lago, ed.(1985), The Crisis of Social Security and Health Care: Latin American Experiences and Lessons, University of Pittsburgh Press. 25

concluded in the strategic plan that 5 : The structure shows an even split between junior and senior/managerial grades. This amply demonstrates that the structure is top heavy and may be an indication of inefficient use of personnel(emphasis ours). The above self-indictment could not be better put about inefficient resource usage with extremely high administrative costs. Although this phenomenon is not peculiar to Ghana, the increasing trend in the costs shows that drastic measures must be taken to curtail costs to make some efficiency gains. The containment of these administrative expenses can significantly influence the amount of benefits that SSNIT can ultimately pay out. The choice is not to “privatize” SSNIT or dismember it into smaller pension schemes. There are sufficient bases of both economies of scale and economies of scope in SSNIT’s operation. Once SSNIT can demonstrate that the cost of administering the Fund declines with increasing operations, then it would be much more cost effective to allow SSNIT to operate as the first tier of social security for all Ghanaians. Much as the operations of social insurance are fairly routine as appears to be the case with SSNIT, the case for a unified scheme is stronger than where operations involve great difficulties of definitions of rights. Indeed, it must be emphasized that privatization and competition in practice do not guarantee that there will be no dishonesty in economic set ups. History has taught us that both private and public pensions can be problematic and mismanaged for private gain. Robert Maxwell 6 who siphoned at least£400 million from various employee funds in the U.K is a classic example of private dishonesty in pension management. Similarly, various scams have affected private pensions in the United States to the extent that Government was forced to use between$200-500 billion of taxpayers money to bail out these private companies and to ensure that depositors receive their pensions 7 . 10 The Way Forward To be able to make recommendations for overhauling SSNIT and also of improving the pensions system in general we look at what has become known as the Swiss Chilanpore, the touted best practices in the pensions reform literature. We do this against the background of problems associated with SSNIT management in general. For emphasis, we recapitulate the core problems 5 SSNIT, May 2002, Medium-Term Strategic Plan(2002-2006), Page 16. 6 “An Honour System Without Honour”, Economist(December 14, 1991) and also“The Maxwell Mess- Worsening”, Economist(February 1, 1992) 7 For further details on private inefficiency of pension funds, see Robert Kuttner,“Privatization Is Not A Cure-All”, Wall Street Journal, April 30, 1992. 26

that face SSNIT as follows: • Poor Investments • Negative Rates of returns on overall investment portfolio • Excessively high administrative costs • Low coverage with marginalization of informal sector • Inadequate level and amounts of benefits actually paid • Excessive government control and interference in the activities of SSNIT • Inadequate products The experiences of other countries suggest that Ghana is not the only country with such problems in the running of social insurance schemes. However, when these problems are identified, it is important to devise an effective set of mechanisms for reviewing the general operations of the social insurance scheme and especially for assessing the financial integrity of the budgetary operations. This is done in different ways in different countries through a variety of boards, committees, advisory groups, actuarial reviews and parliamentary oversight committees. The answer to the poor performance of SSNIT does not lie in the example of anyone country’s experience. It is our firm conviction that we can draw from the experiences of what has become known as the Swiss Chilanpore standing for the social insurance programmes for Switzerland, Chile and Singapore. We review the essential features of these schemes for any useful lessons for Ghana. This is done bearing in mind that no one approach is perfect as each country has its own unique culture, history and level of development to cope with. We need to fashion out the appropriate mix of social insurance suitable for Ghana. 10.1 Swiss First Pillar The essential features of the Swiss system are based on the three-pillar system. The first pillar is a social insurance scheme that pays defined basic benefits or pay as you go system and financed with a total contribution rate of 8.4% equally divided between employers and employees. Government makes an additional contribution to the first pillar from general revenue to cover 20% of pension payments so as to facilitate a redistributive scheme in favour of low-income workers. While the first pillar has useful lessons for developing countries, it must be emphasized that the Swiss economy is a first world economy that has the capacity to generate the required revenue to support the pension system. We must be mindful of the fact that in Ghana, at least for 27

