PENSIONS AND SOCIAL SECURITY IN SUB-SAHARAN AFRICA ISSUES AND OPTIONS1 Luca Barbone Luis-Alvaro Sanchez B. Summary This paper presents an overview of Social Security reform issues in Africa, based on the examination of a number of case studies for both Francophone and Anglophone Africa. There are a number of conclusions that are drawn by the analysis of case studies: With very few exceptions (Mauritius, Botswana, and to a certain extent South Africa), formal social security institutions have not been successful in fulfilling their main mission, broad-based coverage of the population. What is more important, it is unlikely that pensions and disability coverage will be extended to the informal sector, which represents the vast majority of the employed population in Africa, within the next generation. As has happened in Europe over the past century, only sustained economic growth, formalization of the economy and reduction of the relative weight of the agricultural labor force will create the basis for some form of universal coverage. Thus, formal social security systems in Africa are at present a perquisite of the middle class. This has implications for the strategy of reform, and, in particular, for claims of public resources that can be made by existing institutions. Improving governance is the first order of business for existing formal social security institutions, which have all too often failed to deliver on promises to their members due to mismanagement (and sometimes outright pillage) of assets. The main tasks involve increasing transparency, curtailing opportunities for corruption, and most importantly protecting beneficiary rights. An important component of improving governance is the creation of protective barriers around the social security organizations, to prevent undue interference from outside interests, facilitate precise allocation of responsibilities and their oversight and enforcement. The creation or strengthening of regulatory institutions is an important step in this direction. Regulation is important whether management of these institutions remains in public hands or is transferred to private hands. Within the context of a strong regulatory framework, it will then be possible to grant social security institutions administrative and legal autonomy within clearly defined objectives. Further unbundling institutions according to the different services or activities they provide helps fine-tune the relationship between objectives, responsibilities, and incentives and thus improve governance. 1 Paper presented at the XIII International Social Security Association African Regional Conference, Accra, Ghana 6-9 July 1999. Luca Barbone is the manager for Central Africa and the Indian Ocean in the Poverty Reduction and Economic Management Network of the World Bank. Luis-Alvaro Sanchez B is a consultant for the World Bank. The views expressed in this paper are entirely those of the authors and do not represent the view of the World Bank. 1 Over the long term, formal social security institutions need increased membership to thrive. This buoyancy will not be possible if an adequate set of incentives is not in place. In many cases, rules today were developed in a context where government and its enterprises were the main providers of formal employment. As a consequence, contributions are high and benefits, at least as written, are generous. As the environment changes and the private sector takes the leadership in economic development, economic agents become very sensitive to the rules of the game. Social security schemes designed for the stronger economic agents who can contribute more, end up penalizing those at the margin--small and medium economic concerns--leading to increased informality. It thus seems appropriate that compulsory contributions go to finance minimum benefit standards and that additional benefits are the result of voluntary agreements between the parts. In pensions there are no boiler plate solutions. Much will depend on the initial conditions (benefit defined, provident fund, the level of contributions, the stock of reserves etc), the strength of the supporting institutions (finance), and the fiscal situation. This paper emphasizes the need for pension systems to be fully funded because of reduced fiscal risk and improved savings. But investing pension reserves is a difficult issue and there are no easy solutions. Investing abroad, which probably would be the best financially for the beneficiaries, is often politically unpalatable. Countries need thriving economies and strong financial sectors to provide profitable investment opportunities. When this is so, pension fund reserves can profit from the opportunity. When financial sectors are weak, concentrating fund reserves in public hands to protect them often leads pension institutions to try and become financial intermediaries with dismal results, including the further weakening of the financial institutions. This is an area where much work needs to be done. Hopefully, as pension reserves, whether under public or private management, grow, financial institutions will improve and the two can develop together. Already, as reported, countries in the continent have managed to develop significant voluntary private pension sectors. The transition to new designs has problems of its own and can affect the design of the most adequate solutions to be adopted. It is important to undertake rigorous technical work before settling down on a design. It is easy to find solutions that are appealing in the short-term, but that have detrimental long-term effects. The experience of some of the countries in the region already illustrated this trap. Improving administrative performance is another challenge that has been difficult to meet. Reducing costs, better service, and good record keeping is at the core of providing better services and eliminating the opportunities for corrupt behavior. But this is easier said than done. Worldwide experience shows that efforts to improve information management have to be accompanied by changes in the way institutions are organized to be effective. Thus, improving administrative performance will have to be part and parcel of the efforts to improve governance. Institutions with fewer and clearer objectives are in a better position to improve their performance and deliver better services. Lastly, issues like pension portability will become more important as labor mobility increases. 2 I. Introduction This paper reviews some of the main issues in social security reform in Sub-Saharan Africa, with emphasis on pensions, and evaluates reform options available to countries in the continent. This preliminary report is based on work the Bank is currently undertaking to study the pension and social security systems in Sub Saharan Africa in an effort to help countries arrive at improved social security designs suited to local characteristics. The Bank is not alone in these efforts. Governments and other donors also seek ways to address the social security concerns of broad segments of the population. Consensus exists with regard to some critical areas. The importance of finding institutional designs that improve governance and enhance credibility is readily acknowledged by all. The need to have a broad view of social security, including those outside the formal sectors, is also a point of convergence. But, in spite of all the substantive and abundant work undertaken so far, much still remains to be done, especially when it comes to the design of equitable and sustainable institutions. These are auspicious times as many countries are engaged or thinking about substantial reforms of the social security systems. The paper is organized as follows. It begins by outlining basic characteristics that frame social security consideration in the continent. The population is young and working mostly outside the formal sectors. Formal security institutions in many countries have lost credibility and it seems as if only the culture of social solidarity extended throughout the continent provides the basis to build a broad social protection umbrella. The paper then provides a brief overview of the main characteristics of the current social protection schemes. Finally, the paper presents an overview of the main challenges ahead and comments on the options available and their relevance in the African context. Improving governance is the most critical challenge. Given the problems in the past, often business as usual is not the best option and new institutional setups must be found. The aim at the end is to protect the rights of the beneficiaries by providing better services, reducing administrative costs, raising the returns to investment, and contributing to overall development by carefully considering the links with the rest of the economy. The overhaul of social security institutions carries with it important transition considerations that if not taken into account, will risk the sustainability of the designs. 3 II. Background Well-designed social security policies can contribute to a better environment for economic growth in Africa. It would be erroneous to think that social security is a luxury to be afforded only when growth has taken place or when countries have reached a certain level of per capita income. The fact is that there is a mutually reinforcing relationship between economic growth and society’s ability to deal with the consequences of social and economic uncertainty. Adequate frameworks to deal with uncertainty will help improve the allocation of resources and hence contribute to economic growth. At low income levels there are substantial market imperfections that prevent people from dealing adequately with risks, and this determines the type of activities they undertake and the efficiency of their investment. But it would also seem unwarranted not to take into account the limitations that the level of income imposes on the ability of government to effectuate transfers across groups. As income increases, it is easier for governments to tax and transfer resources to help the less fortunate deal with the results of economic uncertainty. Disregard of the budgetary effects and limitations will have harmful effects. Straightforward application of external social security designs that do not take account of the specifics of the African continent are likely to be inadequate, as the crisis of the systems inherited from colonial times has shown. This section highlights some basic facts that frame social security considerations in Sub-Saharan Africa. These pertain to the demographic characteristics of the population, the basic organizational structures of the African economies and the sources of economic uncertainty, the loss of credibility that characterizes formal social security systems, and the cooperative culture of the African continent. Demographic considerations. Demographic characteristics are important for social policy to identify and measure the target populations. Social security difficulties are often associated with aging populations, be it from the viewpoint of pensions or health. African populations on average, however, are young compared to other continents, as is shown in Table 1. The ratio of the population over 60 to total population stood at 4.7 in 1995 in the continent, the lowest ratio in the world. Likewise, the ratio of population over 60 to the population between 15 and 60 stood at 9.3 -- also the lowest ratio in the world. African populations are even younger that those of the low-income countries of the world. The ratio of population over 60 to total population in low-income countries worldwide is 6.1. 4 Most of the countries in the continent have young populations. 2 Upper middle countries have older populations, with ratios of population over 60 to total population at 7.1, but there are only five such countries with small populations. (See Table 2 and Figure 1.) Over the long-term, expected changes in fertility and mortality rates will reduce population growth rates and eventually increase the proportion of the elderly in the population. But this transition will take a long time; the bulk of the African population will continue to World: be young for a while. In fact, in Low Income 6.1 11.2 many countries the age dependency Lower Middle 9.5 15.4 will decrease over the next twenty Upper Middle 8.7 14.5 years. It will increase thereafter to High Income 18.0 28.8 reach, by 2040, levels similar to those of Latin America and the Caribbean today. The aging of the population is expected to take place in the second half of the next century. (See Table 3.) The variation among countries, however, will increase for a while and decrease in the long run, 3 meaning that countries will get older at different rates. Table 2. Aging and Dependency Indicators By Income Level, 1995 Dependency Ratios 60+/Total 60+/15-59 Sub-Saharan Africa: 4.7 9.3 Low Income (n=38) 4.5 9.0 Lower Middle (n=4) 5.4 10.4 Upper Middle (n=5) 7.1 12.2 Table 3. Projection of Percentage of Population Over 60, Sub-Saharan Africa 1995 2015 2040 2060 2080 2100 Sub-Saharan Countries 4.7 4.4 7.5 12.8 19.4 22.9 Low Income (n=38) 4.5 4.2 7.1 12.3 19.1 22.7 Lower Middle (n=4) 5.4 5.3 10.0 16.3 22.1 24.4 Upper Middle (n=5) 7.1 7.7 15.0 21.0 24.7 26.6 Within countries, populations split into different demographic profiles. Fertility and mortality are higher among the poor and life expectancy lower. These differences affect behavior and the demand for social security services by the different income groups. Health and prevention for epidemic diseases may be a greater concern for the poor. AIDS is shortening life expectancies in all groups and again altering the demand for social protection. Still, differences in mortality rates tend to disappear as people get older. Life expectancy for elderly people is similar regardless of the level of income. Thus, people that have reached retirement age are likely to live over a decade more, regardless of socioeconomic status. So provision for the old is also a concern for the poor. 2 A set of appendix tables presents detailed population information by country. The variance over the mean of the population over 60 total population increases for a while to converge in the long-term. (See Table 3.) 