• social security reform;
  • social security financing;
  • social security administration;
  • poverty;
  • developing countries;
  • southern Africa


  1. Top of page
  2. Abstract
  3. Introduction
  4. Supply-side and demand-side considerations
  5. The role of International Organizations (IOs)
  6. Social pensions in practice
  7. Learning about social pensions
  8. Conclusions
  9. Bibliography

In the last decade and particularly since the publication of the Millennium Development Goals, social pensions have captured the interest of those concerned with the well-being of older people across that large part of the world where formal, contributions-financed, old-age benefit systems cover only a minority of the population. International organizations have turned their attention to such schemes and some see them as having a valuable role to play. However, information about what they are and how they work, and about their efficacy in meeting the objectives set for them, is still limited. Learning has been taking place not only in the international organizations but also in the region where they are most prominent – southern Africa. Such learning should be encouraged and the International Social Security Association has a part to play in this learning process.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Supply-side and demand-side considerations
  5. The role of International Organizations (IOs)
  6. Social pensions in practice
  7. Learning about social pensions
  8. Conclusions
  9. Bibliography

Understanding why, at any given time, certain programmes dominate the policy agenda is important. Those who strive for such understanding are sometimes described as being “concerned with the relationship between power [actors] and ideas”. They seek to establish to what degree ideas have the power to drive change on their own, or whether they have power only “to the extent they are actively taken up by powerful individuals or groups” (McNeill, 2006, p. 354). However, a key element missing from many of the relevant studies is the political economy notion of context. From this perspective, the relative power of ideas and actors is not sufficient. If too much importance is accorded to the ideas, and to the role of the actors who promote them, rather than to the underlying cultural, social, economic, demographic and political environment in which these ideas are being discussed, there is a risk that the analysis and assessment of any policies that appear to have resulted from the ideas will be deficient. The reasons for their possible failure might well be missed.

This article looks at the issues of ideas, actors and context in the field of social pensions. The objective is to discern the influence of policy learning in the diffusion of ideas about social pensions and in the widely observed predisposition of policy actors to propose or introduce such programmes in developing countries. Here, social pensions are defined as tax-financed programmes granting periodic benefits, mainly but not exclusively, in cash to older people.1 Social pensions can be universal, in so far as they are payable to all who meet what is normally an age requirement, or they can be means-tested (see HelpAge International, 2004). The growing policy and academic interest in social pensions stems from the recognition of their potential as an adaptable policy tool for combating deep, widespread and persistent poverty.

According to those who advocate them, social pensions have the following virtues.2 They

  • raise the living standard of older people
  • provide health care for older people
  • help improve the education and health of (adult) dependants
  • support orphaned children (living with grandparents)
  • reduce gender inequalities (by giving women an income), and
  • replace support previously provided by adult children.

The last virtue is increasingly stressed. Commentators point out that, as families disperse and children move from rural to urban areas, the ties that have ensured that adult children support their parents have weakened. Even if they have not, adult children might have died – a consequence of the HIV and AIDS pandemic in many countries – or, more recently, have lost their jobs – a consequence of the current global downturn.3

It is to be noted that whilst older people are supposedly the principal beneficiaries of social pensions, there are also secondary beneficiaries. That this is so is used to bolster the argument that social pensions not only reduce poverty but also help foster economic development – in particular by improving the health and education of younger dependants. In this respect, at least some advocates of social pensions appear to take a “productivist” view of them (for example, Barrientos, 2009a). By this they share the view of those, mainly but not exclusively in Europe, who argue that “social protection is a productive factor” and, as such, should not be considered as a burden on the economies that provide it (see, for example, CEC, 1999).

Supply-side and demand-side considerations

  1. Top of page
  2. Abstract
  3. Introduction
  4. Supply-side and demand-side considerations
  5. The role of International Organizations (IOs)
  6. Social pensions in practice
  7. Learning about social pensions
  8. Conclusions
  9. Bibliography

Although much of the discussion of social pensions has dwelt upon the broader welfare-enhancing consequences for individuals and households of such programmes, they also have a political economy dimension that is often ignored. A study that investigated the “politics behind the non-contributory old age social pensions in Lesotho, Namibia and South Africa” made the important distinction between “demand-driven” and “supply-driven” innovation – between innovation that was made to satisfy the demands of populations for increased social security, and innovation that was made to serve political ends. The government of South Africa introduced an old-age grant for “white” and “coloured” South Africans in 1928 in order to emulate the provisions already made by many European countries, and as a colonizing power, it subsequently extended the scheme to Namibia. The introduction of a social pension in Lesotho in 2004 has been described as “a product of regional geopolitics and a concern for equity”: to provide ethnic Basotho living in Lesotho an equivalent to the pension received by Basotho families living in neighbouring South Africa (Pelham, 2007, p. 7).

That “politics matter” is not specific to Africa. An analysis of the political economy of the introduction of the Bonosol pension in Bolivia in the mid-1990s shows that reform “was not motivated by social policy considerations but by the desire to privatise state-owned enterprises and pensions” (Müller, 2008, p. 164).4 In efforts to build consensus in support of the planned, and duly implemented, reform in 1996, the Bonosol (from “Solidarity Bond”) programme was a home-grown idea appended to the proposed legislation “to complicate opposition to pension privatisation” (Müller, 2008, p. 165).5

Even where demand-side considerations are more relevant, it is by no means clear that the reasons for introducing a social pension were well thought through. Lesotho's Commissioner for Pensions stated that the social pension was not introduced as a welfare response to HIV and AIDS but it was in realization of the fact that the target group has now become a “generation of carers” (Thulo, 2007). The same applies in Swaziland, where the King explicitly attributed the country's adoption of the social pension to the challenges exacerbated by HIV and AIDS (RHVP, 2007). Whilst some argue that Botswana adopted a universal pension in 1996 in order to “facilitate family interdependence in the face of urbanization and industrialization” (Johnson and Williamson, 2006 p. 56), others also stress that the high incidence of HIV and AIDS was important to the subsequent introduction of in-kind benefits to the elderly carers of “AIDS orphans” (Fultz and Pieris, 1999, p. 23, fn. 52).