the moment, well over 40% of the national budget is funded from donor sources. Indeed, the total revenue is just enough to finance recurrent expenditures. Consequently, it is not pragmatic to recommend unfunded schemes for implementation in Ghana at the moment. The Swiss second pillar consists of compulsory based plans that pay complementary pensions aimed at achieving a satisfactory replacement rate. The third pillar comprises voluntary savings including fiscally supported pension plans for self-employed people and other workers not covered by company schemes. These two pillars are replicated by the Chilean and Singaporean examples vividly. Consequently, we shall discuss these as bases for finding useful lessons for Ghana. 10.2 The Singaporean System The pension system of Singapore is organized on national provident fund principles. The scheme is designated as the Central Provident fund(CPF). All workers, except self-employed people, are required to participate in this Central Provident fund(CPF). The CPF is a public board that administers the system, collects contributions, keeps records, pays out benefits, and invests the accumulated funds. The investment aspect of the fund is a simple straightforward function for the CPF since, practically, all the funds are invested in government instruments. The investment decisions that matter are taken by two other very important government institutions, the monetary authority of Singapore and the Government of Singapore Investment Corporation. The system was a pure mandatory retirement savings scheme, forcing workers to save for their old age and allowing lump sum withdrawals on reaching age 55. Contributions, which were 10% and divided equally between employers and employees during the inception of the scheme in 1955, were raised to 13% in 1968, when a decision was also made to allow interim, but controlled, withdrawals for the purchase of houses. Since then, there have been several increases in contribution rates, which reached a staggering total 50% in 1984. Subsequently, however, because of the negative impact on employment creation during the recession of 1985/6, the total contribution rate was lowered to 35% by setting the employer’s rate to 10%. More recently the contribution rate has stabilized at the still very high level of 40%, with a long term aim of dividing this equally between employers and employees. An innovation of recent years is the institution of lower contribution rates for people aged over 55, while, from the very beginning, workers earning less than a specified minimum were exempt from making contributions. Over the years, additional investment opportunities for investment in approved securities and for spending for education were allowed, while health insurance was also included among the 28

benefits of the system. Also since1987, workers are required to keep a minimum sum in their account after reaching 55. This is fixed by CPF and is adequate to purchase on retirement at age 60 a minimum life annuity equal to about 25% of average earnings. The CPF is a defined contribution system with no intentional redistribution. Its primary objective is a forced saving for old age. These days it is not a purely retirement savings scheme since it allows use of funds for several other purposes. Thus, its secondary objective is to encourage spending on merit goods(health, housing, education 8 ). Although redistribution is not among its objectives, it is often argued that the CPF creates perverse redistribution because of the low rate of interest credited on account balances. This is particularly so because only high-income workers can avail themselves of the opportunities to invest in other approved but high yielding assets. Although the CPF is a mono-pillar system, the Government of Singapore operates a public assistance pension scheme that offers to destitute old people a small pension that is half the size of the minimum pension imposed under the CPF and amounts to about 12% of average earnings. One of the strengths of the CPF is its high efficiency and very low operating costs. In 1990, total operating costs, including depreciation provisions, amounted to 0.53% of annual contributions, 0.21% of wages and 0.10% of accumulated assets(CPF1990). These ratios are very low by international standards and compare very favourably with those achieved by large employerbased company scheme in Britain and the US. For instance, Postel, the company managing the pension funds for the employees of British Telecom and British Post Office, has total operating and investment management costs of 0.1% or 10 basis points of total assets. This is divided between 6 basis points for operating costs and 4 basis points for investment management cost. Apart from the lack of redistributive objectives, two fundamental weaknesses of the CPF are its very high total contribution rate and the low rate of interest credited on account balances. 10.3 The Chilean System The new Chilean system was introduced in 1981. It is a mandatory fully funded retirement savings scheme that was created to replace an insolvent social pension system that operated on a “pay-as-you-go” basis. The scheme requires all employees to contribute 10% of the first $22,000 of earned income until their normal retirement age, which is 65 for men and 60 for women. No contribution is imposed on employers, although they are required to withhold employee contributions and transfer them to the account holding companies. On retirement, 8 Our emphasis 29

