Draft 03

1 February 2004

Comments welcome

 

 

 

 

 

 

 

                                    Reaching out to 100 million poor?

 

Designing resource mobilization strategies

to combat old-age poverty through universal pensions

 

 

 

                                    A concept note[1]

 

 

 

 

 

 

 

 

 

 

 

Michael Cichon

Financial, Actuarial and Statistical Services Branch

Social Protection Sector

International Labour Office, Geneva

January 2004

 

 

 

 

 

 

 


1.                  Introduction: The problematique

 

Today about 1.2 billion people are living on an income of less than $1 per day. According to ILO estimates about 100 million of these extremely poor people are older than 60.  Being old and poor means that people have little hope and few means to improve their lot. Poverty and destitution tend to become permanent. The risk management strategies that are open to younger members of a society (holding multiple jobs, subsistence farming, selling of assets, acquiring new skills in an effort to improve employability) are generally no longer available to the elderly.  In many developing countries, notably in Africa, the elderly – due to the high prevalence of HIV/AIDS in the active generation - also often have to take charge of caring for sick family members or orphaned descendants and bear the educational responsibility which is normally borne by the active generation. They might also have to use whatever little earnings capacity that they have left to generate some income in kind and in cash for their family, neglecting inevitably some of the caring needs in the household.  In addition they might have to send children to work rather than to school.

 

There are studies from South Africa and Namibia which demonstrate that the payment of a small old-age pension to the elderly on a universal or means-tested and tax-financed basis, not only improves the life of the elderly and helps to reduce old-age poverty, but also has beneficial effects for whole family – notably for families in rural settings or the informal economy.[2]   They probably raise school attendance rates[3] and there is evidence that average weight, notably for girls, is highly correlated to the receipt of basic pensions by the grandparents.[4]  The elderly are thus used as “agents” for social assistance in their families, as they are often the ones deciding on spending priorities in the households if they are the only persons that are bringing in cash income.

 

A study on South Africa’s social pension undertaken by Case and Deaton uses a US$1 poverty line to examine the impact of the pension programme on poverty. Using data for 1993, the authors estimate that the poverty headcount would have been 5 percentage points higher (at 40 per cent) if “the pension incomes were removed” (Case and Deaton 1998, p.1342). It thus appears that tax-financed basic pensions could be a powerful tool against poverty.  The 100 million poor elderly live in families with about 400 to 500 million members who would indirectly benefit from the payment of a cash benefit to the elderly.

 

So far, basic tax-financed pensions have been deemed unaffordable in most developing countries.  Hence, social insurance pension schemes covering initially only part of the workforce in the formal economy were advocated in many developing countries. It was expected that, as the formal economy grew, the coverage of these pension schemes would also automatically increase.  So far, it has turned out that the process of extension of coverage was much slower than expected and that, in addition, many of these schemes are inefficient.  They spend, for example, far too big a proportion of their contribution income on administration. The latter is the consequence of an – inevitably - complicated social insurance administration, as well as often simply bad governance in the absence of proper administrative monitoring by the countries’ executive or legislative organs.  While these schemes still may help to prevent some proportion of potential future old-age poverty, the process is lengthy and has almost no impact on immediate poverty levels.  Recent indicative calculations by the ILO and others[5] seem to indicate that modest universal basic pensions could probably be financed in at least some African countries by domestic resources.[6]  Others would probably require transitional foreign financing.

 

International awareness-raising and national promotion campaigns are necessary to convince governments and key stakeholders in the societies that universal pensions are a key tool in combating poverty and should enjoy high priority when it comes to implementing social protection mechanism.  Such campaigns need credible bases in the form of sound national feasibility studies.  These studies need to ascertain:

 

a)                  what would be the likely effect of such a pension on the standard of living and poverty levels of the families/households in which the elderly live (establishing the social effectiveness of universal pensions);

 

b)                  what the likely cost of the benefit would be over at least five decades (establishing benefit  cost);

 

c)                  that there is no other social transfer mechanism, or other use of public resources, that would be more “welfare enhancing” or “poverty reducing” - if furnished with the same resources (establishing the social efficiency of universal pensions); and

 

d)                  whether a sound resource mobilization strategy could be devised that would enable the country to meet the cost of the benefit in the foreseeable future (establishing financial and fiscal feasibility), and if so, in what time frame.