For all countries, and regardless of the motivation for its implementation, once a social pension programme is in place it often creates its own dynamic. Thus, although the implementation of the pension programmes in South Africa and Namibia were initially supply-driven, demand side factors became more important, particularly in the post-apartheid and post-independence period. This shift can be explained in relation to three mutually-reinforcing characteristics attributed to the receipt of pensions: it fosters social solidarity, not least at the household level and restores a sense of citizenship to the elderly; it builds up expectations amongst the population about what the state can do for the elderly; and it encourages governments to recognize both that providing social welfare is a moral duty and something that can enhance their legitimacy. Moreover, to the extent that they can be seen as effectively delivering benefits to current pensioners, governments can strengthen the credibility of the promises made to future pensioners. This factor – the process of having learnt to expect – may help explain the resilience of social pension programmes, in southern Africa and elsewhere (Pelham, 2007).6

The role of International Organizations (IOs)

  1. Top of page
  2. Abstract
  3. Introduction
  4. Supply-side and demand-side considerations
  5. The role of International Organizations (IOs)
  6. Social pensions in practice
  7. Learning about social pensions
  8. Conclusions
  9. Bibliography

The story of the international development and expansion of social security in the twentieth century can be presented as one of a relatively limited gathering of ideas and actors at the international level and as one of great diversity in context and institutional forms at the national level (Rys, 1974). For much of the twentieth century ideas about social security were largely, but not exclusively, defined and driven by the hegemonic influence of the International Labour Organization (ILO). The ILO was a key source of diffusion for a specific welfare regime (Strang and Chang, 1993, pp. 239-241). Although it did not invent social insurance, as a powerful international actor the ILO has actively promoted it as a model of social security over much of its 90-year history. In the post-war period, the influence of the International Labour Organization's “standards”, and especially of the Social Security (Minimum Standards) Convention 102 (1952), has been significant (Tamburi, 1969; Strang and Chang, 1993). The ILO's standards promote social insurance, to a very large degree, as synonymous with “social security”. Anecdotal evidence suggests that new-member and future accession and aspirant-member states of the European Union ratified Convention 102 not only because it provided the only comparative international benchmark upon which to base and further develop social security provision, but also because it was considered a powerful signalling mechanism to demonstrate these countries' willingness to adopt wider international norms.7

In the middle of the 1990s the social policy hegemony of the ILO was seriously shaken by the stellar-like arrival of the World Bank with its proposals about why retirement systems needed reform and what form pension systems should take (Charlton and McKinnon, 2001). The Bank's pro-market agenda led it to emphasize privatization and the individualization of risk and to give preference to pre-funding over pay-as-you-go. Its much-cited, three-pillar pension model was presented as a remedy to what it perceived to be the shortcomings of many public pension programmes (including low coverage rates, inadequate benefits, weak administration, and political risk in fund management). However, in the final, and much overlooked, chapter of the Bank's seminal volume, Averting the old age crisis, there is a recognition that fully pre-funded systems were unlikely to be suitable for countries without sufficiently developed financial markets or regulatory capacity, even if there is also the suggestion that reformers should concentrate upon trying to develop such an infrastructure (World Bank, 1994, pp. 255-292).8 Consequently, this meant that what has conventionally been referred to as “the Wold Bank model” was as inappropriate for most low-income developing countries as was the ILO's social insurance model.

In part in developing a response to the Bank, the International Labour Office had been turning its attention to alternatives to the social insurance model. In its volume Social security pensions (Gillion et al., 2000), it remarked that “[u]niversal schemes, whether based on residence or on means, are the most effective ways of ensuring high coverage, particularly of the poor”, although it conditioned this remark with a caveat that “such schemes are seen by most developing countries as beyond their resources and often as being of relatively low priority” (2000, p. 421). More important, the publication of the Millennium Development Goals (United Nations, 2005) placed poverty alleviation at the forefront of the international agenda and gave the Office, as the secretariat of a major United Nations specialized agency, the task of seeking ways for their realization.9 Indeed, in the last decade, the views of the Bank and the ILO have been converging – and in a more explicit way than earlier commentators (for example, Queisser, 2000) had had the opportunity to observe – and that convergence has been evident as much as anything else with respect to providing basic income for the elderly poor without adequate, or any, social security protection. In a 2005 publication Old age income support in the 21st century, the Bank also substantially rewrote its three-pillar system into one of five pillars whereby it proposed a “zero pillar” (Holzmann and Hinz, 2005) that would serve as a mechanism for poverty alleviation for those excluded from contributory approaches to social security.10 Equally, the ILO has paid substantial attention to the development of a “Global Social Floor” that could provide a minimum level of social protection not only to the elderly but also to the sick and disabled and to those caring for children (Cichon and Hagemejer, 2007; ILO, 2009a). As with the Bank's zero pillar, all items of that floor, including a basic pension, would be tax-financed.11 Often commented upon is the way in which the Bank's rethinking coincided with debate in Chile about the efficacy of its mono-pillar system of mandatory individual accounts that had been introduced in 1981. Following a process of national dialogue in 2006-2008, the government of that country complemented the system of individual accounts with a new, tax-financed Pensión Básica Solidaria (PBS or solidarity pension) that would help provide adequate pensions to those (particularly, but not only, women) with low earnings or broken work histories (see Kritzer, 2008; ISSA, 2010 forthcoming).

The International Social Security Association (ISSA) was a relatively late entrant to the debate on social pensions. This is not surprising. Its membership consists largely of public social insurance institutions. Its prime concern is ensuring that these institutions function efficiently, and with assisting member organizations to achieve this goal. Poverty alleviation has, thus, not been a core concern in the same way as it has been for both the Bank and the ILO. Nevertheless, in fulfilling its more technical mandate, it has become increasingly aware of the problem of inadequate coverage. As is widely recognized, only a fraction of the global population participates in any contributory pension system (World Bank, 2006, Ch.7; Cichon and Hagemejer, 2007). If the sort of social protection that the ISSA represents was not to remain a marginal phenomenon, or even one that would gradually wither away, a redirection of thinking was necessary.12 Some of the rationale for this redirection was provided by the concept of “Dynamic Social Security” (see McKinnon, 2007a) that emphasized the accessibility of social protection systems and their social inclusiveness. More concretely, the Association came to see tax-financed basic pensions as providing a potential mechanism for significantly and rapidly extending basic pension coverage in countries where coverage under contributory programmes was at best limited or at worst negligible. Recent ISSA Developments and Trends reports summarize this thinking (ISSA, 2007; 2008a; 2009; 2010 forthcoming).