workers must either purchase a life annuity from an insurance company or arrange a schedule of programmed withdrawals from their account. Lump sum payments are allowed only if account balances exceed the sum required to purchase an annuity equal to 70% of final pay. Like the Singaporean CPF, the Chilean system is a defined contribution system based on individual capitalization accounts, where pension benefits depend on the contributions made over a person’s working career and the investment income earned on accumulated balances. Workers are required to purchase term life and disability insurance and pay an additional commission to cover the premiums for these insurance policies as well as the operating costs of the system. It is managed by the private sector. More important, Chile created a retirement system that, by giving workers clearly defined property rights in their pension contributions, a) offers proper work and investment incentives; b) acts as an engine of, not an impediment to, economic growth; and c) enhances personal freedom and dignity The Chilean system has some unique characteristics. It is government mandated and regulated, but completely privately managed by a number of authorized private sector pension management companies, known as Adminstradoras de fondos de Pensiones or Pension Fund Administration Companies Superintendency or AFPs for short. The AFPs are permitted only to administer pension funds. As at December 2001, there were eight such AFPS. Furthermore, the number of affiliates including pensioners and individuals who have contributed to the system at least once stood at 6.4 million: active contributors stood at 3.4 million or 14.8% of Chile’s total population of 23 million. To ensure simplicity and transparency, regulations impose a strict limit of“one account per worker” and“one pension fund per AFP” Affiliates receive regular statements with information about the credited contributions and the investment income of their fund. To ensure the solvency of the system, the pension fund is legally separated from the management companies. This means that in the extreme case, an administrator could go bankrupt but the individual funds of each affiliate would remain unaffected, since they belong to the worker and not the AFPs. The affiliate would only have to transfer his or her funds to another fund manager or administrator. Strict rules are imposed on AFPs, regarding their capital reserves, the investment of pension fund assets, and their performance relative to the average for the AFP industry as a whole. Investment emphasizes safety and profitability. A certain amount of diversification is required and for this purpose maximum limits are imposed on portfolio shares in different classes of instruments as well as in instruments of different issuers. No attempt is 30

made to direct the investment of funds in high priority economic or social projects. The system is subject to strict, even draconian, regulation and to very close and effective supervision. A very important feature of the system is the individual choice granted to affiliates to transfer their accounts between AFPs. Individual choice is expected to maintain pressure on AFPs to compete and operate efficiently, though experience has shown that unlimited choice to transfer accounts may result in very high operating costs, mostly because of publicity and marketing expenses and the actual account switching costs. Individual choice in the purchase of annuities has also given rise to high publicity and selling costs. Workers have the freedom to switch between AFPs, taking with them the entire amount of their accumulated funds. This portability freedom was abused in the mid-1990s when increased competition led to unbridled campaigns by AFPs to woo entrants. Some AFPs offered bicycles as rewards for affiliates moving their funds to them. Some others offered cash to those who switched from one company to another. Consequently, the system displayed a significant affiliate “rotation rate”, which had the effect of increasing administrative costs of the system, and by default, the commissions of the AFPs. At its height, during 1997, the AFPs registered a total of 1.6 million affiliate transfers and employed 18,000 sales people. In response to this situation, in the late 1990s, the authorities created a series of administrative procedures that must be fulfilled when an affiliate wishes to switch between AFPs. Furthermore, the AFPs rationalized their sales forces, greatly reducing the number of sales personnel. These measures led to sharp decline in the rotation of affiliates and a sizeable drop in commission rates. During 2001, only 200,000 affiliate transfers occurred among the AFPs and the system employed about 2000 sales personnel. As a defined contribution system, the Chilean system is ostensibly a mono-pillar system with no intentional redistribution. Some redistribution may take place, however, through the government guarantee that workers with at least 20 years of contributions will always receive the minimum pension. It is not clear how many workers will benefit from this guarantee. Clearly, this depends on the future relationship between wage growth and real returns. In addition to the guarantee regarding minimum pensions, the authorities impose on AFPs, and guarantee in case of AFP failure, a minimum investment return relative to the average for all pension funds. The government also guarantees, subject to specified limits, the value of life annuities with insurance companies. In any case, so far, the funds managed by AFPs have 31