 

            Once positive social effects and financial and fiscal affordability is established, then the implementation of universal pensions – if necessary gradually – would be difficult to challenge politically.

 

            This paper takes the positive social effects on households for granted and focuses on the techniques that can be used to establish the affordability and devise credible resource mobilization strategies to finance universal pensions and hence to demonstrate that such schemes are affordable.  These are the factual bases for national and international promotion campaigns of universal pension projects.  The paper ends with a proposal on how an international promotion campaign can be backed up with research results.

 

 

2.         The size and nature of the financial challenge

 

            Relatively little is known about the typical overall national (net and additional) resource needs of universal pension schemes with modest benefit levels.  The following table summarizes the existing evidence in six countries with existing non-contributory pensions (including universal and means-tested schemes).  The costs of these schemes are mirrored against crude ILO hypothetical cost estimates for four further countries.  The data and estimates are not completely compatible and further research is required to confirm the estimates and the statistical findings. However, it can – tentatively - be concluded that the cost of a basic universal pension scheme does not seem to be extravagant.  Depending on the concrete country case, one can delineate from the table that a basic universal pension scheme with a benefit level of about 20% of per capita GDP could probably cost between 1 and 2% of GDP and would require probably between 5 and 10% of overall government revenues.

 


Table 1:   Summary of observed and estimated resource needs of non-contributory  pensions
in selected developing countries, 2000 and later

 


            Hence, any national and international resource mobilization strategy to combat old-age poverty in developing countries would ultimately face the challenge to mobilize resources in the order of around 1 to 2 % of GDP, which would be the equivalent of 20 % of the current overall national social expenditure levels.  This may sound reasonable but it requires budgetary shifts in the order of up to 10% of government budgets. Such shifts are never easy and may have to be phased in over years.  However, as the example of Nepal shows, one can introduce universal pension schemes gradually by initially only covering subgroups of the elderly (e.g. people of age 75 and above) as well as survivors and possibly disabled people.

 

            Regardless of whether such a transfer system is introduced at once or gradually over time, a national resource mobilization strategy has to be developed and agreed upon by the government and society at large. A credible resource mobilization strategy has to consist of two major elements:

 

a)                  a study that establishes the nature and the feasibility of a universal pension scheme, containing:

 

-                 a sound estimate of the overall cost of a modest universal pension scheme;

 

-                 a  proposal on which types of revenue items in the government budget (i.e. direct taxation, indirect taxation, other charges and contributions, borrowing, foreign aid) or which budgetary shifts (i.e. freeing resources from other uses) can be used to finance the new benefit;

 

-                 estimates as to what amount of resources can be mobilized by each item; and

 

-                 within which time frame; and

 

b)                  a strategy to put the issue of universal pension on the national social policy agenda and  promote the use of national resources for a universal pension scheme.

 

The following sections explore these points in more detail.

 

 

3.                  Tools and steps of feasibility studies

 

            We assume now that the social situation in a country has been analysed in full detail and that universal pensions have – in principle - been identified as one potential measure that can help to close some of the social protection and poverty gaps in the country, i.e. the social effectiveness of the instrument has been established.

 

            The analysis of the feasibility of a universal pension transfer scheme should then proceed in five logical steps:

 

(1)                Undertaking the initial system diagnosis through social protection expenditure and performance reviews

 

            This step is a stock-taking exercise which establishes the performance of the existing social transfer system with respect to established social needs, inter alia, the alleviation or reduction of old-age poverty. This should be done in the form of a Social Protection Expenditure and Performance Review (SPER) which compiles all social transfer schemes in a country and analyses their cost and their performance in relation to their effectiveness (for example in population coverage and the degree of benefit adequacy) and efficiency.[7]  The objective is to identify all existing transfers that aim at the alleviation of poverty and the creation of social security in order to identify protection gaps as well as potential efficiency gains in the existing system and to avoid incompatibilities and duplication within the system that could occur through the creation of a new transfer system.  The performance analysis is based on a set of explicit indicators which help to measure the performance of individual transfer schemes or benchmark their performance against the experience in other countries.  The ILO has developed a methodological concept which has been tested in several countries, including Benin, Chile, Poland, South Africa, Thailand and Uruguay.[8]