Social pensions in practice

  1. Top of page
  2. Abstract
  3. Introduction
  4. Supply-side and demand-side considerations
  5. The role of International Organizations (IOs)
  6. Social pensions in practice
  7. Learning about social pensions
  8. Conclusions
  9. Bibliography

Information on the incidence of social pensions can be drawn from many sources. International non-governmental organizations (NGOs) – particularly Global Action on Ageing, HelpAge International (HAI) and the Chronic Poverty Research Centre – and international organizations – particularly the World Bank and the ILO – have built up databases. HAI claims that 72 countries across the world have a social pension of which 46 are low or middle income.13 However, the definition used is somewhat elastic. Not only are the public pension schemes of countries such as Australia and New Zealand included, so too are many social assistance programmes and social safety nets that older people can access. Other NGOs are more restrictive in what they count as a social pension. Global Action on Ageing lists eight African countries – Botswana, Ethiopia, Lesotho, Mauritius, Namibia, South Africa, Uganda and Zambia – as having relevant programmes, although at least one (that of Zambia) is an experimental programme limited to a few areas, and a similar experimental programme in Kenya is not mentioned.14 The World Bank, which developed a relatively comprehensive inventory (see Palacios and Sluchynsky, 2006), listed only 21 in “developing” countries (but this list includes middle-income countries such as Estonia and Russia) – less than half the number claimed by HelpAge International. Only five of these are in Africa – Botswana, Egypt, Mauritius, Namibia and South Africa.15

Table 1 seeks to summarize the schemes in southern Africa and also those in selected South Asian and South American countries. The examples from the latter two regions are included primarily because the literature on social pensions often refers to one or more of them. The basic parameters of the various schemes are given, including eligibility conditions, coverage and costs and how they are delivered. Nonetheless, and although a large number of sources have been consulted, the Table does not pretend to be perfect. Moreover, with respect to many of the parameters, even the most assiduous collectors of information have problems. In particular, as Barrientos admits, “[t]here is very little information on the costs of administering schemes and delivering benefits” (Barrientos, 2009b, p. 76).

Table 1. Selected social pension programmes
CountryProgrammeStart yearAge of entitlementMeans-tested or universalCoverage (% of pop'n of pension age)Cost (% of GDP)Non-health public social protection expenditure (% of GDP)Programme administrationDelivery mechanismOperating cost
South AfricaOld-age grant192865 (men) 60 (women)Means-testedapprox. 941.48.43 (2005)Department of Social DevelopmentCommercial ‘mobile’ ATMs; designated post offices; automated transfers to a bank accountNA
MauritiusBasic old-age pension195065Universal1001.75.3Ministry of Social Security, National Solidarity, and Senior Citizens Welfare and Reform InstitutionsAutomated transfers to a bank account2.5% of benefits paid out
NamibiaNational pension1990a60Universal93-1000.71.7Ministry of Health and Social ServicesCommercial ‘mobile’ ATMs; designated post offices; automated transfers to a bank account15% of benefits paid out
BotswanaOld-age pension199665Universalapprox. 960.4NASocial Benefits Division, Department of Social Services, (Ministry of Local Government)NA4.5% of benefits paid out
LesothoSocial pension200470All elderly, except those receiving a government pension93-962.4NAPension DepartmentPost officesNA
SwazilandOld-age grant200560Means-tested80NANASocial Welfare DepartmentPost officesNA
KenyaUniversal pensionProposed55Projected between 1.0 to 2.00.02 (2005)
South America
CountryProgrammeStart yearAge of entitlementMeans-tested or universalCoverage (% of pop'n of pension age)Cost (% of GDP)Non-health public social protection expenditure (% of GDP)Programme administrationDelivery mechanismOperating cost
BrazilSocial assistance (Prêvedencia Rural)199160 (men) 55 (women)Rural elderly poor (means-tested) not covered by a contributory social security programmeapprox. 35b0.7NAFederal Government of BrazilCommercial banks and post officesNA
BrazilSocial assistance (Beneficio de Prestaçao Continuada)199365 (men) 60 (women)Elderly poor (means-tested) living in households with per capita income below a quarter of the minimum wageapprox. 35b0.3NANational Institute of Social SecurityCommercial banks and post officesNA
ChileSolidarity pension (Pensión básica solidaria)2006-200865Means-testedAims to cover 60 per centNA5.37 (2007)Institute of Social WelfareLocal offices of the Institute of Social WelfareNA
BoliviaUniversal pension (Renta dignidad)200860Universal931.34.36 (2007)Universal Old-age Pension Fund/ Ministry of FinancePaid in cash or in-kind on a monthly, quarterly, semi-annual or annual basis; paid out by authorized financial entities, and fixed and mobile units of the Bolivian military.NA
South Asia
CountryProgrammeStart yearAge of entitlementMeans-tested or universalCoverage (% of pop'n of pension age)Cost (% of GDP)Non-health public social protection expenditure (% of GDP)Programme administrationDelivery mechanismOperating cost
  1. a The programme's earlier roots stem from 1949 during the era of South African rule.

  2. b Figure combines coverage under the Prêvedencia Rural and the Beneficio de Prestaçao Continuada programmes.