displayed attractive rates of return, with average annual real rates of return of 10.7%. The profitability of the funds has, however, been reduced, principally, due to the relatively poor showing of the local stock exchange. The total amount of funds managed has increased significantly since the inception of the new system from$236 million in 1981 to$35.4 billion by December 2001. Some features of the Chilean scheme give rise o regressive redistribution in favour of highincome workers. This arises from two main sources. First, because the structure of commission charges includes both a flat fee and a per valorem fee, low-income workers are effectively credited with a much lower rate of return than high-income workers. Data for the first ten years of the AFP system show that low income workers obtained a real rate of return of 7.5% against 10.5% for high income workers and 13% for the totality of the pension fund(Vittas and Iglesias, 1992). The difference is caused by the imposition of the flat fee, which in the initial years of the scheme was quite high, but which has declined significantly in real terms. In fact several AFPs have now abolished their flat fees altogether. Unintentional redistribution may also arise from variations in returns among AFPs, though these are limited by the required minimum relative investment returns on pension funds. The second perverse redistribution arises from the forced use of annuities. In theory, life annuities should take account of the shorter life expectancy of poorer people or people from particularly arduous occupations. In practice, however, it seems that low-income workers, not only do not benefit from lower annuity prices, but may also pay much higher commission charges for their life annuities than high-income workers. The extent of perverse redistribution through the use of life annuities is not known, but appears to give rise to concern among Chileanpolicy makers and analysts. However, commentators have noted that, through their pension accounts, Chilean workers have become owners of the means of production in Chile and, consequently, have grown much more attached to the free market and to a free society. This has had the effect of reducing class conflicts, which in turn has promoted political stability and helped to depoliticise the Chilean economy. The only real drawback is excessive government regulation(Testimony of Mr. L. Jacobo Rodriguez of the Cato Institute to the Ways and Means Committee of the US House of Representatives, July 31, 2001). 11 Similarities And Differences In The Swiss Chilanpore Schemes The basic lesson derived from the pension systems of the three countries discussed is the hardheaded softness of the Swiss scheme, the expensive yields of the Chilean scheme and the 32

ruthless efficiency of Singapore. These three have a number of features in common, but they also exhibit some important differences. Notable ones are: • The first similarity is that they all have compulsory systems that cover nearly every worker, except self-employed people. • The second similarity is that they rely to a substantial degree on funded schemes. The financial resources accumulated in pension funds are large in relation to national income in all three countries. • The third similarity is that they represent relatively successful economies with high levels of national and household saving. This is particularly so in the case of Singapore and Switzerland. Chile has suffered from the high inflation that has long characterized the economies of most Latin American countries. But allowing for the negative effect of high and volatile inflation on national saving, especially on financial savings, the financial performance of the Chilean pension funds has been quite remarkable. Of course, several other countries exhibit characteristics similar to those of these three countries. For instance, Britain and the US have large funded pension schemes, though unlike Switzerland, employers in these countries are not compelled to offer pension schemes to all their employees. In Britain and the US less than half of private sector employees are covered, by company pension schemes, against 100% in Switzerland. Korea, India, China, Italy, Greece. Other countries have high rates of household saving, but they do not have compulsory funded pension schemes. Finally, France and, perhaps to a lesser degree, Germany, impose compulsory participation in pension schemes on their residents, but these schemes are not based on funded pension plans. On the other hand, the three countries also exhibit some important differences as follows. • The Swiss system, like those of most OECD countries, is extremely complex and opaque. The complexity of the system makes it difficult to measure its cost or to assess the investment performance of the funded components of the system • • The Singaporean system is quite simple and operationally very efficient. However, it suffers from lack of transparency and produces relatively low returns and benefits to its affiliates • The Chilean system is very simple and highly transparent and is also supported by very effective regulation and supervision. It has produced very high real returns, but suffers from very high operating costs. These afflict not only the pension system itself but also the private annuity market on which it is partly based. • Unlike Switzerland, neither Chile nor Singapore incorporates in their pension systems 33