 

(2)        Building a social budget as a financial governance tool

 

            This step sets up a social budget as a national budgeting tool that allows the assessment of the feasibility of a new social transfer scheme using a mathematical budget model.   A social budget model can, inter alia, simulate and project the cost and effects of new or modified social transfers without undertaking costly real-life trial and error experiments.  It can also be used for the analysis of diverse financing alternatives for specific social protection benefit schemes. One of the key result variables of the model is the effect of specific transfers on the government budget.  A social budget builds on the data base that has been assembled by the SPER.

 

            A social budget is essentially a family of simulation and projection models that map:

 

-             the country’s economy;

-             the demographic structure;

-             the labour market;

-             all existing social transfer systems (e.g. pensions, health care systems, short-term cash benefit schemes and social assistance schemes); as well as

-             the government budget.

 

            The concept of social budgeting is developed in detail in Scholz et al. (2000). For the feasibility of national pension schemes social budget models usually contain elaborate actuarial pension protection models (with long-range projection periods) as one module.  The ILO has developed or supported the development of national social budget models since the mid-1990s for Luxembourg, Panama, Poland, Slovakia, Turkey and the Ukraine.  It just started further initiatives to apply the methodology in Ghana and Tanzania. Pension models set in a limited social budget model context have been developed for and applied in over 30 countries. The methodologies are sufficiently generic to be applied to all countries after appropriate adaptation.

 

(3)               Using the tool to project cost and identify financing options

 

            This step aims at establishing the principal financial and fiscal feasibility of a new universal pension scheme. Once the social budget model has been compiled and tested in status quo runs (i.e. modelling the existing social transfer schemes) it can be used to project the cost of universal pension schemes at different benefit levels. Alternative financing methods (direct taxation, indirect taxation, other charges and contributions, foreign aid, savings in other old-age benefit systems etc.) can also be tested.

 

 

 

 

(4)                   Using the tool to establish social efficiency through a cost-benefit analysis

 

            This step tries to ascertain that a universal pension scheme is – from a social point of view – the most efficient use of scarce public resources. Before political negotiations are undertaken to establish the exact nature of the universal pension scheme with respect to benefit levels and entitlement conditions, the social budgeting technique can be used to establish whether an alternative use of the resources needed for the universal pension scheme could achieve a higher “return” in terms of poverty alleviation (that could be measured for example, by the reduction of the overall poverty headcount index or the closure of the aggregate national poverty gap).[9]  Once that is established in principle the final stage of the planning process can commence.

 

(5)        Using the tool to fine-tune the features of the scheme

 

            This step is a trial and error process between planners an political stakeholders aiming to create a national consensus on the detailed features of the universal pension proposal.  In a dialogue-based process within government and between government and other stakeholders the design can be finalized.  The calculation of multiple variants will be needed during this stage. If the system cannot be introduced completely from the beginning a longer-term transition period during which the new benefit is introduced can be designed.  One could envisage, for example, starting out with the very old or disabled and then developing a scenario where an increasing share of national revenues could be devoted to the new transfer system over a period of – say - 10 years.

 

            Once a feasible design for the national universal pension proposal has been established, national promotion strategies can be established.

 

 

4.                  The promotion strategy

 

            Should the simulation and analytical results show that a universal pension scheme is socially efficient and can be financed with national resources, promotion strategies can be developed.

 

            A critical phase in the national resource mobilization process is when the facts on the expected cost of the scheme have been established and appear manageable. Concrete budgetary shifts and/or the tapping of additional sources of revenue (for example, through increased VAT) require - in the first place - concrete political will. Political will needs to be generated. It is a prerequisite that the scheme can be portrayed as a net gain for society as a whole.  The transfers will largely benefit the informal sector and hence they may be perceived as crowding out benefits for the formal sector. It appears crucial for soliciting public acceptance that existing social insurance schemes for the formal sector are not threatened but are rather assured of their continued existence as a second-tier scheme for those with regular wage incomes. A key element in national resource mobilization strategies will be to promote the inclusion of universal pensions in the national PRSP processes.