NepalOld-age allowance199470Universal520.230.71 (2007)Ministry of Local DevelopmentVillage development committeesNA
IndiaSocial assistance pension199565Means-tested200.01NANational Social Assistance ProgrammesVillage councils and municipalitiesNA
BangladeshOld age allowance (Boishka bhata)199857Means-tested120.030.27 (2007)Department of Social Services (Ministry of Social Welfare)Local branches of the government-run Sonali BankNA

Learning about social pensions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Supply-side and demand-side considerations
  5. The role of International Organizations (IOs)
  6. Social pensions in practice
  7. Learning about social pensions
  8. Conclusions
  9. Bibliography

Learning about social pensions is a relatively new discipline. Much of it has been a self-learning process on the part of international organizations that have begun to recognize the inappropriateness of the approaches they had traditionally advanced (the case of the ILO) or that were consistent with their broader approach to fostering growth and combating poverty (the case of the World Bank). International NGOs have played an important role in enhancing awareness of the extent of poverty in old age in the developing world and of advocating minimum incomes in old age as a “right”. However, whilst through their publications the international organizations and NGOs have been able to show what exists, many more fundamental questions of what works have barely begun to be answered. In particular, much less is known about whether social pensions are the most efficient way of helping to achieve the goals that have been set for them and, in this respect, whether they are affordable and whether the delivery structures that they require for their success are available in the areas where the need for them is greatest.

What began as a process of inventory building is gradually becoming one of analysing and evaluating. Nonetheless, evidence itself is limited. With respect to the costs of social pensions, most of what is known is not about actual costs but about what a social pension system would cost if it were to be established. Simulations have been undertaken for a number of countries. Some of these suggest that the costs are relatively low – less than one per cent of GDP or even only a fraction of this (for example Pal et al., 2005; Mizunoya et al. 2006; Behrendt and Hagemejer, 2009). Moreover, on the assumption that the countries running such programmes continued to grow richer, costs would fall over time, possibly quite quickly (for example, HelpAge International, 2009a). Others have placed costs somewhat higher – between two and three per cent of GDP (Kakwani and Subbarao, 2005). No single lesson can be learnt because different simulations involve different assumptions about the level of benefit to be paid and about which people would be eligible for them.

Far fewer attempts have been made to estimate the administrative costs that are associated with delivery. Experience from developed countries underlines how social assistance programmes involve far higher administrative costs than social insurance programmes. In these latter countries, the difference is due to the tighter eligibility rules that apply and that often involve examination of the means of those applying for them and of their willingness and/or availability to work. Were social pensions to be other than universal, there would be the requirement to establish equivalent, if by no means identical, structures. Even if they were universal, proving age is not unproblematic – registration of births was not general in many developing countries until relatively recently. ILO simulations set administrative costs as the equivalent of 15 per cent of benefits paid out when programmes are universal and 30 per cent when they are conditional (Behrendt and Hagemejer, 2009). Others, reporting on actual experience, have cited lower figures, sometimes much lower, as Table 1 testifies. What explains the differences is unclear.

Little has also been said about the costs associated with the establishment and maintenance of a delivery infrastructure and in many developing countries this does not exist – either for cash transfer or for other social programmes. It might well be unnecessary to build completely new structures and advocates of social pensions are keen to show how existing points of delivery for much more basic services can be utilized. Surprisingly, few have yet sought to investigate the role played by mobile phone technology, despite the fact that it is recognized as a means of intra-familiar cash transfer and remittance payments and as a way of linking crop producers to markets and improving the prices they can obtain.16

Experts have devoted attention to discussing who might be the recipient of a pension and whether it should be universal or targeted. Targeting is presented as a means of reducing benefit costs and ensuring that the resources that are available go to those who most need them. However, it is recognized that the societies in which most of the people for whom social pensions are conceived are those where access to in-kind resources are at least as important as access to cash resources and where definitions of household might well not coincide with those used by welfare agencies in developed countries. As a consequence, considerable effort has been devoted to reviewing the relative merits of various proxies for well-being (for example, see Samson et al., 2006). Concrete examples of decision-making and eligibility determination are offered by some of those cataloguing the procedures employed in countries where there are social pensions. Often these illustrate the role ascribed to local communities or local sources of authority in the decision-making process.17 There might be benefits to be achieved by adopting such a route. On the other hand, even fierce advocates of social pensions and transfers to other poor groups recognize that results can be arbitrary, and can give rise to favouritism or discrimination (HelpAge International, 2004). Aid donors have, in recent years, become increasingly concerned with “governance” and good governance is a prerequisite to the effective operation of a social pension system.

Little evidence is yet available on whether directing cash towards older people per se is the most effective means of reducing poverty or promoting growth. Much of the argument is by assertion. Advocates of social pensions point to trickle-down effects – particularly to the benefits to children where grandparents have become the providers. Of course, to the extent that older people themselves pass on a part of their pension to their family or household members, the extent to which it alleviates their own poverty is reduced. The question of whether it might not be more effective to target resources directly to children remains, at best, an open one. Indeed, simulations for two African countries carried out for the International Poverty Centre (IPC) of the United Nations Development Programme (UNDP) suggest that social pensions are not as efficient as some of their proponents claim.18 This is not to deny that advocates of social pensions make considerable play of the way they assist members of “skip-generation families” and make use of this role to mobilize government and donor finance. However, that social pensions can be shown to have secondary beneficiaries is not a sufficient reason to advocate them, any more than it is to justify the American space exploration programme because it led to the production of non-stick frying pans.19

Simulations can show the costs of social pensions. However, even if these appear relatively low, there are good reasons for not examining social pensions in isolation. One of the merits of the work recently undertaken in the ILO (Behrendt and Hagemejer, 2009) is that it looks not only at the costs of benefits for the elderly but also at the costs of benefits for children and for poor households and the costs of providing “essential” health care interventions for the whole population of a country. In other words, it attempts to cost the whole of the proposed Global Social Floor. When such an exercise is undertaken it becomes clear that costs rise substantially – to between five and 10 per cent of GDP for selected southern African countries. The last component of the package (health) is particularly expensive. Given the low levels of spending most developing countries currently make on social protection, this finding indicates the size of the challenge the ILO has set itself.20 Behrendt and Hagemejer pose the question of whether the resources required can be raised domestically. So, too, does a recent study originating in the World Bank (Jousten, 2009). However, whilst Behrendt and Hagemejer make clear that it is difficult to direct reform effort only to social pensions and that they have to be part of a wider social reform, the work undertaken in the Bank implies that yet wider reform is needed. Such reform would not only involve a revision of tax structures so that governments can call upon sufficient resources but also, and more important, the need for the economies of the countries concerned to grow to the extent that they can support the sort of social protection systems that campaigners are increasingly calling for.