3 5 Structure of the economy and sources of economic uncertainty. Social security policy addresses problems that arise when markets do not provide adequate mechanisms to deal with economic uncertainty or when shortsighted behavior exposes individuals to controllable risks. It is then necessary to ascertain the nature of the risks individuals face in a given society to tailor adequate policies. Many of the operating instruments for social protection in the world today have been designed to address problems characteristic of industrial economies, where large segments of the population have regular employment. In Africa, rural and informal economies are very important. The formal sector (where people have continuous employment and income) is very small. Moreover, in some countries, employment in the formal sectors continues to be highly dependent on the state, be it directly or through state enterprises. Table 4. Percentage of Economically Active Most of the people in Sub-Saharan in Agriculture, Sub-Saharan Africa, 1990 Africa are still employed in agriculture, Sub-Saharan Africa 68 whether they are classified as living in Low Income (n = 38) 75 Lower Middle Income (n = 4) 48 the rural or the urban areas. In 1990, 68 Upper Middle Income (n = 5) 32 percent of those economically active derived their income from agriculture or agriculture related activities. This percentage was even higher, 75 percent, for low-income countries in the region. (See Table 4) Even when people move to urban centers, they are likely to continue to derive their income from agriculture. Urban centers have yet to generate significant sources of steady employment. So most people’s incomes are subject to the variations proper of agricultural activity or the fortuity of informal activities, which themselves may fluctuate considerably. Add to this environment, the regular calamities of nature and man, and the uncertainty of life is enormous. It is often said that traditional societies have developed mechanisms to deal with these uncertainties, and considerable amount of work has been done to describe the social protection mechanisms. The adequacy and coverage of these mechanisms is not sufficiently understood. Whatever the social protection mechanisms in traditional societies, the case is that the elderly work longer later in life in low income countries, where over 60 percent of the people over 60 years of age are economically active. (See Appendix-Table 3) That is, the elderly have to provide for themselves either totally or partially. Moreover, traditional societies are in the process of transformation and social cohesion may be deteriorating. Already it is reported that poverty among the elderly remaining in the countryside may well be on the increase. 6 Lack of credibility of the formal systems. In many of the countries in the continent, formal institutions, public or private, lack credibility. This extends to social security institutions, especially pensions, and derives from the poor services provided to beneficiaries and the mismanagement of pension reserves. People came to view social security contributions more as a tax and sought to develop, within their means, alternative protection schemes. At the same time, financial institutions are not trusted as a result of periodic sector crises. This generalized lack of credibility severely constrains the development of social protection mechanisms that necessarily have to rely on savings and insurance to meet economic uncertainties. Rebuilding credibility in social security and finance institutions requires addressing severe governance shortcomings in the public and the private sphere. A culture of solidarity. A common observation made by both local and external observers is the importance social solidarity plays in the workings of African societies and how traditional societies have developed mechanisms to deal with economic uncertainty. With the transformation that is taking place and people moving to the urban areas, reportedly this culture of cooperation has been transformed into new mechanisms of cooperation. The question is how strong a role can these informal mechanisms play on improving social security among the population. The Bank now seeks to identify and classify the different types of risks that characterize these economies and the social protection mechanisms that exist. Hopefully this work will serve as the basis to help improve social protection policies in the continent. In any case, this culture of solidarity stands in contrast to the credibility problems of the formal institutions, and could be a base upon which to build new schemes of social security protection. These ideas remain to be fully developed and applied. 7 III. The Current Situation. While most Sub-Saharan countries share broad demographic, economic and institutional characteristics, as discussed above, there are also significant variations in the design and performance among the existing pension and social security arrangements, formal or informal. This variety of institutional designs and outcomes prevents easy generalizations as to the most adequate social security designs. Careful study and review of existing systems is an important pre-requisite for reform. This section looks the variety in designs, the following sections will seek to identify key issues. 4 Variations in formal social security systems arise from how they were initially designed and how they have evolved over time. Many of the arrangements were inherited from colonial times or adapted from foreign designs, often from industrialized countries. Still, it is the case that even countries that shared a common origin have tended to diverge over time. Most African countries have some form of social security arrangements for people working in the public sector, in state enterprises, or in the enterprises in the modern sectors. The parallel systems operating in each country are often dissimilar. Table 5 describes several features of the main social security arrangements for a broad set of countries, taken from indirect sources.5 Social security arrangements. Most of the countries provide a broad range of social security arrangements that include pensions and other benefits. French-speaking countries, for instance, tend to fall in this category. In these systems, pension arrangements are of the defined-benefit type.6 Besides pensions for retired and pensions for widows and surviving children, other benefits include work accidents, maternity and children allowances. Ejuba, Ewane J ( “Pension Schemes in Africa” 1999 ) presents an overview of the current social security systems of the African countries (including North Africa). 5 Additional information is provided in Appendix Tables 4, 5, 6 6 The pension benefits in the defined-benefit approach are calculated based on formulae that include the length of contributions, reference salaries, minimum number of years for eligibility, etc. In defined contribution systems resources available for retirement come from savings accumulated during the lifetime plus their return minus administrative charges. 4 8 Table 5. Contribution Rates for Social Security Programs Types of Total Contributions Employer Contribution Shares Soc. Sec. All Pension Social Share of Share of Programs Soc. Sec. Pension as % of Security Pension Soc. Sec. Pension Incl. 1/ (a) (b) Soc. Sec. (c) (d) (c/a) (d/b) Benin 4/ O, S,W,F 23.2 10.0 45 19.6 6.4 84 64 Burkina Faso O,W,F 23.0 9.0 39 18.5 4.5 80 50 Burundi O,W 10.5 8.5 81 7.5 5.5 71 65 Cameroon O,W,F 19.0 7.0 37 16.2 4.2 85 60 Cape Verde O,S,W,F 27.0 10.0 37 20.0 7.0 74 70 Cent. Af. Rep. O,W,F 20.0 5.0 25 18.0 3.0 90 60 Chad 4/ O,W,F 14.3 6.0 41 12.3 4.0 86 67 Congo, D. O,W,F 18.28 7.0 56 15.88 3.5 87 50 Congo, R. O,S,F 18.48 6.0 32 13.83 3.6 75 60 Côte d'Ivoire 4/ O,S,W,F 15.1 4.0 27 12.9 2.4 85 60 Eq. Guinea O 26.0 21.5 83 Ethiopia O 10.0 6.0 60 Gabon 2/ O,S,W,F 22.6 7.5 33 20.1 5.0 89 67 Gambia 3/ O,W 20.0 19.0 95 19.0 19.0 95 100 Ghana O 17.5 12.5 71 Guinea O,S,W,F 23.0 6.5 28 18.0 4.0 78 62 Kenya O 10.0 5.0 50 Liberia O,W 7.75 6.0 77 6.0 3.0 77 50 Madagascar O,W,F 14.0 4.5 32 13 3.5 93 78 Mali 4/ O,S,W,F 23.0 9.0 45 21.0 7.0 91 78 Mauritania O,S,W,F 16.0 3.0 19 15.0 2.0 94 67 Mauritius 5/ O, 9.0 6.0 67 9.0 6.0 100 100 Niger O,W,F 17.0 4.0 24 15.4 2.4 91 60 Nigeria O,S, 7.5 5.0 67 Rwanda O,W 8.0 6.0 75 6.0 3.0 75 50 Sao Tome O 10.0 6.0 60 60 Senegal 4/ 6/ O,S,F 35.0 14.0 40 8.4 72 60 South Africa O,U 2.0 0.0 1.0 0.0 50 0 Sudan O,W 27.0 25.0 93 25.0 17.0 93 68 Swaziland O 10.0 5.0 50 Tanzania O 20.0 20.0 100 20.0 10.0 100 50 Togo 4/ O,S,W,F 20.5 6.0 29 18.1 3.6 88 60 Uganda O,W 15.0 15.0 100 10.0 10.0 67 67 Zambia O 10.0 5.0 50 Zimbabwe O 6.0 3.0 50 Source: U.S. Social Security Administration, 1997, except where indicated. 1/ Only includes social security programs financed through earnings related contributions. O is for old age, disability and death. S is sickness and maternity. W is for work injury. F is for family allowances. When contributions consist of a range, the highest value is shown. Blank spaces indicate that some or all of the contributions in column (a) are not earnings related. 2/ Gabon also has the “retraite viellesse” (ACE) program, where the insured pays 5% and the employer pays 2%. 3/ Pension scheme shown. Gambia also has a provident fund where insured pay 5% and employer pays 10% 4/ Source: CIPRES, 1999 5/ Earnings related pension shown. Mauritius also has an universal pension which is entirely financed by the government. 6/ Regime General shown. Senegal also has a pension regime complementaire (insured 2.4%, employer 3.6%) 9 The overall contribution rates in these systems tend to be high, as they cover pensions and other benefits. They reach up to 31/35 percent of salaries in Senegal and quite a few countries hover above 20 percent. The average total social security contribution stands at around 19 percent, and the median at 20 (See Table 5 column 1.) 7 Pension contributions can be quite low in some countries. The lowest are Mauritania, Côte d’Ivoire and Madagascar, where the contribution for pensions stands at 3, 4 and 4.5 percent respectively (Table 5, column b). Of the 35 countries included in Table 5, 28 have pension contribution rates at or below 10 percent. The low pension contributions follow from the fact that initially the resources needed to pay pension benefits were small. The receipts from higher social security contribution rates have gone to finance generous family benefits. As pension outlays have grown at a rate faster than the allotted contributions, internal transfers have been necessary to finance pension obligations. (More on this below) Employers pay the bulk of the contributions, both for overall social security and for pensions (See Table 5). Retirement ages range between 55 and 60, but most systems allow early retirement at about 50 years. The minimum number of years to qualify for a pension can be quite high, reaching 20 years in some countries. Failure to reach this benchmark means that the contributors can get back only his/her nominal contributions. In some countries pensions for widows and survivors can be quite generous. Replacement rates vary with contribution rates. It can reach 80 percent for public servants in some countries (Côte d’Ivoire, etc.). In most countries the difference between basic and total wages (basic plus allowances) is a source of ambiguity that creates opportunities for evasion. Provident Funds. Provident funds focus exclusively on retirement and do not include other benefits. They operate as compulsory individual savings accounts, with beneficiaries entitled to a lump sum at retirement. Although annuities are possible, often people prefer to take the lump sum. Several provident funds have been transformed into social security arrangements with pensions organized around a defined benefit principle. This has already happened in countries like Ghana and Nigeria. Most recently (1998) in Tanzania, Parliament approved a law transforming the National Provident Fund (NPF) into a broad social security arrangement covering pensions and other benefits (including possibly health insurance) —National Social Security Fund. This transformation has yet to be fully implemented. Kenya and Uganda retain the provident fund design. Contribution rates among provident funds vary considerably and are very high for the NPF in Tanzania--20 percent of salaries.8 Contributions to the Uganda provident fund stand at 15 percent--10 for employers and 5 for employees. Kenya’s contribution stands at 10 percent, but with an effective ceiling for most workers--around three dollars, half for the employees and half for the employers. This extremely low level is the result of contribution rates fixed in nominal terms during the years. Additional information on SSA’s pension systems is available in the appendices. The NPF is now the NSSF. Besides the NPF, there is the PPF, covering parastatals, with a defined benefit scheme since its creation in 1987. As parastatals are privatized, PPF has sought to continue covering the emerging private concerns and to extend its coverage beyond in the private sector. 7 8 10 Other type of systems. Mauritius has a different history. The National Pension Fund (NPF), created in 1978, is a contributory arrangement, partially funded, where the members purchase pension points with their contributions, to be redeemed later for pension benefits. The basic ratios are roughly indexed to inflation and have remained relatively constant over time. In Mauritius, the standard contribution to the NPF is 9 percent--6 for employers and 3 for employees. However, there is a cap at a relatively low level of income and the effective contribution for medium and high salaries is much lower. Public Pensions. Civil servants often have one or more social security arrangements. In a good number of cases pensions are paid out of the budget and there is no pension fund per se. Often the pension rules covering civil servants or public employees vary considerably. In Cameroon around seven different regimes for public servants can be found. In countries like Kenya, Uganda, and Tanzania, local governments have their independent pension arrangements. In French speaking countries, not all public employees are covered by the civil service rules; some of them belong to the funds covering private employees. All of this creates a lot of confusion and inequities. Universal pensions. Mauritius pays a pension out of the budget to anyone over 60 years old. Initially Mauritius experimented with a means-tested pension scheme out of the general budget. Application complications led to the introduction of a universal pension, which has become pretty much ingrained in the population. This pension is adjusted regularly, although no precise formula exists. The universal pension covers roughly one third of the average wage. Total expenditure on the universal pension stands at 1.3 percent of GDP, and surveys show that it is an important instrument to sustain income of the less fortunate. South Africa has a means-tested universal pension. Gabon also has a universal pension. Private sector arrangements. There is a wide variety of private pension arrangements in Africa. Some are set up by individuals and some by companies, with contributions both by employers and employees. Three factors seem to favor the emergence of private pension arrangements: (1) low contribution rates to the public funds, (2) adequate tax treatment, and (3) a working financial sector with some tradition in the management of pension accounts. These pension accounts can be of different types: defined-contribution, definedbenefit, and provident fund. Private non-compulsory private arrangements are strong in Kenya, where more than a thousand can be found. Reportedly, the preferred private option in Kenya is defined-contribution. Defined benefit arrangements have decreased in popularity, because they require the employer to balance the funds, if needed after periodic actuarial examinations. The assets under management by private pension funds stand at 10 percent of GDP, roughly the same percentage as the public provident fund (NSSF). In Kenya there is a welldeveloped infrastructure of brokers, fund managers, and insurance companies and significant competition in the sector. As already mentioned, the contribution levels are low and tax law allows a deduction for investments in pension funds both for employer and employees. This industry could play a very important role in the region. These remarkable 11 developments not withstanding, some concerns remain regarding (1) the coverage of the combined public/ private arrangements, and (2) the transparency of the existing private schemes--costs, for instance. Still, quotes obtained from different sources would suggest that the total costs of administering pension funds in the private sector stand at around 1.5 percent of assets, which seems reasonable given the size of the market. Mauritius has also developed a significant sector of complementary private pensions. The office of the Income Tax Commissioner reports over 900 independent arrangements, additional to around 70 covered by the Parastatal Pension Act. The information on the private arrangements is limited and it is difficult to estimate its size from available information. It is likely that jointly, the private and the parastatal arrangements have assets equal to half of the assets under NPF management. The private arrangements are complementary to the public schemes and reportedly cover mostly medium and high level employees, even though the law orders full employee coverage to be eligible for tax benefits. Zimbabwe also has significant private sector pension arrangements. They are not customary in Tanzania and Uganda, probably as a result of the high contribution rates to the public pension systems, (which are 20 percent for Tanzania and 15 percent for Uganda), and credibility problems of the financial sector. In French-speaking Africa and where contribution to the public pension system is low, private pension funds can also be found. Countries in these regions tend to have favorable tax treatment for private pension funds. Overall, there is limited information of the extent to which private arrangements are used to provide for pensions. 