intentional redistribution in favour of low-income workers. On the contrary, both may inadvertently cause unintentional redistribution that may be perverse by penalizing lowincome workers. Nevertheless, both countries offer some forms of minimum pensions. 12 Recommendations For Improved Retirement Schemes From our review of the SSNIT scheme and the lessons drawn from best practices across the world, we make the follow recommendations. • SSNIT should have the force of government legislation behind it as a First Tier National Pension Scheme. However, the heavy presence of government must be removed from its operations. Indeed, many of the shortcomings and weaknesses of SSNIT derive from the structure of the Law establishing it, PNDC Law 247, and how the law is implemented. In other words, Government’s role must be largely regulatory. • The SSNIT Board should be re-engineered: how it is constituted, its responsibilities and associated corporate governance issues must be recast in the light of best practices and SSNIT’s own history. • The Board should be reduced in number from fourteen to eleven. SSNIT must be purged of excessive government presence and interference by reducing Government representation on the Board to about four • SSNIT lacks an effective formal oversight body. True, the Board and external auditors ostensibly play this role. However, the indications are that the oversight responsibility is weak, probably because of politics in the appointment of the Board. It is therefore recommended that SSNIT must be made accountable to Parliament through annual submission of report and/or appearance in the House to answer questions. The Report which should be accompanied with audited statements should be subject to the Scrutiny of the Public Accounts Committee of Parliament. This will bring transparency and accountability to its operations and may protect SSNIT from excessive government intervention. • As it is SSNIT is virtually a private fund. However, since government has an intrinsic responsibility for social protection, it must bear residual responsibility for some minimum pension; in other words, it should provide some form of guarantee for the Scheme. The current objectives of SSNIT are not explicit; they are sometimes contradictory and indistinct. Not surprisingly, they have engendered confusion, which has impacted negatively on accountability and responsibility among the Board, Staff and Government. It is recommended that this needs to be reviewed. Concise and unambiguous objectives should be clearly specified relating to what SSNIT does and does not do. SSNIT must be purged of inconsistent laws, typically the Students Loans Law, that compromise its true objective. • The Administrative costs of SSNIT are too high compared with best practices. The organizational structure of SSNIT would have to be clearly defined, especially as regards 34

senior staff and junior staff ratios, which is one major source of inefficiency. This would entail staff rationalization as a cost reduction strategy to infuse efficiency into SSNIT operation. • The core business of SSNIT entails collection of contributions, investing such collected funds and making pension payments. This can be done using IT solutions as ways of capturing economies of scale and scope so that these services are delivered accurately and efficiently. • Stakeholders need not fill their representation on the Board from among their membership. In this connection, qualifications for membership of the Board should be set taking into account the core business of SSNIT and its allied responsibilities to guide the stakeholders in choosing their representatives. • Since the Board is made up of representatives, they must have the legal authority to report back to their constituency. This will suggest that the Oath of Secrecy required presently to be taken by Board members should be abolished because it may have a chilling effect on members. If need be, confidentiality requirements which protect SSNIT’s trade secrets and competitiveness may be introduced. • The Chilean example is overstretched and cannot replace SSNIT. The Chilean example is recommended as second tier private retirement scheme with the following features. • CAP 30 as it exists today should be preserved only for the following security service personnel who are already under the scheme: Armed Forces, Police Service, Prisons Service, and Fire Service. This conforms to best practices worldwide where these security services are under unfunded schemes. This does not however preclude them from taking part in other schemes in so far as they have not exceeded the allowable tax incentives. They should be free to join the SSNIT scheme. • As best practices have shown, a second tier private retirement scheme with wider objectives tends to enhance social protection, increase national savings, promote investment and facilitate economic growth for poverty reduction and wealth generation. Consequently, it is recommended that Government facilitates the establishment of second tier long term savings plan, probably designated as Private Retirement Plan • The proposed Private Retirement Plans must be employer-sponsored and must be tax deductible. • The Funds of the Plans must be managed by approved private fund managers regulated by the Securities and Exchange Commission(SEC) under the Securities Industry Law, 1993(P.N.D.C.L. 333), as amended by Act 590. Fund managers must be required to offer an Individual Retirement Plan as a way of reaching out to persons in the informal sector. It is recommended that fund managers should offer Home Ownership Plans and Education Plans. • A third tier private scheme is also recommended to meet the needs of merit goods such as housing and education, along the lines of the Singaporean example. Alternatively, the 35