Building the evidence base is not, however, solely the task of external experts or external organizations. At a regional level, cooperation is beginning. In 2006, member governments of some 13 southern African countries, met in Zambia under the auspices of the African Union (AU) to discuss poverty and called for the greater use of cash transfers including to vulnerable children, older persons and people with disabilities, and the “Livingstone Call for Action” that resulted made specific reference to “social pensions”. On top of this, however, the governments stressed the need for greater cooperation and the exchange of experience, both with each other and with outside countries, and for the institutionalization of a bi-annual conference on social protection to be organized by the Union. The call for action was repeated some months later this time at a meeting in Cameroon.21 Three sub-regional meetings of experts took place and a first AU Conference of Ministers of Social Development was held in Windhoek, Namibia, in late 2008. Pensions provision, however, did not feature strongly in the subsequent declaration – perhaps the result of higher priority being given to other issues, perhaps because the participants were much less homogenous than those at earlier meetings and had less common experience with cash transfer programmes for the elderly. A steering committee on ageing was however constituted.

The Southern African Development Community (SADC) has also started to involve its Member States in examining social protection. In early 2009, it held a meeting to investigate how its Parliamentary Forum and its member parliaments could promote the incorporation of social transfers in both national and regional development agendas. External as well as local experts made presentations of principles and of specific practices. Later in the year parliamentarians met again for a workshop on poverty and social transfers. Both meetings considered, inter alia, the position of the elderly and also called for increased oversight of the implementation and impact of social protection in general and social transfer programmes in particular.22

The Zambian meeting of 2006 was of interest in that, as well as involving representatives from Africa, it also had some from Brazil in an observer/expert status. The meeting appears to be one of the first of what has subsequently been referred to as ‘south-south cooperation’. At the Windhoek forum, the Brazilian Minister for Social Development gave an address and explained cash transfer programmes in his country. Subsequently, African experts were invited to Brazil. The latter country has now started similar exchanges with the People's Republic of China, concentrating initially upon poverty reduction and with the objective of “promoting knowledge sharing on best practices”.23

ISSA, itself, entered the learning process at a late stage. As part of a new approach to better understand and address key challenges facing ISSA member organizations in different regions of the world, the first Regional Social Security Forum for Africa was held in November 2008. One of the aims of the forum was to consider how the “[e]xtension of coverage of social security pensions [could serve] as a basis for old-age security in low-income countries” (ISSA 2008b). Although examples of social pensions were discussed in the course of the meeting, and were described in the accompanying report on Dynamic Social Security for Africa (ISSA, 2008a), the concluding press release did not use the term and instead referred to the role played by “universal pensions” in improving the life of older people (ISSA, 2008c). The forum provided a means of information exchange and transmitted findings of others, but it has not yet engaged in original research on social pensions in the same way as has the ILO or the International Poverty Centre. However, the fact that the 2010 World Social Security Forum will take place in Cape Town, South Africa, means that the Association will continue to be confronted by the issue and it is likely to face some pressure to play a more active role in the debate.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Supply-side and demand-side considerations
  5. The role of International Organizations (IOs)
  6. Social pensions in practice
  7. Learning about social pensions
  8. Conclusions
  9. Bibliography

Learning about social pensions is still at a relatively early stage. The issue of the well-being of older people in developing countries was little discussed until well into the start of the second millennium. It was given a boost by the propagation of the Millennium Development Goals, even if they did not make older people the subject of a goal in their own right. This highlighting of poverty gave campaigners a foothold in which to advance their cause and provided opportunities for a much more intense research programme to be undertaken in which academics from across the world participated. At the same time, principal international organizations began to rethink their approaches. They, too, went through a learning process. Neither conventional public social insurance pension systems nor more radical alternatives based upon fully pre-funded individual accounts seemed either adequate or relevant for a vast part of the world's population. Both the ILO and the World Bank have devoted increased resources both to finding out what is being done and how well it is being done and to considering what might or what should be done if the objectives they proclaim are to be fulfilled.

In this respect, it appears as if a community of experts concerned with social pensions is being built up. This community is made up of donors, aid organizations and academics. Almost all of them are from outside the developing world. The capacity of the developing world is, almost by definition, much more limited. The scope for learning is large and a growing willingness to participate in learning processes is apparent. Most interesting is the beginning of an interest in mutual learning within the developing world, which goes beyond the more usual approach where external participants come either from the developed world or from the major international organizations.

However, although there is much more experience being exchanged, there is still scope for a better understanding of simply what exists. This provides a foundation for analysis. However, there is also a need for deeper understanding of the contribution that social pensions can make. Only then will it be possible to judge if they are the right way forward for all the countries for which they are being proposed or in which they have been initiated. A fuller understanding is still lacking. The evidence base is currently weak, even if it is being continuously built up. Much of the material currently available has a ring of “advocacy” about it. To an extent, this is not surprising, if standards of living are seen as a “right”. Nonetheless, this means that evidence is collected to show the conditions under which older people are living and to make a plea for help, rather than to show that the means proposed for the alleviation of the perceived injustices would be effective, affordable and sustainable if they were to be introduced on the scale that was needed.

Moreover, it is also apparent that social pensions, where they do exist, have been introduced as much as an emotional response as a strategic response to the situations in the countries concerned. This is not to suggest that policy-makers in the countries concerned are more opportunistic than those elsewhere, or that social pensions are less effective than other policies directed to other situations that have been introduced under such circumstances. Nevertheless, where resources are severely constrained, the need to use them effectively is paramount. In this respect, there is a pressing need for evaluation. Countries can only learn from the experience of others if they are able to look at what has been achieved as opposed to at what has been aspired to. The development of local evaluating capacity is therefore central and this, as much as anything else, might be something that countries can learn together.