 

            In the process of generating the political will a coalition of NGOs, the ILO, UNDP and other agencies and interest groups have a crucial role to play.  Recent ILO experience in the context of the Global Social Trust Initiative shows that the existence of a “national agent of change” is necessary. Without an institution, an agency or an NGO that is leading the process, no awareness and promotion campaign can be sustained.[10]

 

            If national financing is not sufficient international aid may be sought for a transitional period.  The following section explores some of the options.

 

 

5.         Mobilizing international resources

 

            The ILO has estimated that only about 2.0 per cent of the global Gross National Income (GNI) would be needed to furnish all of the world’s poor with a minimum of income security, access to basic educational services and access to basic health care.[11]  If one were to promote universal pensions as a partial solution, then the cost could be reduced to about one-tenth of that order of magnitude for the group of the aged poor.

 

            If one were to avoid a complex means-testing procedure and would provide a $1 per day pension to all the elderly over 60 in the less developed world, about 0.5% of the global GNI of around US$31,500 billion would be required.  Limiting the pensions to persons over 65 would probably reduce the cost to about 0.4% of the global GDP.

 

            Again, the order of magnitude appears small but given the fact that the developed world is presently not meeting the target of 0.7% of GDP for all international aid, straightforward international financing  will remain a dream for decades to come.

 

            However, international aid could contribute to the transitory or the seed financing of the new transfer system.  A reduction of the interest payments in Nepal by only 16% in the process of a debt relief process could double the amount of resources available for the existing modest universal pension programme in that country.[12]  In other countries the order of magnitude would probably be higher.  In any case the absolute cost per country appears rather small in an international context and the case can be made for carefully designed international transfers to support the introduction of universal pension schemes in a limited number of countries for a limited period.

 

            The ILO is presently testing another global resource mechanism for the financing of investments into the build-up of national social transfers, known under the working title the Global Social Trust Initiative. This initiative is not exclusive designed to sponsor universal pension schemes but can be used to support national resource mobilization strategies for universal pensions.

 

            The Global Social Trust is an innovative concept in which individual contributors in donor countries are brought together in partnership with recipient countries to extend sustainable social security provision.  In November 2002, the Governing Body of the ILO discussed the feasibility of the Global Social Trust[13] and authorized the Director-General to establish a Global Social Trust pilot project, the funding for which should come from extra-budgetary resources.[14] The Governing Body accepted this recommendation.

 

            Work commenced immediately. The first step was to identify two countries which could become partners for the pilot; countries of a size which was sufficient to test the concept, whilst keeping the project manageable. A round of tripartite negotiations following the November 2002 Governing Body led to the identification of Luxembourg and Namibia as donor and recipient countries respectively.

 

            The priority need identified by the Government and social partners of Namibia was a family income maintenance benefit for survivors, most of whom are affected by the AIDS pandemic, although the benefit will be available for all survivors and thus constitute a systemic extension to the existing state grant benefit structure. The project will pilot the benefit in one region of Namibia, with payments to begin in mid-2004.

 

            The social partners of Luxembourg agreed to collect contributions through a project managed by two existing union-based NGOs.  The joint project is to be known as the National Social Trust Project of Luxembourg.  The Government of Namibia would take over funding of the benefit, and its extension to the whole of Namibia, when the pilot ends in 2008, provided the pilot proves to be successful.  It is assumed that the pilot benefit will reach around 4,000 households by 2008, and 70,000 households when the benefit is extended nationally.

 

            The Global Social Trust idea could be used in future to create funding for a global campaign for universal pensions in different countries.

 

6.         A proposal: A universal pension project

 

            International resources could be sought to launch a global project ("The Universal Pension Project”) that would initiate feasibility studies along the lines of the above procedure in about 50 to 100 developing countries that do not yet have an old-age pension system with wide population coverage.  The studies could be undertaken by national research groups which could be trained in the use of the ILO methodology.  If the introduction of universal pensions at national level is deemed feasible, the formation of national action groups could be organized to undertake promotion activities. A global NGO (HelpAge?) or a consortium of NGOs in the field could co-ordinate the research and support for promotion activities.  The ILO could provide technical training and guidance. The ILO’s International Financial and Actuarial Service (ILO-FACTS) has just requested internal funding for three pilot studies (probably in Ethiopia, Laos and Chile). If approved these studies have to be finalized by the end of 2005.