12 IV. Key Issues This section reviews key pension and social security issues that have been identified in the sample countries under study. These issues are not intended as generalizations that apply to all African countries. Exceptions can readily be found. The intent rather is to present a sample of the challenges faced by countries that seek to reform or improve their social security systems. Credibility Problems. A good many formal social security systems of Sub-Saharan Africa have experienced serious credibility problems due to many factors, such as low benefit levels, long lead times to process claims and poor services. Credibility problems have affected all systems--provident fund or defined contributions. Provident funds problems derived from the fact that the benefits they provide correspond to contributions plus returns, so that all risks are shifted to the beneficiaries. Thus the mismanagement of reserves, administration costs, and poor administrative performance led to meager pensions, in the best of cases. In many other cases, records can’t even be located. These problems help explain the move away from provident funds that has taken place in some countries. The credibility problems of the defined benefit schemes are similar to those of the provident funds regarding services provided. However, since benefits are defined, the risks of poor performance, mismanagement of funds, etc. are transferred to the government or to future generations. Unfortunately, however, beneficiaries have not been protected as well as one would expect from theory. The reason is that often governments, either by fiat or neglect, adjust benefits downward by failing, for instance, to index them to inflation. In such cases, the consequences of poor performance are partially transferred to beneficiaries. Serious governance problems underlie the loss in credibility. At the root of the poor governance is a faulty institutional design. In many countries, the key problem has been the interference of the government in the management of the funds. This interference has been encouraged by institutional designs that give government control of governing boards and the social security administrations. Besides government interference, administrations themselves are often times captured, distort the social security agencies and their statutory objectives and use them for their own purpose. Limited coverage of the formal social security systems. Formal social security systems cover a small percentage of the population in most African countries. In pensions, coverage is low, both in terms of the total number of contributors over the economically active labor force, and the number of beneficiaries over as compared to the population over 60 (Table 6.) That the coverage in relationship with the overall population is low should not be surprising, since, as we have seen, the size of the formal labor force is small to begin with. Coverage in terms of the relevant target population is substantially higher, especially when formal sector employment is dominated by the public sector. Pension expenditure in 13 relative terms is still small in Africa. The average ratio of pension expenditure to total GDP for the countries included in Table 6 stands at 0.55 percent. Formal social security systems highly dependent on the state. The state has played and continues to play an important role in the development and the characteristics of the formal pension and social security systems. The oldest arrangements have been those covering public servants. Some of these remain from former colonial times, with few changes. Moreover, in many countries a considerable percentage of the people covered by formal social security schemes are either employed by the state or by state enterprises. In Togo, for instance, only one third of those covered by formal social security arrangements can be considered to be working for the private sector. Although this percentage is higher in other countries, public employment policies still have a significant effect on the development of formal social security schemes. High Administrative costs and poor administrative performance. The cost of administering social security systems in SS-Africa appears to be high. In the countries reviewed the ratio of administration costs over contributions varies considerably. It can be as low of 12 percent in Mauritius when considering just the public pension fund and around 3 percent if the transfers for the universal pension are taken into account (1995). It reaches over 100 percent in Kenya (1997)9. Certain care should be taken in interpreting this type of data. This indicator, administration costs over total collections, depends not only on administrative performance but also on the contribution rates. Thus, for instance, in Kenya administrative costs exceed contributions, but the contribution rate is extremely low. Thus while indeed it may be a fact that administrative expenditures are relatively high, the indicator exaggerates the extent of the problem in this case. Also, systems that provide a broad range of benefits will have to have, on average, lower ratios of administrative expenditures over revenues. While it may be difficult to judge the efficacy of one organization compared to another (within a country or across countries) using broad indicators, it is still the case that in most countries social security organizations in SubSaharan Africa face considerable administrative difficulties. A complete set of consistent comparative administrative costs for the countries in the continent have yet to be put together. Explanations of the poor administrative performance will vary from country to country. Often there has been a lack of incentives for the boards and management to improve administrative performance. Existing incentives on the contrary have led to extensive employment, nationwide networks, and considerable expenditures on building and infrastructure. The absence of hard budget constraints on administrative expenditures has contributed to the deterioration of pension reserves. In between, the ratio of administrative expenditures to revenues stands at 13 percent for Uganda’s NSSF, 24.5 percent and 18 percent for Tanzania’s NPF and PPF respectively, and around 40 percent for Cameroon’s CNPS. 9 14 Table 6 Labor Force and Pension Coverage Benin Burkina Faso Burundi Cameroon Cen. Af. Rep. Chad Congo, D. Côte d'Ivoire Gabon Ghana Guinea Kenya Madagascar Mali Mauritania Mauritius Mozambique Niger Nigeria Rwanda Senegal Tanzania Togo Uganda Zambia Labor ContribForce Year utors Coverage Year 1989 1989 152,443 3.7 1993 1990 4.7 1993 1989 597,452 13.7 1993 1989 1990 1.1 1990 1992 1989 9.3 1989 1991 1989 1,100,000 13.3 1993 1993 1990 1,400,000 14.7 1993 1990 261,469 5.4 1993 1990 2.5 1990 1989 1989 1986 1,248 0.5 1993 1990 108,656 2.8 1992 1990 1,000,000 2.4 1993 1989 315,217 9.3 1993 1990 220,542 6.9 1992 1990 642,600 5.1 1992 1993 1989 1989 359,620 13.8 1994 Covered Contrib- PensPensWage Pension utors/ ioners/ Ioners/ Bill/ Spending Labor persons ContribGDP /GDP Force over 60 Utors 3.5 0.4 0.3 3.1 2.4 9.5 5.0 0.2 3.3 8.7 23.4 5.5 0.3 13.7 3.9 2.3 0.0 1.1 0.9 5.8 11.0 9.3 5.7 12.3 7.3 5.7 0.1 7.2 1.5 6.8 0.5 25 5.4 4.1 8.0 0.4 2.5 6.5 0.2 20.3 2.8 100 0.0 5.0 0.1 1.3 0.1 1.3 9.7 9.3 5.6 4.9 6.9 4.3 0.5 6.6 5.4 0.1 0.8 0.1 10.2 Source: International Patterns of Pension Provision. Palacios/Pallares 1998, ILO, EPF, Staff Estimates and CIV. Besides the high administration costs, there have also been problems with the quality of the services delivered. Good record keeping, which is crucial to a well-managed social security administration, has been lacking. Funds have had problems tracking individual accounts or producing updated valuations of their assets. The end result has been long lags between retirement and the time beneficiaries begin receiving pension payments. Also, the lack of adequate record keeping provides an opportunity for corrupt practices as it makes it difficult to control the oft-found practice of creating fake employment files to collect pensions. Many social security institutions have sought to improve their administrative practices, often with external technical assistance and have centered on information management and record keeping. While progress is reported, continued administrative difficulties suggest focusing first on the basic institutional design, before engaging in micro administrative reforms. For instance, social security institutions undertake a wide variety of tasks including collection, account management, investment of reserves, benefit payments, and 15 often other tasks on behalf of the government. To carry out multiple missions within a single organization has proven complex. Difficulties managing pension reserve funds. Most pension funds in Sub-Saharan Africa, provident or otherwise, have generated surplus resources. Management of these reserves has proven problematic and has contributed to existing and forthcoming difficulties. There are multiple reasons why management of funds has been difficult. To begin, governments have had a considerable degree of say on how pension surplus funds are utilized. In almost all countries, government has either borrowed or appropriated resources from the pension funds. The ways in which this was done vary by country. Only in a few countries and very recently, have treasuries issued debt at market interest rates to the pension funds. As a consequence, in many cases the interest rates were below market returns, and in some cases below inflation. Countries are finding it increasingly necessary to clarify the accounts between the government and the pension funds. 10 Besides the direct use of resources by the public treasury, governments have also directed pension funds to invest in specific projects or companies. These investments have not always been fortunate. There have also been problems with the decisions of fund management regarding where to invest. A favorite area of investment has been real estate. Problems in this area have come from inflated purchase prices and high construction costs. The end result has been that pension funds are not adequately funded. This shows up and poses different problems in provident funds and in defined-benefit arrangements. In provident funds, the impact of fund mismanagement falls on the contributors. Thus, the returns to investment that are credited are often below the market rate of return. Even so it may be that not enough funds are available to cover the obligations in the books to beneficiaries—the sum total of savings plus the returns. In this case the provident fund itself is unfunded.11 In cases like Kenya and Uganda, the extent to which provident funds can actuarially cover their obligations to the beneficiaries depends on successfully finishing on-going real estate projects and the returns to previous real estate investments. The problems that follow in defined benefit schemes are slightly different. Poor reserve fund management will affect the extent to which the funds are balanced, but not necessarily the benefits extended to the beneficiaries, at least in principle. The fact is however, that loss in reserves ends up pressuring an increase in the level of the contributions or lowering benefits indirectly, such as failing to adjust pensions for inflation, or adjusting them too late. 10 In addition to the misuse of pension reserves, in some countries, governments or their enterprises often fail to pay their contributions to the funds. 11 This is so, because it has been difficult for provident fund administrators to keep an accurate and up-to-date value of their investments and their returns. 16 Lack of investment opportunities. The problems with the management of pension reserve funds do not arise exclusively from bad governance. It has also been the case that over the last decades investment opportunities have been very limited. It appears, in fact, that investment returns in the sub-continent have been low if not negative over the last decades. The profitability crisis and the crisis of the financial sector have not been any less dramatic than the social security crisis. The result has been a loss of credibility in both the public and private sectors. Issues specific to the defined benefit schemes. Defined benefit schemes as found in Frenchspeaking Africa face their own specific problems. These systems, as mentioned, combine pension and non-pension benefits, and pension contribution rates have been low, as pointed out above. Initially, these schemes generated operating surpluses. Over time, however, pension obligations have grown faster than contributions. This may appear odd in countries with young populations and fast growing economically active populations. However, the demographics of the formal pension systems are considerably different from the population Figure 2. Pension Contributions and Benefits in Cote d'Ivoire. 