Second tier Private Retirement Plan could be made not entirely forced saving for old age but to meet other social things that workers desire to acquire before retirement. • It is recommended that any restructuring of the SSNIT Scheme should enable it to cover all the 9 products listed in ILO Convention 102 as noted. The recommendations including those in respect of SSNIT, all require tax incentives. Obviously, this has revenue loss implications for Government. However, considering the enormous benefits to be derived from these schemes and as best practices show, the tax revenue that will be lost will be quite insignificant. It is recommended that there should be a contribution ceiling of 35% of a contributor’s annual income as tax deductible for all long term savings plans including SSNIT. If this recommendation is adopted, the following tax treatments are important as tax incentives • Contributions made by an employer to a retirement plan on behalf of an employee, should be tax-deductible; • Contributions made by an employer to a retirement plan on behalf of an employee, should not be treated as part of the employee’s assessable income for any tax year; • Contributions made by an employee or a self-employed person to a retirement plan should be tax-deductible to the extent of such contributions; • The income(including investment income) of a retirement plan should be tax-exempt; • Withdrawals from a retirement plan on or after the statutory retirement age should be taxexempt. Lump sum withdrawals after 10 years of contributions and before retirement should be subject to tax liability and a penalty. No withdrawals should be permitted before 10 years of contributions. • Withdrawals from a Home-Ownership Plan or an Education Plan on or after 5 years of contributions should be tax-exempt. Earlier withdrawals should be subject to tax liability and a penalty. • Early withdrawals from a retirement plan by a member on account of proven total physical or mental disability should enjoy some level of tax-reliefs These recommendations would require amendments to the current tax regime under the Internal Revenue Act 2000(Act 592) as amended by the Internal Revenue(Amendment) Act 2002(Act 622), and a complete overhaul of current applicable legislation on SSNIT. 36

APPENDIX 1 The Essential Thrusts of SSNIT Pension and CAP 30 Essential Thrust Legislation Administration Coverage Contribution Normal Retirement Age Early retirement Age Minimum Eligibility Conditions Social Security& National Insurance Trust(SSNIT) Social security Law 1991(PNDCL 247) with precursors as the Provident fund under the Social Security Act, 1965; and 1972 Social Security Decree,(NRCD 127): Management Board comprising representatives of Government Bureaucracy, Employers and Unions. Private and self- employed Civil Service and public sector employees from January 1, 1972. Employee Employer Total Age 60 Age 55 5% 12.5% 17.5% 240 months of portable contributions National Pension Scheme Or CAP 30 1946 Pensions Ordinance; Chapter 30 of Ordinance No.42, 1950; Teachers Pension Ordinance, 1955; Pension and Social Security(Amendment) Decree, 1975 SMCD 8. Management responsibility under the controller and Accountant General’s Dept. of the Ministry of Finance. Civil Service, Public Sector and Civilian Employees of the Armed Forces who were pensionable and non-pensionable officers prior to January 1, 1972 and who opted to stay on cap 30; Armed Forces, Police Service, Prisons Service and some Staff of the Judiciary Service. Non – contributory for Armed Forces, Police Service, Prisons Service and Judiciary Service. Age 60, Compulsory Age 45 for Voluntary, but earlier retirement permissible under “special circumstances”. 10 years continuous contributions, except under“ special circumstances” Final Pension Salary(FPS) Average annual base salary for the best 3 years in employees working life Final pensionable Emoluments: final gross annual salary earned by officer at date of retirement used in computing pensions. For Reduced Pensions, i.e. on drawing a gratuity, FPS equals 50% of minimum salary in first point in the officer’s salary range (effective June 1995). 37