The fact that more evidence is needed about social pensions means that there is a contribution that international organizations and outside experts can make. This presents a challenge to the ISSA in particular. As has been made clear, the Association is currently not deeply engaged in the discussion about social pensions. It has viewed its remit quite narrowly and has left much of the discussion and most of the analytical work to other bodies. On the other hand, it has an interest in the extension of social security and it has also, implicitly if not explicitly, accepted that social pensions might contribute to this. Moreover, because it is not regionally specific, it can contribute to the south-south cooperation process as much as to any north-south process. There are lessons to be learnt about social pensions in South America that might indicate not only how, and how well, these function in their own right, which is relevant for some of the countries of Africa and of Asia and the Pacific, but also how they interact with existing contributory social security systems – be these embryonic, stable or decaying. Hereby, the Association can see both whether its approach is sustainable and how much it needs to reorient itself.

Alongside the wider political questions associated with social pensions, there are narrower, more technical questions that are relevant to the Association and where it has expertise that it can contribute. There are many unknowns relating to the delivery and administration that are common to pension systems, whether or not they are “social”. The ISSA has access to extensive experience and is well placed to enable “twinning” between member organizations and to offer practical assistance. The latter is something that many developing countries need. In assisting in the development of effective management of social pension systems, it can assist in enhancing their legitimacy and in ensuring that they do benefit the population for which they are intended. In being present, it can also ensure that the concept of social insurance and contributory approaches to social security more generally do not become forgotten, and it can show how basic pension systems might gradually be transformed. Moreover, by coordinating new social pension systems with existing contribution-based systems, it can ensure the survival, at least in a modified form, of the latter rather than see them being discredited as bastions of privilege for the few.