40,000 35,000 Millions of Franc CFA 30,000 25,000 20,000 15,000 10,000 5,000 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 at large. The number of beneficiaries contributing to the formal pension funds stagnated over the last decade, while the average age of its membership increased. As a consequence, the ratio of beneficiaries over contributors increased dramatically over time12, resembling the characteristics of a quickly maturing pension system. In Cameroon, between 1991 and 1996 the ratio of total beneficiaries to contributors increased from 8.8 to 16. In the public sector the same ratio has gone from 9.7 in 1990 to 31.25 in 1997. In Côte d’Ivoire the 12 Also, the ratio of beneficiaries to contributors in the population is much higher than the ratio of over 60 to the overall labor force. 17 number of beneficiaries has grown over the last ten years at 7.9 percent, while the number of contributors has remained stagnant. It is likely that this situation will worsen in the near future, as membership in the formal pension systems is still linked to public sector employment and civil service reforms and privatization are slowing or even reducing the employment provided by the public sector and its institutions. Over the longer term, much will depend on the evolution of the formal sectors as economic recovery gets underway. In some countries, like Cameroon, recent growth has taken place outside the formal sectors, and so membership and contributions to the social security institutions has not picked up. So far the solution has been to redirect resources from other benefits to pay for pensions. In some countries this is no longer feasible. In countries like Cameroon, Togo, and Côte d’Ivoire, pension expenditures already equal pension revenues, so that they do not cover administrative expenditures. (See Figure 2 for the case of Côte d’Ivoire.) Now countries face hard choices. The expedient solution is to increase the contribution rates. This however, can have undesirable side effects in the medium term, as it can contribute to a continued stagnation of the membership. Also, it can create pressure to increase the benefits (replacement rates) with the consequence that the beneficial effect of the increase in rates will only be short-lived. A more considered approach will have to include rethinking the level of other benefits, reaching an understanding with government on the outstanding obligations and the form of payment, and, most importantly, taking measures to improve governance and performance before committing to increase pension contribution rates. Public service pension obligations. Social security and pension obligations for public employees raise important problems of their own. When there are independent pension systems for public employees, a problem has been that governments did not pay their contributions, thus de facto appropriating the surplus that would have been generated. With pension funds in problems, it is necessary to clear the accounts between the government and the funds, and agree on mechanism for reimbursement. In the absence of a public pension fund, all of the current obligations fall on the budget. The end result has been that the payments for pensions have been increasing. This has been further accelerated by retrenchment. Additionally, civil service reforms have implied substantial salary increases, leading to higher pension outlays now and in the future. This has led some countries to try and set new public pension funds. Salary increases have been used to allow employees to contribute to the funds. Uganda is going through this process now. Provident fund difficulties. Some of the problems with provident funds have already been mentioned. The consequence of administrative and financial mismanagement falls on the beneficiaries. Low pension benefits made the contribution seem like a tax. The result has been the tendency to move away from defined benefit schemes. It is clear that the beneficiaries did not have the mechanisms to monitor management and the government. 18 An option would have been to allow employees and/or employers to exit and form complementary or substitute funds. While this could have been possible (it has happened in other countries outside Africa with provident fund designs) it did not happen to a significant extent. In some countries, like Uganda, the public pension fund successfully argued against these options. Pension Portability. In most countries under review, movement from one pension system to another is likely to imply a loss of pension benefits. This results from two facts. Usually it takes a long-time to be eligible for a pension. Also, countries do not have provisions allowing contributors to carry benefits from one pension system to another. The most obvious example is when moving from the public to the private sector or vice-versa. Unless the contributor has met the minimum years of work to qualify for a pension, movement will imply a loss, since often the person can only take his/her contributions and not the employer’s. There are other examples. Movement across countries, as in West Africa, will imply loss of accrued benefits. Portability is also important in company pension schemes. Movement from one pension to another will imply a loss of pension benefits, when the employee can only take its contribution and not the employer’s. Lack of pension mobility then becomes a mechanism to retain labor. Regulation. There is very limited experience in pension regulation in Sub-Saharan Africa. Public pension schemes are subject to the laws that created them. The Ministries of Labor and Finance often exercise oversight. Private pension funds often fall under the jurisdiction of insurance regulators. Pension specific regulation has recently been introduced in Kenya. In French-speaking Africa, CIPRES plays the role of a regulator, but is still in a developing stage. 19 V. The Path of Reform: Challenges and Options The challenge in many African countries continues to be the building up of credible and sustainable social security institutions, anchored on a clear concept of equity and supported by rules of the game that assure adequate governance. Development of a broad approach to social security that includes all persons both inside and outside the formal system, can only be developed gradually and as a result of the efforts countries undertake to design and implement new social protection schemes. The initial focus will in most cases be on the reform of formal social security arrangements, to an extent greater than is probably warranted. But, there is hardly any choice given the current difficulties with formal social security systems. In doing so, it is necessary to proceed gradually and by taking careful stock of the existing situations. Social security reform is a difficult area of social policy, since the entitlements it creates are difficult to reverse later on, and long-term effects are difficult to visualize without the support of detailed technical analysis. Solutions that may appear attractive in the short-term can end up having very negative long-term effects. Countries must avoid at all costs strategies that can not be sustained over time. Improving social protection for the informal and traditional societies is going to take time, as it requires identifying mechanisms that are tested and feasible. This can only be done by using the lessons that will be gathered from the multiplicity of efforts to improve social protection coverage in the continent. Overtime, it is likely that the problems with the elderly in the informal sector will increase. By then, hopefully the mechanisms will be in place to allow societies to supply them with adequate protection. The overriding motif of what follows is governance understood in a broad sense. Welfare considerations: Equity The objective of social security policies is to help individuals and families deal with the uncertainties of economic life and correct myopic behavior. Equity considerations are at the heart of social security policy. The poor are the most exposed to the uncertainties of economic life and the least prepared to deal with them. The government has basically two instruments to carry out social security policy. It can use the budget, which is the basic element of solidarity in an economy. Taxes are collected to be spent according to a variety of criteria that should include protection of the most vulnerable--children, mothers and the elderly. Besides the budget, countries can set up mechanisms to provide protection for limited groups. The state participates by creating an enabling environment and overseeing or enforcing its implementation, without necessarily providing budgetary support. Classic systems are Pensions, health, accident insurance, and family support. 20 The limited coverage of the formal social security systems in Africa raises serious equity issues, in the sense that valuable resources and efforts may be targeted to a small and privileged segment of the population. This is a very valid concern and certainly exclusive emphasis on the formal system would be unwarranted at the present time. Unfortunately, however, countries need to meet head-on the difficulties of the on-going social security concerns and, in some cases, doing so requires substantial budgetary transfers to meet past obligations—contributions that were not paid or resources taken that have not been returned. Moreover, honoring entitlements in the current benefit schemes may require additional budgetary transfers, if increases in contribution rates are not desirable or feasible. Towards the future, states will have to depart from previous practices that have used arguments of equity and solidarity to the benefit of a few. This is very germane for countries considering significant overhauls of their social security systems. The danger now as in the past is to increase entitlements without regard for budgetary limitations or concern for those excluded from the formal systems. For even though social security arrangements appear as mechanisms of solidarity amongst clearly defined populations, the fact is that often the state ends up guaranteeing the benefits promised. For instance, in the case of compulsory defined benefits pension schemes the state, in many countries, has in effect been liable for pension benefits when the funds have not been available to meet them. Governance The foremost priority in the reform of formal social security systems in SS-Africa is to improve governance and by doing so, build up their credibility. If governance issues are not forcefully addressed, other actions to deal with the present problems will only have a shortterm and limited effect. For instance, contribution rate increments to eliminate deficits will serve only as temporary palliatives and problems will resurface later on. Indeed, it is advisable that governments first address governance issues before or as part of engaging in social security reform. Experience worldwide shows that social security systems (particularly the pension components) have easily been captured by special interests that proceed to distort the original intent of the institutions. These special interests include the government-- Ministry of Finance and Labor or even higher offices--and the administrations of the social security services. The end result has been high administrative costs, low returns to investments, poor benefits and service for the beneficiaries. To this there are exemptions, as is the case of Mauritius. The institutional design should seek overall to protect the interests of the beneficiaries. Since these interests are dispersed, it is necessary to set up the mechanisms to monitor performance, enforce the rules of the game, and provide transparent information to the beneficiaries. The arrangements so far in many countries have fallen far short of meeting 21 these criteria. Often yearly balances are not produced or mandatory audits are not performed. Today, in many countries, it is difficult to know the value of the assets or the outstanding obligations to beneficiaries. Clearly, the mechanisms to enforce the rules of the game have not been effective. Governments lacked the incentive to pressure for transparency, as they themselves were culprits in the mismanagement of these institutions. Employer and employee representatives, most often appointed by government, are in a minority position. The only option for beneficiaries has been to exit the system, if possible, or simply take contributions as a tax and make alternate arrangements. On-going social security reform activity all over the world, especially in the pension area, is seeking to grapple with these issues. To do so, reforms have sought to radically overhaul the governance structures. Key elements have been: Redefinition of the role of the state, focusing priorities on the provision of adequate social security frameworks and limiting direct management and financial exposure. Unbundling the provision of social security services, to facilitate management, increase transparency and facilitate clear assignment of responsibilities. Redefinition of the role and composition of the managing boards providing administrative and legal autonomy, defining clear responsibilities, and introducing oversight and control mechanisms. Making explicit any intra and inter generation transfers. Introduction of competition into the management of social security reserves. Providing beneficiaries with a greater menu of options, either by allowing them to select the pension provider or by having various pillars to introduce diversification. Introduction of regulatory frameworks to guarantee that resources are well managed, that the state is kept at arms length, and that beneficiaries are well informed on funds performance. Taking account of the consequences of social security reform on other sectors, like labor and financial markets. The adequacy of these reforms can only be considered on a country by country basis. However, in all there is an urgent need to improve governance. Unbundling A useful first step of where to begin the reform of the formal pension system is to take stock of the different services provided by the various social security institutions and evaluate the convenience to have the same institution provide them. In the simplest case, that of a provident fund, the institution: (a) collects contributions, (b) maintains contributors and the beneficiary’s records, (c) assesses the pension benefits and pays them, and (d) manages the reserves. More complex social security organizations like those found in French-speaking Africa, must additionally administer family benefits and work accident insurance. All of these functions are often not undertaken with separate administration and accounting. As has been shown, unintended cross-subsidies between activities (pensions, family benefits, and work accidents) can emerge. The multiplicity of objectives and the 22 lack of clearly defined responsibilities among different activities make it very difficult to manage these multi-purpose institutions. To address these difficulties, the tendency has been towards unbundling the management of the different services provided by social security institutions. To begin this can be done by breaking up the provision of the different services into separate organizations or under an umbrella organization. Some further unbundling may be necessary at the level of each service. Countries like Cameroon and Côte d’Ivoire have been considering reforms along these lines. The size of the Formal Social Security Sector As has been shown, the formal economic sectors are small in most Sub-Saharan African economies. Social security contributions center on a limited number of incorporated enterprises and the public sector. Over the long-term, countries should look for the social security systems to expand at rates equal to or greater than economic growth. High and sustained growth will come under the leadership of the private sector. A buoyant social security sector will come about only if the economic agents have the incentives to incorporate and work in the formal economy that provides continued employment and salaries. This will not happen if the appropriate set of policies is not in place. There is a long-term trade-off between the size of the formal social security system and its coverage. Policies that seek extended and sizeable benefits for formal sector employees will have the effect of reducing the contributory base. The reason is that higher replacement rates, for instance, require larger contribution rates, that in turn have negative consequences on the labor market. Participation in the formal system decreases. This is a problem everywhere but more so in Sub-Saharan Africa where the supply of capital and labor to the formal sectors is very elastic.13 Solutions can be tried to mitigate these effects. For instance, they can increase investment in enforcement, but at some point the higher administrative expenses can significantly reduce the resources available for benefits.14 Benefits can be tailored to induce people to remain in the system—minimum pensions and family benefits are an example. But this itself is limited by the budget available to provide these incentives. 