 

            Based on the tentative budget for the ILO pilot studies, it must be assumed that the cost per country study (including the initializing of national promotion campaigns if the introduction of universal pensions is found feasible) would be in the order of US$200,000.  The global volume of the initiative could thus be in the order of US$10 to 20 million - a fraction of the amount of money that annually goes into development aid.[15]

 

 

7.         Conclusion

 

            There is a significant body of evidence that modest universal pensions have positive social effects for whole families in developing countries.[16]

 

            In macro- and micro-economic terms, universal pensions have few negative side-effects.  They redistribute consumption, but would probably only have minor effects on the overall levels of consumption or savings and investments. The effect on labour market participation would also most likely be negligible since pensions are paid to people that will have retired from regular labour market participation. There are indications that the overall cost of such schemes is not prohibitive at least for a number of developing countries.  There are also examples that show that the benefit can be introduced gradually.

 

            All the technical tools are available to design and manage national resource mobilization strategies and thus build the factual basis for credible promotion campaigns. If transitional international financing is necessary in individual countries there are mechanisms emerging that can be tapped and are presently developing in the context of the PRSP process and possibly the Global Social Trust approach or similar facilities.

 

            Once the theoretical financial and fiscal feasibility and the social utility of basic pensions have been established, the promotion of universal pensions on a broad scale could become a substantial contribution to the achievement of the first Millennium Development Goal (MDG) (“halving $1-dollar poverty by 2015”) as well as a major component of the ILO’s campaign on the extension of coverage and the Global Social Trust Initiative.  The suggested Universal Pension Project would also be a concrete follow up to the Plan of Action that the Second World Assembly on Ageing adopted in Madrid in April 2002.

 

            There may be a need for further global intelligence building but the battle for hearts and minds has to be won in each country. This can only be done if the feasibility and affordability of universal pensions can be proven by credible feasibility studies.  Compared to overall amounts of technical aid that flow into developing countries, the cost of such studies, and thus show whether we can reach out to 100 million poor elderly, are minute.  There is no reason to wait any longer.

 

 

*     *     *


References

 

Arenas de Mesa, A. and Benavides Salazar, P.:  Proteccion Social en Chile, Financiemento,      Cobertura y Desempeno, 1990 –2000, ILO, Santiago de Chile, 2003.

 

Case, A.: (2001) Does money protect health status? Evidence from South African         pensions.          NBER Working Paper 8495, National Bureau of Economic Research, Cambridge, M.A.

 

Case, A. and Deaton, A.: (1998), “Large Scale Transfers to the Elderly in South Africa”,           Economic Journal, vol. 108, No. 450, pp. 1330-1361.

 

Charlton, R. and McKinnon, R.: Pensions in development, Aldershot, 2001.

 

Cichon, M., Scholz, W. (et al.):  Social protection financing, ILO, Geneva, 2004 (forthcoming).

 

ILO:  Global Social Trust – Investing in the world’s social future. Report and documentation of a           feasibility study (forthcoming, Geneva, 2002).

 

Moller, V. and Devey, R.: (2001) “Trends in living conditions and satisfaction among poor older            South Africans: objective and subjective indicators of quality of life in the           October           household survey”. Paper presented at the 17th Congress of the International Association   of Gerontology, Vancouver, 1-6 July 2001.

 

World Bank: World Development Report 2000/01 – Attacking poverty, Washington, 2000.

 

Lund, F. and Srinivanas, S. (ILO-STEP and WIEGO): Learning from experience – A   gendered approach to social protection for workers in the informal sector, Geneva,   2001.

 

Scholz, W., Cichon, M., Hagemejer, K.:  Social Budgeting, ILO, Geneva, 2000.

 

Willmore, L. (UN): Universal pensions in low-income countries, Discussion paper IPD-01-05, (mimeo), 2001.