Accordingly, there are two good practical reasons for the ISSA to engage in the discussion of social pensions. It can maintain its role as a recognized authority on social protection rather than cede it to other bodies. It can also make a positive contribution to its objective of improving the level of social protection enjoyed by people worldwide.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Supply-side and demand-side considerations
  5. The role of International Organizations (IOs)
  6. Social pensions in practice
  7. Learning about social pensions
  8. Conclusions
  9. Bibliography
  • Adema, W. 2006. Social assistance policy, development and the provision of a decent level if income in selected OECD countries (OECD Social, employment and migration working paper, No. 38). Paris, Organisation for Economic Co-operation and Development – Department for Employment, Labour and Social Affairs.
  • African Union. (undated). AU policy framework and plan of action on ageing. Nairobi, HelpAge Regional African Development Centre.
  • Barrientos, A. 2009a. “Introducing basic social protection in low-income countries: Lessons from existing programmes”, in P.Townsend (ed.), Building decent societies: Rethinking the role of social security in development. Basingstoke, Palgrave.
  • Barrientos, A. 2009b. “Social pensions in low-income countries”, in R.Holzmann, D. A.Robalino and N.Takayama (eds.), Closing the coverage gap: The role of social pensions and other retirement income transfers. Washington, DC, World Bank.
  • Behrendt, C.; Hagamejer, C. 2009. “Can low-income countries afford social security?”, in P.Townsend (ed.). Building decent societies: Rethinking the role of social security in development. Basingstoke, Palgrave.
  • Casey, B. H. 2005. “Security, effectiveness and equity in social protection: Where is ISSA going, and where should it go?”, in ISSA, New and revised approaches to social protection in Europe. Geneva, International Social Security Association.
  • CEC. 1999. A concerted strategy for modernising social protection (Communication from the Commission, COM(1999) 347 final). Brussels, Commission of the European Communities.
  • Charlton, R.; McKinnon, R. 2001. Pensions in development. Aldershot, Ashgate.
  • Chronic Poverty Research Centre. 2008. Social assistance in developing countries database (version 4). Manchester. <> (accessed on 16.09.2009).
  • Cichon, M.; Hagemejer, K. 2007. “Changing the development policy paradigm: Investing in a social security floor for all”, in International Social Security Review, Vol. 60, No. 2/3.
  • Fultz, E.; Pieris, B. 1999. Social security schemes in Southern Africa: An overview and proposals for future development (ILO/SAMAT Discussion paper, No. 11). Harare, ILO/SAMAT.
  • Gillion C.; et al. (ed.). 2000. Social security pensions: Development and reform. Geneva, International Labour Office.
  • Hagemejer, K. 2008. “Can African countries afford basic social security? Can they afford not to have it?”, in Poverty in focus, No. 15. <> (accessed on 14.09.2009).
  • Handa, S.; Stewart, S. 2008. “The orphan targeting dilemma in Eastern and Southern Africa”, in Poverty in focus, No. 15. <> (accessed on 14.09.2009).
  • HelpAge International. 2004. Age and security: How social pensions can deliver effective aid to poor older people and their families. London. <> (accessed on 21.09.2009).
  • HelpAge International. 2005. MDGs must target poorest say older people (Supplement to Ageing and Development). London. <> (accessed on 21.09.2009).
  • HelpAge International. 2009a. Working for life: Making decent work and pensions a reality for older people. London. <> (accessed on 21.09.2009).
  • HelpAge International. 2009b. The universal social pension in Nepal: An assessment of its impact on older people in Tanahun district (HelpAge briefing). London. <> (accessed on 21.09.2009).
  • Holzmann, R.; Hinz, R. 2005. Old age income support in the 21st century: An international perspective on pension systems and reform. Washington, DC, World Bank.
  • ILO. 2009a. Social security for all: Investing social justice and economic development (Social security policy briefing, No. 7). Geneva, International Labour Office. <> (accessed on 30.07.2009).
  • ILO. 2009b. Social security expenditure database. Geneva, International Labour Office. <> (accessed on 16.09.2009).
  • ISSA. 2007. Developments and trends: Supporting dynamic social security. Geneva, International Social Security Association. <> (accessed on 04.08.2009).
  • ISSA. 2008a. Dynamic social security for Africa: An agenda for development (Developments and trends). Geneva, International Social Security Association. <> (accessed on 30.07.2009).
  • ISSA. 2008b. Regional Forum to highlight social security as key to African development (Press release). Geneva, International Social Security Association.
  • ISSA. 2008c. African leaders call for greater political commitment to social security (Press release). Geneva, International Social Security Association.
  • ISSA. 2009. Dynamic social security for Asia and the Pacific: Integrated responses for more equitable growth (Developments and trends). Geneva, International Social Security Association.
  • ISSA. 2010. Dynamic social security for the Americas: Social cohesion and institutional diversity (Developments and trends). Geneva, International Social Security Association (forthcoming).
  • Johnson, J. K. M.; Williamson, J. B. 2006. “Do universal non-contributory pensions make sense for rural areas in low-income countries?”, in International Social Security Review, Vol. 59, No. 4.
  • Jousten, A. 2009. “Financing social pensions”, in R.Holzmann, D. A.Robalino and N.Takayama (eds.), Closing the coverage gap: The role of social pensions and other retirement income transfers. Washington, DC, World Bank.
  • Kakwani, N.; Subbarao, K. 2005. Aging and poverty in Africa and the role of social pensions (Social protection discussion paper, No. 0521). Washington, DC, World Bank.
  • Kritzer, B. E. 2008. “Chile's next generation pension reform”, in Social Security Bulletin, Vol. 68, No. 2.
  • Labonne J.; Chase, R. S. 2009. The power of information: The impact of mobile phones on farmers' welfare in the Philippines (World Bank Policy research working paper, No. 4996). Washington, DC, World Bank.
  • McKinnon, R. 2007a. “Dynamic social security: A framework for directing change and extending coverage”, in International Social Security Review, Vol. 60, No. 2/3.
  • McKinnon, R. 2007b. “Tax-financed old-age pensions in lower-income countries”, in ISSA, Developments and trends: Supporting dynamic social security. Geneva, International Social Security Association.
  • McNeill, D. 2006. “The diffusion of ideas in development theory and policy”, in Global Social Policy, Vol. 6, No. 3.
  • Mizunoya, S.; et al. 2006. Costing of basic social protection benefits for selected Asian countries: First results of a modelling exercise (Issues in social protection discussion paper, No. 17). Geneva, International Labour Office.
  • Müller, K. 2008. “Contested universalism: From Bonosol to Renta Dignidad in Bolivia”, in International Journal of Social Welfare, Vol. 18, No. 2.
  • Pal, K.; et al. 2005. Can low income countries afford basic social protection? First results from a modelling exercise (Issues in social protection discussion paper, No. 13). Geneva, International Labour Office.
  • Palacios, R.; Sluchynsky, O. 2006. Social pensions. Part I: Their role in the overall pensions system (Social protection discussion paper, No. 601). Washington, DC, World Bank.
  • Pelham, L. 2007. The politics behind the non-contributory old age social pensions in Lesotho, Namibia and South Africa (CPRC Working paper, No. 83). Manchester, Chronic Poverty Research Centre.
  • Queisser, M. 2000. “Pension reform and international organizations: From conflict to convergence”, in International Social Security Review, Vol. 53, No. 2.
  • RHVP. 2007. Old age and public assistance grants, Swaziland (REBA Case study brief, No. 6). Johannesburg, Regional Hunger and Vulnerability Programme. <> (accessed on 03.08.2009).
  • RHVP. 2009. Ever upwardly mobile: How do cellphones benefit vulnerable people? (Wahenga Brief, No. 16). Johannesburg, Regional Hunger and Vulnerability Programme. <> (accessed on 12.09.2009).
  • Rys, V. 1974. “Problems of social security planning in industrialised and developing countries”, in International Social Security Review, Vol. 27, No. 2/3.
  • Samson, M.; et al. 2006. Designing and implementing social transfer programmes (EPRI Research Paper, No. 38). Cape Town, Economic Policy Research Institute. <> (accessed on 14.09.2009).
  • Samson, M. 2009. “The impact of social transfers on growth, development, poverty and inequality in developing countries”, in P.Townsend (ed.), Building decent societies: Rethinking the role of social security in development. Basingstoke, Palgrave.
  • Samson, M.; Kaniki, S. 2008. “Social pensions as developmental social security for Africa”, in Poverty in focus, No. 15. <> (accessed on 14.09.2009).
  • Schwarzer H.; Querino, A. C. 2002. Non-contributory pensions in Brazil: The impact on poverty reduction (ESS Paper, No. 11). Geneva, International Labour Office – Social Security Policy and Development Branch.
  • Sluchynsky, O. 2009. “Administration of social pension programmes”, in R.Holzmann, D. A.Robalino and N.Takayama (eds.), Closing the coverage gap: The role of social pensions and other retirement income transfers. Washington, DC, World Bank.
  • SSA; ISSA. (various years). Social security programs throughout the world. Washington, DC, Social Security Administration.
  • Standard Bank. 2007. Standard Bank proposal on the payment of pension grants to the elderly. Johannesburg. <> (accessed on 14.09.2009).
  • Strang, D.; Chang, P. M. Y. 1993. “The International Labour Organisation and the welfare state: Institutional effects on national welfare spending”, in International Organization, Vol. 47, No. 2.
  • Tamburi, G. 1969. “The ILO and social security: The challenge of technical co-operation”, in International Social Security Review, Vol. 22, No. 4.
  • Thulo, T. 2007. Old age grant: Alternative delivery mechanisms (Conference paper, “Delivering Swaziland's Old Age Grant: A Conference on Alternative Delivery Options”, Johannesburg, 2 May). Johannesburg, Regional Hunger and Vulnerability Programme. <> (accessed on 30.07.2009).
  • United Nations. 2005. Millennium Development Goals. New York, NY. <> (accessed on 30.07.2009).
  • United Nations Population Division. 2008. World population prospects: The 2008 revision population database. <> (accessed on 15.09.2009).
  • Vincent, K. 2007. Delivery system issues in Swaziland (Conference paper, “Delivering Swaziland's Old Age Grant: A Conference on Alternative Delivery Options”, Johannesburg, 2 May). Johannesburg, Regional Hunger and Vulnerability Programme. <> (accessed on 14.09.2009).
  • Willmore, L. 2006a. “Universal age pensions in developing countries: The example of Mauritius”, in International Social Security Review, Vol. 59, No. 4.
  • Willmore, L. 2006b. Universal pensions for developing countries. Laxenburg, IIASA. <> (accessed on 14.09.2009).
  • World Bank. 1994. Averting the old age crisis: Policies to protect the poor and promote growth. Washington, DC.
  • World Bank. 2006. World Development Report 2006: Equity and development. Washington, DC.
  • 1

    The term “tax-financed programme” is employed in preference to “non-contributory programme”. In social security discourse, the latter term tends to be viewed as depreciating as well as being technically incorrect. Social pensions are often conflated with “social assistance”, although the latter are directed at a far wider group and, especially if recipients are of working age and are deemed as capable of working, are accompanied by obligations to seek work, to engage in work preparation, or even to engage in some form of community service. Moreover, social assistance is always means-tested (see Adema, 2006).