The Essential Thrusts of SSNIT Pension and CAP 30(continued) Criteria Social Security& National Insurance Trust(SSNIT) Normal Old Age Actuarial Rate: Pension Earned Pensionable Right is defined as 50% of FPS after 240 months contribution, plus 0.125% per additional month. If the employee contributed less than 240 months, a lump sum benefit equal to total contributions made to date adjusted to prevailing Government Treasury Bill rate. Optional commutation or gratuity is at 25% of the employee’s 12 years pension as a lump sum. Maximum Actuarial Rate is 80%. -Minimum Pension varies according to prevailing minimum wage and inflation. Early Retirement Age Survivors’ Benefits -Death Before Retirement Reduced pension at age 55 is 60%, whilst age 56 is 67.5%, age 57 is 75%, age 58 is 82.5% and age 59 is 90% If member contributed for at least 240 months then a lump sum equal to present value of employee’s pensions payable to age 72 is paid to survivor. If contributions are less than 240 then lump sum equal to 50% of the employee’s pensions payable to age 72. Applicable interest rate for computation is a 10% discount rate per annum National Pension Scheme Or CAP 30 Established officers with Pension Rights: -Full pension; 1/480 th FPS times length of service in months: Reduced Pension; ¾ of Full Pension Gratuity: 1/4 of Full Pension times 20 as lump sum payment. Non-established Officers, i.e. General Orders 400 to 410: Full Pension; 1/800 th FPS times length of service in months: Reduced pension; ¾ of Full Pension: Pension; Fixed Annual Allowance(Adjusted Annually): Gratuity; Ex-Gratia Awards. Maximum Actuarial Rate is 100% FPS i.e. 40 years’ continuous service. Minimum Pension is 50% of salary range at rate of retirement There is no actuarial reduction. Death gratuity; Lump sum payment of 20 times 1/3 of officers full pension. Widow’s Pension; pension of 25% of FPS. Each Child(maximum 6) with mother; Pension of ¼ widow’s pension to age 21. Each child(maximum 6) without mother; Pension of ½ of widow’s pension per child to age 21. 38

The Essential Thrusts of SSNIT Pension and CAP 30(continued) Survivor’s Benefit- Lump sum equal to the present value Death After of employee’s unexpired pension Retirement payable until age 72. No benefits if pensioner dies after age 72. Disabilities Pension Entitled to Earned pension if minimum contribution is 240 months. Entitled to 50% of Earned pension if contribution is less than 240 months. Adjustment to Indexed annually at January 1. pension in payment. Marriage Benefits None Lump sum equal to balance of unpaid pensions, until 80 years. Balance is paid to employee’s estate. Treated as early retirement pension without limit of 10 years service under“special circumstance”. Indexed annually to 100% of adjustment to Civil Service salary scales. 1/8 of monthly FPS per month of continuous service, payable to wives in pensionable posts who resign or retire, having served a minimum of 60 months or 5 years. 39