 

Willmore, L. (UN): Universal Pensions in Mauritius: Lessons for the Rest of Us, UN DESA       Discussion paper No. 32, New York, United Nations, April 2003.

 

ID21 insights (R. Disney): Africa in crisis – Hazards rise for prime age adults      WWW.id21.org/insights/insights42/insigths-isss42-art03.html

 

HelpAge International and the Institute of Development and Policy Management (IDPM),                 University of Manchester: Non-contributory pensions and poverty prevention – A      comparative study of Brazil and South Africa, DFID, London, 2003.



[1] Fiona Kilpatrick and Florian Léger of the Financial, Actuarial and Statistical Services Branch of the ILO provided helpful comments on the draft.

[2] Lund and Srivinas state (see Lund and Srivinas (2000) p. 107) “The South African state assistance to pensioners has been vital in ensuring household security, has acted as proxy for unemployment coverage for younger household members with no access to the formal scheme, has helped smooth income and consumption, and contributed to agricultural inputs – apart from assisting with the care of children.  This is a good example of a state subsidy in anchoring the ability of people in households to plan for integrated risk management.”  Le Roux (1995) as quoted by Charlton and McKinnon (2001, p.213) finds that in South Africa: “… people are fed and sent to school out of this pension money, it enables investments in farming activities, and in general it is crucial for the very survival of these communities”.

 

[3] Some studies have attempted to assess the impact of the South African pension on children's attendance at school and pre-school. However, efforts to isolate the effects of the pension have been thwarted by numerous intervening variables, notably the introduction of school feeding programmes in the early 1990s (Moller and Devey, 2001).  However, in South Africa, the pension receipt is of course not conditional on school attendance of children in the household. The feasibility study conducted here might explore whether there is the possibility to transform the nature of the individual old-age pension to a household benefit that is paid to the elderly but tied to certain conditions (such as school attendance or possibly nursing responsibilities).

 

[4] Case (2001) found that for both African and Coloured people, the presence of a pensioner was positively correlated with children's height (the presence of one pensioner is associated with an additional 3 to 4 cm of height).

 

      [5]   See Willmore (2001), p. 14.

 

     [6]  See World Bank: World Development Report 2000/01: Attacking poverty, Washington, 2000.

[7]  The methodology is described in full detail in Cichon et al. (2004), ch. 7.

[8] See for example: Areneas de Mesa, Benavides Salazar (2003).

[9] One main “competitor” for a universal pension scheme could be child benefits conditional upon the school attendance of children.

[10] Of equal importance to the sound design of a social transfer system is maintenance of the scheme through good governance. The effectiveness and efficiency of the new transfer scheme has to be constantly monitored to avoid abuse or the deterioration of benefit levels.  Monitoring can make use of the same instruments as the design process.  Indicators developed for the SPERs can be used to monitor performance.  Medium-term budget planning can utilize the social budget model. Sound financial governance requires, furthermore, well-trained financial managers.  The ILO is presently training financial managers in social protection (Masters of Social Protection Financing) together with the University of Maastricht. However, long-term governance issues – crucial as they are - are outside the scope of this paper.

 

[11] The estimate is based on a benefit of US$0.5 per capita and day as a cash transfer. The health care cost of US$120 per annum and capita (which is an extremely conservative assumption) and educational cost of US$240 per annum per child, as well as the administrative cost of benefit delivery of about 20% of benefit cost.

 

[12] Calculations based on data from Willmore (2003).

 

[13] See GB.285/ESP/4.

 

[14] See GB.285/13, para. 146.

[15] The proposal is in line with the renewed commitment of the ILO – as expressed by the International Labour Conference (ILC) in 2001 - to extend the coverage and improve the governance of social security, as well as the mandate of the ILC’s Committee on the Informal Economy “…to improve and extend social security coverage to all those in need of social protection, especially those in the informal economy, inter alia, through the piloting of innovative ideas …”. 

 

[16] The impact on families could even be strengthened further if one were to tie benefit receipt to certain sharing conditions within families, such as making pension payments to grandparents dependent on the school attendance of grandchildren.  In addition to the long-term investment aspect this might even have positive effects on adult employment.