13 Wage surcharges, either from the employer or the employee, to finance social security contributions have an effect on the labor market leading to a reduction of wages or employment. Some labor groups often think that increased benefits amount to a transfer from employers to employees and disregard the side effects. This analysis may be relevant for a limited group of workers at a given point in time, but not for the aggregate labor force in the long-term. Compulsory rules that cover all the labor force should take account of how policies affect the marginal contributor. Special cases where employers and employees agree to additional benefits, because of the specific industry characteristics, are best left to be arranged between the concerned parties on a voluntary basis. 14 Experience from taxation shows tax rate increases have to be accompanied by increases in enforcement expenditures, if government want revenues to increase proportionally. At some point compensating increases in enforcement expenditure will not be enough. 23 These considerations are quite timely given that African countries face pressure to increase social security contribution rates to close fiscal gaps or improve benefit coverage. There is not a precise mathematical formula that can be used to calculate adequate contribution levels. Experience shows that the level of the social security contributions to the compulsory systems goes up with per capita income, partly because it is easier to collect contributions. From a more substantive perspective, it must be noted that highly competitive economies take care not to distort factor prices unduly. Mauritius, which is a highly competitive economy, has contribution rates that on average are around 6 to 7 percent. There are many ways in which a social security system can be adjusted to meet financial crises without having to increase unduly contribution rates. Options include increasing the retirement age, eliminating exemptions and broadening the contribution base, introducing more stringent rules on early retirement, and streamlining non-pension benefits. Moreover, greater administrative efficiency and improved returns on fund reserves have powerful effects in the long-term. Of course, total social security coverage is not determined exclusively by the compulsory formal system. Complementary arrangements between employers and employees or by employees themselves play an important role in developed and developing countries, and have to be taken into consideration when taking stock of social security coverage. Pensions The three-pillar pension framework has become a standard reference that can serve as a framework within which to analyze the options available to Sub-Saharan countries. Options, however, are rarely independent of the existing conditions, because transition can be costly and difficult. The Three Pillar Framework The first pillar of a pension system seeks to provide a minimum income to the elderly, usually defined as those over 60 years of age. Countries often finance this pillar from the general budget, but do not cover all of the elderly and use means testing to determine eligibility. As mentioned above, Mauritius and to a lesser extent Gabon provide a universal pension. For most Sub-Saharan countries, fiscal limitations and difficulties with means testing prevent consideration of this option for a while. However, limited support can still be given to the elderly in the context of anti-poverty programs. (See below.) A limited version of the first pillar seeks redistribution among those contributing to the formal pension system(s). Different degrees of redistribution are possible, including providing uniform pension to the members. But, redistribution mechanisms have incentive consequences whose effects have to be taken into account when considering the design of the pension system. Certainly a minimum pension provides an incentive to those with low 24 incomes to remain in the system and contribute, if the minimum requirements for eligibility are enforced, which is difficult in countries with poor record keeping. On the other hand redistribution represents a tax for the high-income groups and, as all taxes, negatively affects their participation, limiting the degree and amount of resources available for redistribution. For this reason, countries often cap the contributions to this pillar. These considerations limit the extent of the redistribution and the size of the first pillar. In countries with very broad formal sector coverage the government can transfer resources to help with the pension equalization. When the coverage of the formal pension fund is small, however, financial support from the general budget to the funds would seem unwarranted on equity grounds. As a general rule, then, a first pillar based on redistribution would best be of limited size. Countries with defined benefit pension systems have the basis in place for a first pillar and, in fact, some of countries in the region already have redistributive elements in their pension designs, including a minimum and a maximum pension. Note, however, that a first pillar of the type described above is also possible within a defined contribution system. This is obtained, for instance, by taxing participants above a certain income to provide transfers for low-income members. Thus it is not necessary to have a defined benefit pension arrangement provide for intra generation solidarity. The second pillar is a defined contribution system operating as a compulsory or forced savings mechanism of individual accounts. Beneficiaries receive annuities upon retirement. On the surface there are similarities between defined contribution schemes and provident funds, as these are also schemes of individual savings accounts. A major difference between the two is the institutional design that supports the contributory systems as currently implemented. This institutional design introduces rules or norms to prevent poor pension fund management. Key elements are the creation of a regulatory body to oversee pension funds and the introduction of competition for the management of the funds. Hence, the poor performance of the provident funds is not a sign that defined contribution schemes would not work in the African context. In fact they have performed well as voluntary private arrangements in countries where public provident funds have experienced difficulties, as has been the case in Kenya. The third pillar covers voluntary private pension arrangements. As mentioned, voluntary pension arrangements already thrive in African countries like Kenya, Mauritius, Zimbabwe, and to a lesser extent in several other countries. A key factor that has facilitated the emergence of private pension arrangements has been tax rules. Countries in the region exempt from taxation contributions and interest returns and tax benefits. Even though this form of taxation is formally neutral, it has been the subject of some controversy, as it may end up implying a transfer for those taking advantage of the tax provisions who are likely to be relatively well off. Tax issues aside, the important point to highlight is the ability of the financial sector in these countries to organize and manage private savings accounts. However, because private arrangements may not be as transparent, regulation is a necessary complement. 25 Countries will do best to build on existing private pension systems rather than crowding them out with the introduction of compulsory arrangements, either defined contributions or defined benefits. It is advisable to take stock and account of private pension arrangements when undertaking reform. Doing so will greatly help the development of the financial markets. While the three pillars provide a general framework that allows for a wide variety of designs, much remains to be decided regarding the level of the contributions, the distribution among pillars, the selection between defined benefit and defined contribution approaches, governance, etc. Defined benefit and Defined contribution. Given the controversy surrounding the relative merits between defined benefit and defined contribution approaches, it is appropriate to revisit the issues and try and dispel some misunderstandings. One critical difference between the two approaches is as follows. In defined benefit schemes, the fund guarantees a pension based on a set of rules that include the years of contribution, the levels of contribution, and the reference salary for pension calculation. In the defined contribution scheme the pension to be received by the beneficiary will depend on the contributions, the returns, administration costs, and the costs of setting up the annuities. It could be said that the defined benefit approach seeks to smooth out some of the risks of the financial markets and of administrative mismanagement. The pension fund takes these risks and eventually passes them to the government or to future generations. Unfortunately, in many developing countries, defined benefit schemes failed to keep their promises and have adjusted their balances by reducing benefits; failure to correct for inflation is a typical mechanism. Thus defined benefit schemes have not been a guarantee of more secure pensions. Since effective pensions in the defined contribution approach can be risky as they depend on the financial returns and costs of administration, its designers have sought to address these potential difficulties through regulation. Some countries guarantee a minimum return to investments, often requiring the fund manager to put up capital to cover low or negative investment returns and by imposing strict rules on portfolio allocations. Limits can also be set on administration costs. Lastly, the option for contributors to move among funds can serve to improve management performance. Another critical difference between the two systems is the structure of incentives. In the defined benefit systems participants do not have incentives to improve management or monitor performance if they perceive that their pension is guaranteed. The incentives for doing so are with the government or with coming generations, since poor performance can imply future budgetary transfers or higher contribution rates. But many times governments do not take a long-term view, and future generations do participate in current policy discussions. 26 With defined contribution systems, beneficiaries have a direct incentive to seek improved management and will do so if the adequate oversight instruments are available. Also, defined contribution systems have positive effects on savings and, in a growing economy with high returns to investment, will need lower contribution rates to provide a given replacement rate. Differences between the two systems are not as sharp with regard to other issues. The idea is that defined contribution scheme is fully funded; that is, its assets correspond to the liabilities. But the defined benefit schemes can also be funded to different degrees. At one extreme is a pure pay-as-you-go system that collects enough revenues to meet the obligations and perhaps keep available some liquidity. In the long-term, this approach requires increasing contribution rates, borrowing, receiving government transfers or a combination of the above. This may be difficult to achieve and in Latin American countries, for instance, the need to increase contributions to fund the formal pension arrangements has been one of the reasons for the recent pension reform wave. This illustrates how difficult it is to enforce inter-temporal contracts as future generations may renege on initial agreements, once they see how funds have been performing. At the other extreme is a highly funded scheme that seeks to keep assets equivalent to the expected obligations to the beneficiaries. This requires periodic actuarial analysis to determine needed adjustments to keep the fund balanced. This type of arrangement is found in company pension funds all over the world. It is, however, difficult to manage and implies additional risks for enterprises. Reportedly, this practice is being gradually substituted by defined contribution schemes. As to public defined benefit schemes, while they could be highly funded, the case is that at best in reality they are partially funded. One reason is that when the schemes are introduced, pensions are paid to beneficiaries who have not fully contributed. This generates a transfer across generations, which is not necessarily a characteristic of defined benefit systems but that often accompanies them. This transfer operates as a tax on younger generations and affects the level of benefits the funds can provide. Transition issues. In moving from one pension design to another, there are a variety of transition issues, which themselves impact the availability of feasible solutions. It seems opportune to consider some of these transition issues at this time, as countries in the region are likely to face them at some