References Arenas de Mesa, Alberto and Mario Marcel, 1999,“Fiscal Effects Of Social Security Reform In Chile: The Case Of The Minimum Pension” Paper Presented At The APEC Second Regional Forum On Pension Funds Reforms, Vinar del Mar, Chile, April 26-27. Asher, Mukul G., 1999,“The Pension System In Singapore” Social Protection Discussion Paper No. 9919, World Bank, Washington D. C. Atkinson, A B. and Joseph Stigliz. 1980, Lectures In Public Economics, London& New York, McGraw- Hill. Barr, Abigail and Truman Packard, 2002,” Revealed Preference And Self-Insurance: Can We Learn From The Self-Employed In Chile?” Policy Research Working Paper No. 2754, World Bank, January 2002 Barr, Nicholas, 2000,“Reforming Pensions: Myths, Truths, And Policy Choices” in IMF Working Paper 139, August 2000 International Monetary Fund, Washington D.C. Belan, Pascal, and Pierre Pestieau,1999,“Privatizing Social Security: A Critical Assessment”, The Geneva Papers On Risk And Insurance- Issues And Practice, Vol. 48, pp 5-22. Brown Dorothy A., Karen C. Burke, Grayson M. P. MeCouch-Social Security Reform: Risks, Returns and Race Volume 0 Cornell Journal of Law and Social Policy 633(2000) Burtless, Gary, 2000, Social Security Privatization And Financial Market Risk: Lessons From U.S. Financial History, Center On Social And Economic Dynamics Working Paper No. 10 (Washington: Brookings Institution). Callund, David, 1999,“Chile: Controversy, Difficulty And Solutions” The Geneva Papers On Risk And Insurance- Issues And Practice, Vol. 44, pp 528-33. Dei, G. Henry, 1993,“Implications For Social Security Of Structural Adjustment Policies In English-speaking Africa”, in The Implications For Social Security Of Structural Adjustment Policies, Studies And Research No. 34, International Social Security Association, Geneva Disney, Richard, 2000,“Crises In Public Pension Programmes in OECD Countries: What Are The Reform Options?” Economic Journal, Vol. 100 February, pp F1-F23. Gockel, A.F., 1996, The Formal Social Security System In Ghana, Friedrich Ebert Foundation, Accra, August, 1996. Holzmann, Robert, 1997,“Pension Reform, Financial Market Development, and Economic Growth: Preliminary Evidence From Chile” IMF Staff Papers, International Monetary Fund, Vol. 44(June), pp. 149-78. Holzmann, Robert and Truman Packard, 2000,“Extending Coverage In Multi-Pillar Pension Systems: Constraints, Hypothesis, Preliminary Evidence And Future Research Agenda”, Social Protection Working Paper NO. 0002, World Bank, Washington D.C. 40

James, Estelle, 1998, New Systems For Old Age security: Experiments, Evidence And Unanswered Questions”, World Bank Research Observer, August 1998, World Bank, Washington D. C. Kushner, Michael G., Virginia L. Briggs And Patrick M. Rockelli, 1989, Retirement& Welfare Benefit Plans: A Guide To Understanding The Tax Implications, The Bureau Of national Affairs, Washington D.C. Langbein, John H, and Bruce A. Wolk, 1995, Pension And Employee Benefit Law, University Case Studies, 2 nd Edition, Westbury, New York, The Foundation Press Orszag, Peter, and J. E. Stigliz,“Rethinking Pension Reforms: Ten Myths About Social security Systems”, in New Ideas About Old-Age Security: Proceedings Of The World Bank Pension Reform Research Conference, September 1999, ed. Robert Holzmann and J. E. Stigliz(World Bank, Washington D. C.) Packard, G. Truman, 2001,“Is There A Positive incentive Effect From Privatizing Social Security?- Evidence From Latin America” in Policy Research Working Paper No. 2719, World Bank, November 2001. Shultz, H. James, 1993,“Economic Support In Old Age: The Role Of Social Insurance In Developing Countries, in The Implications For Social Security Of Structural Adjustment Policies, Studies And Research No. 34, International Social Security Association, Geneva SSNIT, 2002, Medium-Term Strategic Plan: 2002- 2006, Social Security And National Insurance Trust(SSNIT), Accra, May 2002. SSNIT, Annual Report, Various Issues US Employee Retirement Income Security Act of 1974 US House of Representatives WAYS AND MEANS COMMITTEE HEARING ON SOCIAL SECURITY AND PENSION REFORM OF 1999 and 2001(DOC. 1999 – 5867& 2001 – 20563) Vittas, Dimitri, 1993,“Swiss Chilanpore: The Way Forward For pension Reform?” in The Implications For Social Security Of Structural Adjustment Policies, Studies And Research No. 34, International Social Security Association, Geneva WASTSON, Camilla E – The Pension Game: Age – and Gender-Based inequities in the retirement System 25 Georgia Law Review 1(1990) 41