  • 2

    The list is a summary taken from the website of Global Action on Ageing <> (accessed on 12.09.2009).

  • 3

    Examples include HelpAge International (2005). The point was also made repeatedly by contributors to the 5th World Ageing and Generations Conference, St. Gallen, Switzerland, 3-4 September, 2009.

  • 4

    The Bonosol programme was replaced in 2008 by the Renta Dignidad programme.

  • 5

    The programme acquired this name because it was financed by proceeds of the privatization of state-owned enterprises.

  • 6

    Mauritius provides a neat case study of this shift from supply to demand. “Mauritius ended up with a system of universal age pensions by accident, not by design”. The non-contributory pension introduced in 1950 was “regarded as temporary, something to take care of the needs of the elderly population until the day a contributory, income-related system of pensions could be put in place”. Sixty years later, it has proved to be “both popular and durable” (Willmore, 2006a, p. 87). Mauritius is also a good example to highlight the idea of the entrenchment of the right to pensions. In 2004, the Government of Mauritius implemented means-testing to its previously universal programme and, subsequently, lost the 2005 election. Needless to say, the new government immediately restored universalism.

  • 7

    Countries that have ratified ILO Convention 102 in the last two decades are Croatia (1991), Cyprus (1991), FYR of Macedonia (1991), Bosnia and Herzegovina (1993), Czech Republic (1993), Serbia (2000), Poland (2003), Albania (2006), Montenegro (2006), Bulgaria (2008), and Brazil (2009).

  • 8

    The chapter did go on to suggest that countries that not only were young but also had limited existing pension systems, and so faced few transition costs, would find the establishment of a system based upon funding easier to contemplate than older countries with mature pay-as-you-go schemes.

  • 9

    Of course, the views attributed to the International Labour Office in Geneva do not always represent the position of the International Labour Organization. The policy stance is dictated by decisions taken by the International Labour Conference that are codified in Conventions or Recommendations. Developing new Conventions and Recommendations, or updating existing ones, can be lengthy and difficult, and it is likely there will be a considerable time lag before the concept of the “floor” achieves any formal status.

  • 10

    The “fifth pillar” in the revised Bank model in 2005 was defined as sources of informal support for the elderly.

  • 11

    The idea of the Global Social Floor can be traced back to ILO recommendations dating from as early as 1944 that requested universal basic income security and universal access to health care. However, these Recommendations did not achieve Convention status.

  • 12

    The role played by private and voluntary pension schemes within retirement benefit systems has grown relative to that of public, mandatory social security schemes, and might be argued to pose a threat to the authority of the ISSA (Casey, 2005). Calls for a greater role to be played by tax-financed social pensions might be seen to question its authority yet further. Effectively, it is being subjected to a pincer-like attack – on one side from second and third pillar systems and on the other from zero pillar systems.

  • 13
  • 14
  • 15.See 

    Palacios and Sluchynsky, 2006, Table A1. The table does claim to be “selective”. Moreover, it was compiled as some countries were first initiating their programmes; for example, Lesotho and Swaziland.

  • 16

    On the role of mobile phone technology for cash transfer purposes, see Vincent (2007) and Standard Bank (2007). With respect to its use by crop producers, see RHVP (2009) and Labonne and Chase (2009).

  • 17

    For example, in Kenya, those who are considered extremely poor and so are potentially eligible for cash benefits are chosen by their local communities. Those who live in a stone house or who grow coffee are, however, by definition, not extremely poor (communication from the Minister for Gender, Children and Social Development at the World Demographic Association conference, St. Gallen, Switzerland, on 4 September, 2009). In Bangladesh, the benefits are paid to the 20 elderly people identified as the “most poor” with in the ward in which they live (Barrientos, 2009a, b).

  • 18

    IPC simulations for Malawi and Uganda show that, for a given spend, programmes targeted at household with children reach a higher proportion of poor children than do those targeted at households that are poor or households with older people. Moreover, programmes targeted at orphaned children are the most efficient (Handa and Stewart, 2008).

  • 19

    The claim is, in fact, erroneous, although it is also pervasive. The technology dates from the late 1930s and was discovered, accidentally, in an attempt to improve fridge-cooling. The material produced did, however, prove useful in the “Manhattan project” to create an atomic bomb.

  • 20

    There are only a few southern African countries that spend more than two per cent of GDP on non-health social protection and most spend under one per cent.

  • 21

    The Livingstone and Yaoundé meetings were not the first time African countries had addressed either the issue of ageing populations or of social protection. In 2002, the nascent African Union had issued its own Action Plan on Ageing that called upon governments to “[p]rovide and enhance access to social assistance schemes for older people including public assistance schemes, old age pensions etc” (African Union, undated, p. 17). Moreover, the ILO reports inter-governmental events in French-speaking West Africa – in Dakar, Senegal, in 1994, and in Abidjan, Côte d'Ivoire, in 1996, where the message for the development of improved social protection appears to have been one of the rehabilitation and restoration of the credibility of existing social security systems (Gillion et al., 2000, p. 526). At Ouagadougou, in Burkina Faso, in 2006, heads of state discussing the challenges of employment and poverty had called for improvements in the living condition of the aged – through better social protection services including improved pensions, health and other social security schemes – but had made no specific mention of social pensions.

  • 22

    Details of these meetings are obtainable at <>.

  • 23

    Details of this initiative can be seen on the website of the IPC <>.