point or another. The type of transition that has been studied the most is that from defined benefit to defined contribution schemes. This has been typical of Latin American countries that have adopted defined contribution systems. The key issue here is the financing of the debt implicit in the defined benefit scheme. This debt corresponds to value of the acquired rights of those that have already retired plus those that have contributed to the pension system, minus the assets of the fund. As the defined benefit is phased out, pension contributions are no longer available to pay for pension obligations, and instead go into the individual savings accounts. There are several options to finance the implicit debt. Existing fund reserves, if they exist, can be used. Some countries have used proceeds from privatization. More 27 generally however, these obligations have to be met with resources from the general budget or through borrowing, thus being paid by either current or future generations. The obligations, of course, need not be met outright and can be honored by issuing bonds to the beneficiaries to be paid as they come due in the future. In countries with small formal pension systems the above strategy will imply a transfer from the general budget to a small group of individuals, a point to be considered carefully. Transition from defined benefit to defined contribution schemes requires very prudent fiscal management and thus a necessary step in before undertaking a reform of this type is to calculate the implicit debt of the existing defined benefit systems to get an idea of the fiscal requirements. Countries in fiscal difficulties may find it difficult to undertake reforms of this type. Transition from provident funds to defined benefit systems has been motivated by the poor performance of the former and the meager pensions or lump sum payments it provided to beneficiaries. The move to defined benefit is attractive because it allows the government to improve outlays without increasing contributions. If the contribution rate is high enough, it will even be possible to increase the number of benefits. This “miracle” will be possible because of the fund moves from a funded to an unfunded scheme. The long-term consequences need not surface for a while. If administration is efficient and if the returns to investments are competitive, the fund can be financially sustainable for a considerable period of time. But eventually the fund will begin using its reserves, and at some point it will have to increase contributions or curtail benefits. A provident fund can easily move to a defined contribution system. Key issues here are the introduction of annuities, some form of life insurance, and adequate governance structure. This transition can come about gradually. A first step can be to open pension management to some competition under a regulatory framework. This can be done, for instance, by allowing large enterprises to set substitutive pension arrangements and later on allow new pension providers to come in. A more difficult problem is found when, within a defined benefit, countries are unable to increase contribution rates without simultaneously raising replacement rates. As a result, the room provided for by the rate increase is only temporary; further increases are looming down the line. This situation is partially alleviated if the government steps in to pay the obligations it owes to the funds, so that they become highly funded. But if the government does so the condition is set to move to defined-contribution systems. Improving management of fund reserves. We have advocated the need for pension funds to be funded, even if they are benefit defined. The advantage of being fully funded lies in reducing the fiscal risk and avoiding negative effects on savings. But as we have seen already, the management of pension reserves has been highly problematic in the past, so it is fair to ask what conditions will guarantee improved results this time around. This is a very difficult question to answer satisfactorily. A first step is to seek professional and specialized management of pension reserves. Some countries seek to do this by setting special investment committees, usually under the 28 Ministry of Finance. A more aggressive move is to contract out management of pension reserves, subject to a regulatory framework. This is roughly what defined contribution arrangements do. Something similar could be done for the reserves of the defined benefit arrangements. Even if an appropriate governance framework is set up, the question of adequate investment instruments remains. In the past, investment has not been highly profitable in Africa. Financial products are very limited, even government paper. Direct lending or investment in companies by the pension fund would require the internal development of evaluating infrastructure, which amounts to converting pension funds in financial intermediaries. This is a path full of difficulties. No doubt, the limited investment opportunities have to be taken into account in the design of the pension systems. One alternative is to invest part of the funds abroad. Some countries allow this. But it is a politically difficult question, since policy makers associate the availability of savings with domestic investment and development. Most countries in Africa have weak and shallow financial sectors. This and credibility problems prevents them from being a strong underpinning of privately managed pension system. On the other hand, reserve funds in the hands of the government or the public pension institutions weaken the financial sectors, as it subtracts valuable resources. If the environment for private economic activity and the financial sector improves, it will be easier to find investment outlets in the future. Since fund reserves will grow only gradually, it would seem that the appropriate strategy is for the financial sector and pensions to develop in parallel. Pension reform and savings. There is a tight link between the design of the pension system and the savings behavior of the population. Two observations are particularly relevant. First, defined benefit systems that are scantly funded have negative consequences on national savings. The reason is that the promise of a pension lowers the savings of the beneficiaries. For this reason it is very important to have funded pension systems.15 Second, compulsory pensions that substitute for voluntary pension arrangements are likely to have only limited net effects on savings. Thus, when introducing compulsory defined contribution systems it is important to take into account the effect on private pension schemes. The crowding out effect on savings will thwart financial development. Administrative issues. Administrative reform is needed to improve the quality of the services provided and to reduce costs. Improving quality will help rebuild the credibility of the social security institutions. Reducing costs will help tremendously the finances of the institutions over the long-term. Most countries need improvements in both areas. Significant improvement most likely will not come about through traditional strategies that have centered their efforts on the information management techniques, for instance. As it is 15 Some may argue that since the introduction of compulsory savings is partly to prevent myopic behavior, then it can be safely assumed that people would not have saved in the absence of the pension scheme and hence that there would not be a drop in income. 29 well known, effectiveness of new technologies will depend on the ability to streamline processes and realign objectives and incentives within the institutions. This often requires major reforms similar to those needed in tax administration. Administrative reforms need to deal with the multiplicity of functions that social security institutions undertake. Some of these areas are hardly related with each other. One of these areas is the collection of contributions. Some institutions considered efficient by international standards, like the social security administration in the US, focus on a limited set of activities, mainly record keeping and providing services to beneficiaries. The Internal Revenue Service collects social security contributions as part of the tax collection. Some countries have recently separated out the collection of social security contributions and assigned the task to a specialized agency. There is a certain reticence in many countries, and not only in Africa, to trust tax administrations with the collection of social security contributions. There are two main reasons for this mistrust. The first is that social security contributions are likely to be a marginal activity for tax agencies and thus they may not assign the sufficient resources to enforcement and book keeping. Moreover, social security often requires keeping personal accounts and tax agencies often do not keep individual records when taxes withheld by employers are considered final. This is not suitable for pension schemes, like defined contributions, that require personalized accounts. This may not even be suitable for defined benefit systems, as accurate records of contribution are also needed. The second reason for mistrust derives from the customary appropriation of pension reserves by ministries of finance. The price for this mistrust is high. Social security institutions fail to take advantage of the economies of scale and specialization provided by the tax agencies. Under these circumstances, it is advisable for countries to consider assigning collection responsibility to one specialized agency subject to a hard budget constraint. 16 Improving portability. Portability is increasingly becoming a problem in most countries and has negative effects on labor mobility. The way pension systems are designed often implies that people end-up losing benefits when they move from one pension system to another. This issue is important within as well as among countries. Within countries, there is the mobility between the public and private and pension systems. Some countries, like Côte d’Ivoire, are taking measures to assure portability. The basic idea is to permit contributors to sum up the number of years worked in each system in determining eligibility for benefits. Since the two systems offer different benefits, final pensions are paid in proportion to the years worked by each one of pension institutions. Pension portability is also highly relevant in countries that have developed complex private pension schemes. When moving from one employer to another, workers run the risk of losing benefits. This mechanism can be used to block labor mobility and hoard labor. Mauritius has introduced legislation mandating mobility in new pension arrangements. Portability problems are also important between countries. This is particularly the case when there is 16 In the case of autonomous civil servants pension funds, a way to minimize costs is to use government infrastructure for the payment of salaries and record keeping, rather than duplicating it. 30 considerable labor mobility among countries, like in French-speaking Africa. Here workers can work a lifetime and not qualify for any pension benefits. The relevance of pension regulation. It is difficult to overstate the importance of adequate regulation in improving governance of social security arrangements, pensions in particular. Regulation is relevant both for private as well as publicly mandated systems. In the case of private arrangements (be it individual or enterprises), regulation is important to increase transparency and protect beneficiaries. Private pension arrangements as they have emerged in some countries often fail to provide accurate information regarding costs, returns and balances, making it difficult for beneficiaries to compare among different providers. Insurance companies as part of their life business often provide private pension arrangements. Insurance regulation not only tends to be new but also does not have the expertise to monitor private pension providers. Pension specific regulation can help. Regulation is also important and relevant for the mandatory social security schemes. Traditionally, the statutes that created them bind formal social security arrangements, and oversight and control corresponds to the oversight Ministry, often Finance, Labor or Social security. In some cases there is dual oversight. In Mauritius, for instance, the general oversight is the responsibility of the Ministry of Social Security, but the investment committee is under the Ministry of Finance. Board composition is often heavily laden in favor of government representatives, with token representation from employers and employees. Efforts to reform these arrangements are underway in several countries. Boards have been given the authority to appoint management in Kenya. The participation of employees and employers through representative organizations has been increased. Some countries are thinking of handing management and responsibility to employees and employers to manage these institutions within the framework of the law. These ideas however will fall short in the absence of appropriate regulation. Regulators can put a fence around these institutions and protect them from external or internal perils. As already mentioned, beneficiary interests are highly dispersed. Employee or employer representation is not enough, as the representatives have limited interests. The regulator has to protect beneficiaries very much as capital market regulation seeks to protect the rights of small stockholders. The pension regulator must pay attention to all business aspects verifying implementation of the regulations. Regulation can set bounds on administrative costs, set investment guidelines, and require minimum financial returns. It can also set rules to increase transparency requiring timely production of yearly balances and mandatory actuarial reports. The regulator can also solve disputes that arise between parties in the sector. Aside from enforcing regulation, the regulator can take a pro-active role in the development of social security framework over time. Kenya has recently introduced pension-specific regulation--the Retired Benefits Authority (RBA)-- to oversee all pension arrangements, including NSSF and public service. The 31 results will provide a valuable input the development of other regulatory agencies in the continent. The tasks faced by the new regulatory agency are considerable, and they have chosen to implement legislation gradually. The Kenyan experience shows that there is considerable overlapping between pension, insurance, capital markets, and banking regulation. This points to the need to consider a broader regulatory agency covering all of these areas, as has already happened in other countries. Indeed, Kenya is planning to introduce a Financial Services Authority that would include all financial sector regulation. Similar plans are underway in Mauritius. CIPRES, in the French-speaking Africa, reportedly plays the role of regional regulator. A Broad Social Security Net A sizeable percentage of the children born today in Africa will spend their lives in the traditional or informal sectors, with little hope they will be covered by formal social security arrangements. Thus, the most pressing challenge in most African countries is the development of a broad social security umbrella that covers a significant portion of the population. Many isolated efforts are taking place. These include the reform of the formal systems, as discussed above, and initiatives to develop forms of social security coverage in the informal and traditional sectors. Communities, NGOs, governments, and international donors have been involved in these initiatives. Comprehensive and feasible social security strategies have yet to be developed. Expansion of formal institutions to the rest of the population does not appear a feasible strategy. The best that can be done in this regard is to try and cover all areas of formal employment. But, even here the administrative costs of enforcing rules among very small enterprises may outweigh the benefits. (Already some countries have introduced special collection methods for small enterprises.) Expansion to the self-employed and the informal sectors is definitely problematic and probably not advisable, especially when the credibility of formal institutions is at the low end. But there are technical problems as well. Traditional and informal sectors do not generate smooth income flows and thus individuals can not commit to steady contributions. This creates difficulties for the publicly mandated systems and for private arrangements as well. Pension/Saving plans where individuals commit to regular contributions over a long period of time have fallen into difficulties when economic activity slows down with heavy penalties for individuals and problems for the financial intermediary. The alternative is the development of different institutions, which build on a culture of cooperation and existing community or extended family arrangements. The relevance of social protection mechanisms in traditional societies has been recognized for a long time, and research in this area has been increasing. These institutions could help solve problems of credibility and enforcement. Unfortunately, most information on existing arrangements is descriptive and thus it is difficult to ascertain their coverage and effectiveness. Still, some broad observations can be made. Traditional or informal social security arrangements generally focus on specific products rather than broad coverage schemes. For instance, tontines (in West Africa) represent alternative saving mechanisms. Often the members of a 32 tontine make a cooperative arrangement to help procure medicaments or pay for basic medical expenses. Community organizations may procure medical service packages for their members for specific interventions—malaria, etc. Pooled resources are often used to address critical bottlenecks such procuring ambulance services to guarantee reaching hospitals in time. All of these isolated packages point to the areas of importance to the communities, as well as the ease to contract and enforce obligations. Income smoothing and health concerns are definitely a priority. The parallel financial arrangements found in the informal sector result from the lack of credibility in the formal institutions and the high transaction costs in using them. That is, people perceive a high risk of losing their savings and an inability to provide resources when in time of need. Thus, the development of credible and efficient financial systems that takes into account the needs of the poor and those in the informal sector is a necessity to help people smooth their income flows and have resources to deal with uncertainty. This requires creating legal and institutional frameworks that allow the informal institutions to operate without burdening them with excessive formal requirements, and providing security for the participants. Community or group financing of health services must be an integral part of country health strategies. As opposed to income smoothing and health insurance mechanisms, the emergence of pension arrangements in the informal and the traditional sectors is far more difficult. Reportedly, traditional agricultural societies have mechanisms to provide for the elderly. These mechanisms vary from place to place, but generally are tied to the ownership of land. The elderly retain control of the land as a mechanism to guarantee care from their children.17 These mechanisms not withstanding, anthropological studies show that increasingly the elderly are becoming destitute in the countryside. In the informal sectors, pension arrangements are difficult to agree to and enforce because they are by nature longterm. 18Aside from enforcement difficulties, populations with low life expectancy will be less willing to undertake long-term agreements. Moreover problems like AIDS introduce uncertainty into formal or informal insurance. The challenge for social policy is how to introduce designs that improve on the voluntary agreements that develop naturally. The danger is that policy intervention will worsen outcomes. These developments can happen when public policy alters incentives in such a way that they reduce the social protection efforts now being undertaken. For instance, children may stop providing for their parents because of availability of public support. A consensus seems to be emerging that the government must operate through the development of regulatory frameworks that facilitate the operation of community organizations that provide social protection. These organizations tend to have credibility and capacity to enforce agreements at the local level. The question is how to aggregate 17 In some cases the youngest will have the responsibility of caring for the parents and as compensation inherits the main house and adjacent land. 18 Even governments have failed to enforce pension contracts in the formal sectors. 33 these efforts in such a way that credibility and ability to enforce are maintained in the larger organizations. The larger organizations will be able to benefit from economies of scale and scope and be in a better position to work with formal sector organizations. It is likely that problems with the elderly in the traditional and informal societies can not be solved through voluntary agreements and that at some point, states will have to provide some support through budgetary resources. Given the limited availability of budgetary resources and the multiple claims on them, it is likely that only very selective interventions are possible. These selective interventions can take place in the context of anti-poverty programs, be they developed by the countries or linked to debt-reduction programs. These interventions would have to be based on some form of means testing and carefully targeted to avoid perverse incentives. Clearly this is an area where there is ample scope for countries to learn from each other experiences. Social Security, Pensions and the Bank The Bank is increasingly focusing attention on social security and protection in SubSaharan Africa. Two initiatives are particularly relevant. The first focuses on issues of social protection in traditional and informal sectors. To this effect, the Bank has developed a framework identifying the risk face by populations and the mechanisms available to mitigate them. It has also issued guidelines for social protection work in the region. This work is intended to identify the effectiveness of social protection institutions and existing gaps. The challenge is for governments to build institutions that protect broad segments of the population with the cooperation of civil society. An in-depth pilot study has been initiated in Togo this year. The second area of work is pensions. The Bank has already provided or is providing support to several countries for the reform of their pension systems. The Bank intends to deepen its involvement in Pension and Social Security Reform, as Sub-Saharan African countries have begun to focus on the reform of their formal social security systems. To begin, the Bank is taking stock of the pension and social security issues in Sub-Saharan Africa. Some of the observations presented here are the result of this work. As the diagnosis of these issue advances, the Bank will involve countries in the region in a dialog to develop broad social protection strategies and improve governance of formal social security institutions. 34 Conclusions This paper has emphasized the urgency of improving governance of formal social security institutions and doing so within the broad framework of a social protection strategy. A broad social protection strategy will take time to develop, and in the meantime the priority is with the formal arrangements. Improving the social security governance is truly a test of a government's commitment to institutional reform. The main tasks involve increasing transparency, curtailing opportunities for corruption, and most importantly protecting beneficiary rights. Since governments and even social security administrations have contributed to the mismanagement of social security institutions, an important component of improving governance is then to create protective barriers that foster good governance and penalize poor governance. The creation or strengthening of regulatory institutions is an important step in this direction. Regulation is important whether management of these institutions remains in public hands or is transferred to private hands. Within the context of a strong regulatory framework, it will then be possible to grant social security institutions administrative and legal autonomy within clearly defined objectives. Further unbundling institutions according to the different services or activities they provide will help fine-tune the relationship between objectives, responsibilities, and incentives and thus improve governance. Over the long term, formal social security institutions need increased membership to thrive. This buoyancy will not be possible if an adequate set of incentives is not in place. In many cases, rules today were developed in a context where government and its enterprises were the main providers of formal employment. As a consequence, contributions are high and benefits, at least as written, are generous. As the environment changes and the private sector takes the leadership in economic development, economic agents become very sensitive to the rules of the game. Social security schemes designed for the stronger economic agents, who can contribute more, end up penalizing those at the margin--small and medium economic concerns--leading to increased informality. It thus seems appropriate that compulsory contributions go to finance minimum benefit standards and that additional benefits are the result of voluntary agreements between the parts. In pensions there are no boiler plate solutions. Much will depend on the initial conditions (benefit defined, provident fund, the level of contributions, the stock of reserves etc), the strength of the supporting institutions (finance), and the fiscal situation. This paper has emphasized the need for pension systems to be fully funded because of reduced fiscal risk and improved savings. But investing pension reserves is a difficult issue and there are no easy solutions. Investing abroad, which probably would be the best financially for the beneficiaries, is often politically unpalatable. Only countries with thriving economies and strong financial sectors can provide profitable investment opportunities. When this is so, pension fund reserves can profit from the opportunity. When financial sectors are weak, concentrating fund reserves in public hands to protect them often leads pension institutions to try and become financial intermediaries with dismal results, including the further weakening of the financial institutions. This is an area where much work needs to be done. 35 Hopefully, as pension reserves, whether under public or private management, grow, financial institutions will improve and the two can develop together. Already, as reported, countries in the continent have managed to develop significant voluntary private pension sectors. The paper presented a brief review of the three-pillar framework and commented on each pillar, taking into account the experience in African countries. A broad first pillar, financed by the budget and based on means testing, would seem to be out of the reach of most countries in the region. Moreover, some would argue that there are more important priorities where such resources could be spent. A first pillar limited to those contributing to the formal sector is possible within defined benefit or defined contribution schemes. The size and convenience of such a scheme will depend on the need and possibilities for redistribution. If the pillar is targeted only for redistribution, as suggested here, the pillar will be small. The second and third pillars are highly related. Often purely voluntary pension schemes do not provide extended coverage, and leave people out in lower ranks or in small enterprises. Compulsory defined contribution schemes address this difficulty, but place the resources in few hands. The solution in some countries has been to engage the private sector in the management of the reserves. Management in a third pillar is more dispersed and hence could lead to high administration fees. However, it provides muchneeded diversification and can help the development of the financial sector. This subject has to be further researched by studying conditions in countries with extended voluntary pension plans, like Kenya. The transition to new designs has problems of its own and can affect the resulting institutional arrangement. It is important to undertake rigorous technical work before settling down on a design. It is easy to find solutions that are appealing in the short-term, but that have detrimental long-term effects. The experience of some of the countries in the region already illustrated this trap. Improving administrative performance is another challenge that has been difficult to meet. Reducing costs, better service, and good record keeping is at the core of providing better services and eliminating the opportunities for corrupt behavior. But this is easier said than done. Worldwide experience shows that efforts to improve information management to be effective have to be accompanied by changes in the way institutions are organized. Thus, improving administrative performance will have to be part and parcel of the efforts to improve governance. Institutions with fewer and clearer objectives are in a better position to improve their performance and deliver better services. Lastly, issues like pension portability will become more important as labor mobility increases. This paper has merely scratched the surface of social protection issues in Sub-Saharan Africa. A broad agenda remains ahead. Work under way will examine in detail individual country cases in an effort to develop the capacity to support countries in their reform efforts. 36 Appendix Table 1. Demographic Indicators, Sub-Saharan Africa, 1997 GNP Total Growth Infant Fertility Life Death Rate, HIV Income Per Cap Populatio Rate 2/ Mortality Rate 3/ Expectancy crude AIDS Class. Atlas (mills) 1997-2010 Rate (thou.) at Birth (per 1000) Rate 4/ n (1000) Angola L 260 11.7 48.6 125 6.7 46.5 18.6 1.0 Benin L 380 5.8 45.4 88 5.8 54.7 12.2 0.9 Botswana UM 3,310 1.5 35.2 58 4.2 50.3 13.2 12.5 Burkina Faso L 250 10.5 53.9 99 6.6 46.0 17.9 3.3 Burundi L 140 6.4 41.4 119 6.3 47.1 16.6 4.1 Cameroon L 620 13.9 44.5 52 5.4 56.5 11.2 2.3 Cape Verde LM 1,090 0.4 38.8 56 3.6 66.5 6.9 Cen. Afr. Rep. L 320 3.4 32.1 98 4.9 48.6 16.6 5.3 Chad L 230 7.2 29.4 100 5.5 48.6 16.9 1.3 Comoros L 400 0.5 93.2 65 4.6 59.8 9.1 Congo, D. L 110 46.7 51.4 92 6.2 52.9 13.7 2.0 Congo, R. L 670 2.7 46.9 90 5.9 50.9 14.9 3.6 Côte d'Ivoire L 710 14.2 35.6 87 5.0 53.6 12.3 4.9 Djibouti LM 1/ 781 0.6 106 5.4 50.3 15.0 5.2 Eq. Guinea LM 1,060 0.4 43.5 108 5.5 50.0 16.4 0.6 Eritrea L 230 3.8 28.4 62 5.8 55.0 11.8 Ethiopia L 110 59.8 51.6 107 7.0 50.0 16.2 4.3 Gabon UM 4,120 1.2 31.8 87 5.0 55.1 14.2 2.0 Gambia L 340 1.2 28.6 78 5.2 53.2 13.0 1.1 Ghana L 390 18.0 46.4 66 4.9 59.1 9.7 1.2 Guinea L 550 6.9 53.1 120 5.7 46.5 17.3 1.0 Guinea Biss. L 230 1.1 32.6 130 6.0 43.8 21.2 1.1 Kenya L 340 28.6 38.2 74 4.4 58.1 9.3 5.6 Lesotho L 680 2.0 48.4 93 4.5 58.6 10.4 4.0 Liberia L .. 2.9 55.5 116 6.3 51.5 15.4 1.8 Madagascar L 250 14.1 68.7 94 5.7 58.5 10.5 0.1 Malawi L 210 10.3 39.5 133 6.4 42.8 20.5 7.0 Mali L 260 10.3 64.8 118 6.6 50.4 15.9 0.8 Mauritania L 440 2.5 33.3 92 5.0 53.5 13.5 0.3 Mauritius UM 3,870 1.1 19.9 20 2.1 71.6 6.3 Mozambique L 140 16.6 53.1 135 6.1 45.5 18.2 6.6 Namibia LM 2,110 1.6 38.9 65 4.8 55.6 11.7 9.3 Niger L 200 9.8 52.4 118 7.4 46.9 18.0 0.7 Nigeria L 280 117.9 44.4 77 5.3 53.5 12.3 1.9 Rwanda L 210 7.9 25.1 124 6.0 42.1 19.2 6.3 Sao Tome L 290 0.1 50 4.7 64.0 9.5 Senegal L 540 8.8 41.1 70 5.6 50.7 13.8 0.9 Seychelles UM 6,910 0.1 15 2.3 71.4 6.5 Sierra Leone L 160 4.7 30.7 170 6.5 37.5 25.7 1.5 Somalia L * 120 8.8 81.0 122 7.0 49.0 17.6 South Africa UM 3,210 40.6 41.1 48 2.8 65.2 7.4 6.7 Sudan L 290 27.7 35.1 71 4.6 55.0 11.6 Swaziland LM 1,520 1.0 28.9 65 4.5 57.4 10.3 9.3 Tanzania L 210 31.3 42.9 85 5.5 50.1 14.3 4.4 Togo L 340 4.3 43.3 86 6.1 50.1 15.0 3.9 Uganda L 330 20.3 50.6 99 6.6 42.5 19.7 4.5 Zambia L 370 9.4 26.8 113 5.6 43.4 18.9 9.1 Zimbabwe L 720 11.5 35.4 69 3.8 55.4 11.0 12.8 Source: World Development Indicators Central, except where indicated. 1. RDB (Total GNP per capita divided by population). 2/ 2010 figures from ILO. 3/ LDB. For 1996. 4/ UNAIDS. * For 1990 37 Appendix Table 2. Projected Dependency Ratios, Sub-Saharan Africa. Angola Benin Botswana Burkina Faso Burundi Cameroon Cape Verde Cen Afr Rep. Chad Comoros Congo, D. Congo, R. Côte d'Ivoire Djibouti Eq. Guinea Eritrea Ethiopia Gabon Gambia Ghana Guinea Guinea Biss. Kenya Lesotho Liberia Madagascar Malawi Mali Mauritania Mauritius Mozambique Namibia Niger Nigeria Rwanda Sao Tome Senegal Seychelles Sierra Leone Somalia South Africa Sudan Swaziland Tanzania Togo Uganda Zambia Zimbabwe 1995 2015 60+ / 60+ / 60+ / 60+ / total 15-59 Total 15-59 4.6 9.7 4.2 8.1 4.5 9.5 4.3 7.9 3.9 7.5 3.6 5.8 4.8 9.8 3.5 6.5 4.3 8.7 3.6 6.6 5.6 11.1 4.9 8.9 6.3 12.1 4.3 6.7 6.0 11.9 4.5 7.9 5.2 12.1 3.5 6.7 4.5 9.2 5.4 9.1 4.5 9.2 4.1 7.9 5.1 10.4 4.0 7.5 4.5 8.9 4.0 7.0 5.1 9.6 5.7 10.1 4.5 8.9 4.4 7.9 4.6 9.3 3.5 6.6 6.5 12.7 6.0 10.8 8.8 16.8 7.5 13.5 4.9 9.0 5.9 10.5 4.7 9.4 5.5 9.5 4.2 8.6 4.6 8.1 6.5 12.9 5.6 10.2 4.5 9.2 3.8 6.5 6.3 11.9 6.4 10.8 4.4 8.7 4.5 8.4 4.7 9.5 5.0 9.0 4.2 8.5 4.0 7.4 5.2 11.0 3.8 7.3 5.0 9.8 5.3 9.4 8.5 13.2 13.7 21.1 6.2 13.3 5.1 8.8 5.8 11.2 5.6 9.7 4.0 8.4 3.6 7.2 4.0 8.0 4.5 8.2 3.8 7.7 2.9 5.3 9.1 17.1 8.5 14.9 4.3 8.5 4.4 8.0 9.2 15.6 10.5 16.4 4.4 9.0 4.4 7.9 4.0 8.3 3.9 7.8 7.1 12.2 7.7 11.8 4.9 9.0 6.1 10.6 4.1 7.8 4.7 7.9 4.1 8.2 3.6 6.4 4.9 10.0 4.2 7.7 3.8 8.0 2.6 4.9 3.8 7.7 3.5 6.2 4.4 8.4 4.5 7.1 2040 2060 2080 2100 60+ / 60+ / 60+ / 60+ / 60+ / 60+ / 60+ / 60+ / total 15-59 total 15-59 total 15-59 total 15-59 5.9 9.5 11.3 17.6 19.3 32.1 24.7 44.1 7.0 10.9 12.6 19.8 19.2 32.0 22.7 39.6 7.4 11.4 13.7 21.8 18.4 30.5 21.6 37.2 5.1 8.0 9.9 15.3 16.7 27.1 20.7 35.3 6.0 9.4 10.5 16.3 16.7 27.2 20.3 34.6 7.8 12.1 14.0 22.2 21.5 36.8 24.4 43.2 15.2 24.5 22.8 39.5 27.0 49.7 28.1 52.6 7.1 10.9 12.6 19.8 18.3 30.5 20.9 35.8 5.0 7.8 11.1 17.3 18.7 31.0 22.8 39.6 10.6 16.2 18.9 31.3 24.2 43.0 25.7 46.4 5.9 9.6 10.8 16.7 17.8 29.1 23.0 40.3 6.4 10.1 11.7 18.3 19.0 31.7 23.1 40.3 7.3 11.2 13.3 21.0 18.8 31.5 21.4 36.8 9.2 14.5 13.7 21.7 19.8 33.4 22.5 39.0 6.8 10.5 10.1 15.6 16.6 22.7 39.5 5.1 8.1 12.7 19.8 19.6 27.0 32.9 20.6 35.0 8.3 13.1 10.1 15.6 16.6 27.0 23.2 40.5 9.2 14.1 14.7 23.4 21.1 36.0 23.2 40.6 8.0 12.4 13.6 21.5 20.0 33.7 23.1 40.2 10.2 15.7 17.1 27.9 23.2 40.7 25.0 44.7 7.4 11.4 12.6 19.9 18.1 30.0 20.7 35.3 6.8 10.6 11.7 18.3 17.1 28.2 19.9 33.8 8.8 13.5 15.4 24.7 20.7 35.4 23.0 40.2 10.7 16.5 17.5 28.6 22.6 39.4 24.4 43.2 6.6 10.4 11.1 17.2 17.8 29.3 21.6 37.1 8.4 12.9 14.9 23.8 21.7 37.2 24.1 42.6 6.2 9.8 10.9 16.9 16.8 27.4 20.3 34.4 5.7 9.1 8.9 13.6 17.3 28.2 22.4 38.9 8.6 13.3 14.3 22.7 20.4 34.6 23.1 40.3 24.9 44.2 27.5 50.9 28.0 52.2 28.5 53.4 8.1 12.5 5.5 10.6 18.9 31.6 21.3 36.5 9.1 14.0 15.6 25.0 21.8 37.5 24.0 42.4 5.0 8.3 9.6 14.7 17.5 28.6 24.4 43.3 8.1 12.5 13.8 21.7 20.8 35.3 23.6 41.4 6.0 9.3 9.7 14.9 16.4 26.8 19.7 33.3 11.0 17.0 19.4 32.5 25.4 45.8 26.8 49.0 7.1 10.9 12.6 19.7 19.6 32.9 22.6 39.3 21.4 36.8 25.8 47.1 29.1 56.1 28.7 54.4 6.6 10.4 11.0 17.3 15.7 25.6 18.6 31.2 5.9 9.9 11.0 17.0 20.0 33.4 27.0 49.7 15.2 24.5 21.4 36.9 25.0 44.7 27.0 49.4 10.6 16.6 16.1 25.9 22.4 38.9 24.3 43.2 10.6 16.6 17.8 29.3 23.4 41.0 25.6 46.2 6.5 9.9 12.1 18.9 18.5 30.7 21.7 37.3 6.3 9.9 11.5 17.9 18.3 30.2 22.2 38.3 4.8 7.4 10.0 15.5 16.6 27.2 20.1 34.1 6.8 10.5 12.3 19.4 17.7 29.3 20.3 34.4 10.3 15.9 17.0 28.0 20.4 34.6 23.2 40.5 Source: World Development Indicators Central, except where indicated. 1. RDB (Total GNP per capita divided by population). 2/ 2010 figures from ILO. 3/ LDB. 4/ UNAIDS. * For 1990 38 Appendix Table 3. Aging in Sub-Saharan Africa. Angola Benin Botswana Burkina Faso Burundi Cameroon Cape Verde Cen Afr. Rep. Chad Comoros Congo, D. Congo, R Côte d'Ivoire Eq. Guinea Eritrea Ethiopia Gabon Gambia, The Ghana Guinea Guinea-Biss. Kenya Lesotho Liberia Madagascar Malawi Mali Mauritania Mauritius Mozambique Namibia Niger Nigeria Rwanda Sao Tome Senegal Sierra Leone Somalia South Africa Sudan Swaziland Tanzania Togo Uganda Zambia Zimbabwe Population 60+ (thous.) Growth rate 1995 1995-2010 501 51.7 240 47.1 55 52.7 454 41.9 268 33.2 737 45.9 25 202 (4.0) 27.7 364 35.2 24 66.7 2,034 52.3 134 35.1 627 51.4 25 28.0 152 70.4 2,556 58.0 96 34.4 54 68.5 807 68.3 309 39.8 71 26.8 1,210 31.3 123 54.5 117 105.1 620 57.6 414 41.5 445 51.9 118 45.8 94 45.7 888 40.5 89 46.1 364 58.5 5,014 66.0 196 94.4 1/ 7,822 391 53.5 205 41.5 405 56.0 2,754 54.0 1,304 65.4 36 77.8 1,233 59.2 203 40.4 749 22.6 302 21.9 477 24.9 Source: ILO, except where indicated. 1. United Nations. 2. Palacios, 1996 (Data for 1990, not 1995). 39 % of 60+ that are Economically Active 1995 2010 61.1 58.8 58.8 55.5 47.3 44.0 63.9 60.1 57.8 58.5 53.9 51.4 36.0 29.2 76.2 72.9 50.8 45.3 62.5 62.5 54.6 53.0 65.7 61.9 55.0 53.2 44.0 43.8 63.8 62.2 59.2 58.2 53.1 50.4 68.5 64.8 77.6 76.3 59.2 56.9 57.7 55.6 64.6 61.2 50.4 47.4 53.0 50.4 67.1 65.0 80.9 76.6 57.8 55.6 52.5 48.8 11.7 9.5 82.1 77.6 34.8 32.3 62.4 60.3 55.0 52.4 58.7 58.3 50.1 43.4 58.3 24.4 52.5 30.6 77.5 56.2 69.6 49.3 65.6 47.5 40.7 56.2 20.9 50.6 25.0 74.1 54.4 67.5 43.8 60.9 Appendix Table 4. Pension Structure, Sub-Saharan Africa. Benin Burkina Faso Burundi Cameroon Cape Verde Cen Af Rep. Chad Congo, D. Congo, R Côte d'Ivoire Eq Guinea Ethiopia Gabon Ghana Guinea Kenya Liberia Madagascar Malawi Mali Mauritania Mauritius Mozambique Niger Nigeria Rwanda Sao Tome Senegal Seychelles Sudan Swaziland Tanzania Togo Uganda Zambia Year 1986 1986 1985 1986 1986 Pension Pension Spending Spending/ LCU, GDP *** Year mill. 7,040 0.40 4,800 0.30 1981 396 0.20 1981 17892 0.30 874 0.26 0.00 0.90 0.80 1986 1986 114 - 1986 1989 58 824 1987 1986 1986 1990 1986 1986 4021 615 1,043 73 1649 1989 496 0.90 0.70 0.10 1986 0.00 0.50 1989 0.42 0.40 0.20 2.80 0.04 0.10 1980 0.10 1988 0.26 1979 1989 1981 1982 1986 1990 1986 1989 4 1,263 - Reserves Payroll Tax % of recent avg. LCU Reserves/ for Pensions earnings paid as thou. GDP (%) (%) a pension*, 1991 9.0 60 3.6 9.0 40 1.3 8.5 50 11,814 7.0 45 1,190 7.0 5.0 45 6.0 48 6.0 6.5 50 4.0 26.0 10.0 45 0.6 4.0 3,182 11.5 6.0 51 19,764 4.5 0.43 0.21 1979 0.50 1980 0.80 0.10 1987 7.0 3.0 7,653 976 2,879 23,270 101 134 1,413 13,754 - 1.4 0.7 3.0 40 40 45 1.6 10.4 2.0 6.0 10.0 8.8 15.0 14.0 4.3 5.8 6.0 40 84-112.4** 50 5.8 Source: Palacios, 1996 and Chand et al, 1998 * After thirty years of covered employment. ** This is the replacement rate without adjustment for the commuted pensions gratuity (CPG). *** Source: "International Patterns of Pension Provision" Placios/Pallares 1998, ILO, EPF, Staff estimates, and CIV. 40 Appendix Table 5 Summary of Old-Age Pension Programs Special Main No. System for Program of Public Type Types Employees Benin Soc. Insurance 3 Yes Botswana Pub. Employees 1 Only Burkina Faso Soc. Insurance 1 Yes Burundi Soc. Insurance 2 Yes Cameroon Soc. Insurance 3 Yes Cape Verde Soc. Insurance 2 yes Cent. Af.Rep. Soc. Insurance 1 no Chad Soc. Insurance Congo, D. Soc. Insurance 3 yes Congo, R. Soc. Insurance 2 yes Côte d'Ivoire Soc. Insurance 2 yes Eq. Guinea Soc. Insurance 1 no Ethiopia Soc. Insurance 2 yes Gabon Soc. Insurance 3 yes Gambia Provident 2 yes Gambia Pension Scheme 1 Ghana Soc. Insurance 1 Army Guinea Soc. Insurance 1 No Kenya Provident 2 Yes Liberia Soc. Insurance 1 No Madagascar Soc. Insurance 2 Yes Malawi 1 Only Mali 2/ Soc. Insurance 2 Yes Mauritania Soc. Insurance 2 Yes Mauritius Soc. Insurance 2 Yes Mauritius Universal 1 Niger Soc. Insurance 2 Yes Nigeria Soc. Insurance 2 Yes Rwanda Soc. Insurance 1 No Sao Tome Soc. Insurance 1 No Senegal 2/ Soc. Insurance 1 Yes Senegal 2/ Soc. Insurance 1 Seychelles Soc. Sec. Fund 1 No Seychelles Pension Scheme 1 No South Africa Soc. Assistance 2 Yes Sudan Soc. Insurance 2 Yes Swaziland Provident 2 Yes Tanzania Provident * 1 No Togo Soc. Insurance 2 Yes Uganda Provident 2 Yes Zambia Provident 2 Yes Zimbabwe Soc. Insurance 1 No Source of Funds Insured Employer (% of (% of Payroll) Payroll) Total Gov't. 3.6 6.4 10.0 0 4.5 3.0 2.8 3.0 2.0 2.0 3.5 2.4 1.6 4.5 4.0 2.5 0.0 5.0 5.0 2.5 5.0 3.0 1.0 4.5 5.5 4.2 7.0 3.0 4.0 3.5 3.6 2.4 21.5 6.0 5.0 19.0 10.0 12.5 4.0 5.0 3.0 3.5 2.0 7.0 1.0 2.0 3.0 6.0 0.0 0.0 1.6 2.4 2.5 5.0 3.0 3.0 4.0 6.0 5.6 8.4 2.4 3.6 5.0 1/ Volunt. Rs50/mo. 0.0 0.0 8.0 17.0 5.0 5.0 10.0 10.0 2.4 3.6 5.0 10.0 5.0 5.0 3.0 3.0 Max Earnings For Contrib. & Benefit Purposes 9.0 0 200,000 francs 8.5 0 80,000 francs 7.0 0 300,000 francs 10.0 0 5.0 0 6.0 0 7.0 Subsidy 6.0 0 (23,500 francs) 4.0 0 1,647,315 francs 26.0 25% SS 10.0 0 7.5 0 1,500,000 francs 19.0 0 15.0 0 17.5 0 6.5 0 400,000 francs 10.0 0 1,600 shillings 6.0 0 4.5 0 100,000 francs 9.0 3.0 9.0 0.0 4.0 7.5 6.0 10.0 14.0 6.0 25.0 10.0 20.0 6.0 15.0 10.0 6.0 0 0 35,000 ougiyas 0 Rs55,500 All Rs55,500 0 250,000 francs 0 48,000 naira 0 0 0 600,000 CFA 0 600,000 CFA 0 0 All 0 0 400 emalangeni 0 0 0 0 K15,000 0 Z$4,000 Source: U.S. Social Security Administration, 1997 * Moving towards social insurance. 1/ 10% on first Rs1,000 of monthly wages, 20% on second Rs1,000, 35% on next Rs8,000, and 40% on wages in excess of Rs10,000 2/ Source: CIPRES 1999 3/ S is self employed, T is temporary workers, A is agricultural workers, and D is domestic workers. 4/ Parenthesis indicates a minimum instead of a maximum. 41 Appendix Table 6 --------- Qualifying Conditions --------Min. Length of Contribution RetireYrs. Of Mos. of Over How Ment Age Insurance Contrib. Many yrs? Necess.? 55 20 60 10 yes Benin Botswana Burkina Faso Burundi Cameroon Cape Verde Cent. Af. Rep. Chad Congo, D. Congo, R. Côte d'Ivoire Eq. Guinea Ethiopia Gabon Gambia Ghana Guinea Kenya Liberia Madagascar 55 55 60 65(60) 55 (50) 55 63 (60) 55 55 60 55 55 55 60 55 55 60 60 (55) 15 15 15 3 20 15 5 20 10 10 10 20 5 20 15 180 8.3 15 Malawi Mali Mauritania 55 60 (55) 10 20 yes no yes 36 60 60 60 60 10 10 10 5 yes yes yes yes yes 60 10 yes 100 28 quarters 10 Yes yes yes yes 60 10 yes yes 240 Summary of Old-Age Pension Programs ---------------------- Benefits -------------------Old-age grant for % increm. Minimu Maximum Non-qualif. Contrib. m % of Avg. Base For every Pension Pension Benefit: Min. Monthly Period 12 mos. Of (% of (% of avg. Automatic Mos. wages Mos. of Earnings (yrs) Insurance min Earnings) Adjustment per year Contrib. 30 3-5 2 wage) 60 80 Cost of liv. 1 12 * 20 3-5 1.33 60 80 Cost of liv. 2 30 3-5 2 60 80 1 30 3-5 1 50 80 1 20 5 1.50 4200 esc. 30 3-5 2/ 1 60 80 1 12 30 3-5 1.2 60 80 1 1.67 ** 50 Wage index 10 x annual pension 40 3-5 2 60 80 Cost of liv. 1 1.33 ** Yes 40 2 4/ 2 80 30 3 4/ 1 60 Yes 40 3-5 2/ 1 1 Yes 50 3 2/ 1.5 80 contribs. @ T-rate interest 2 ** Yes *** Yes 25 3/ 1 Yes 12 **** 30 min 1 60 40 Cost of Proportion. 100 quarters wage living Reduced * 1.67 5 60 cost of liv. 20 3-5 1.33 60 80 cost of liv. 1 43 Appendix Table 6 Cont. Summary of Old-Age Pension Programs --------- Qualifying Conditions --------- Mauritius Mauritius Niger Nigeria Rwanda ---------------------- Benefits -------------------Old-age grant for % increm. Minimum Maximu Non-qualif. Contrib. m Min. Length of Contribution Retire- % of Avg. Base for every Pension Pension Benefit: Min. Yrs. Of Mos. of Over how Ment Monthly Period 12 mos. of (% of (% of Automatic Mos. wages Mos. of avg. Adjustment per year Age Insurance Contrib. Many yrs? Necess Earnings (yrs) Insurance Min wage) Earning Contrib. s.? s)33.3 min decree 60 12 yrs of residence after age 0.83 Contrib. Gov. adjust. Rs177/ 18. 60 0 yes Rs1,055/mo 60 20 60 10 yes 20 3-5 1.3 60 80 1 60 10 120 yes 30 final 1.5 80 65 12 55 20 60 10 yes 30 3-5 1 50 Cost of liv. 50% of avg. monthly earnings during the last 3-5 yrs of contrib. Sao Tome Senegal Seychelles Seychelles South Africa Sudan Swaziland Tanzania Togo Uganda Zambia Zimbabwe 62 (57) 10 120 no 35 5 1 30 Wage index 55 1 12 yes 1.33 ** 63 5 yrs. residence preceding retirement. Rs1,100 cost of liv. Rs1,100/mo 63 5 yrs. residence preceding retirement. Based on worker’s voluntary contribs. to Pension Scheme. 65(60) Must be citizens R470 60 (55) 12 50 5,250 pnds 75 yes (see sheets) 50 Yes *** 55 5/ 15 Yes *** 55 20 60 10 Yes 20 3-5 1.33 60 80 cost of liv. 1 55 *** 50 *** 1/12 of annual insurable earnings 60 10 1-1/3% ** X no. yrs. of contrib. Up to 10 yrs. Source: U.S. Social Security Administration, 1997. Retirement ages in parentheses are for Women. * Special system for public employees only. ** Multiplied by the number of years of insurance coverage. *** Total employer and employee contributions, plus interest. **** Plus 20% of average earnings during last 10 years. 2/ For every 12 monthly contributions in excess of 240. 3/ For every 10 monthly contributions in excess of 100. 4/ For every year of contribution over 10 years. 5/ Source: Ejuba, 1999 43