THE IMPLICATIONS OF HIV/AIDS FOR SOCIAL PROTECTION

The production of this paper was funded by the Department for International Development. The arguments presented in the paper are the responsibility of the author alone. Comments to r.slater@odi.org.uk.

The myriad of impacts of HIV/AIDS on poor people is increasingly well documented. Poverty research from many different disciplines and sectors has contributed to a growing understanding of the current and potential future impacts of the epidemic. The main outcome of this improved understanding is recognition, on the parts of governments, donors and civil society, of the need to take action in order to mitigate the growing vulnerability and poverty effects of the epidemic.

At the same time, there has been a re-emergence of concern amongst policy-makers regarding the ways in which social protection interventions can help households cope with livelihood insecurity. Debates about social protection focus on new types of interventions and on appropriate targeting and implementation mechanisms. Through a review of current understandings of the impact of HIV/AIDS and a brief analysis of emerging new perspectives on social protection, this paper demonstrates why it is important for policy-makers and practitioners to explore the implications of HIV/AIDS for social protection.

Assessment of the impacts of HIV/AIDS on social protection focuses on four main types of intervention: transfers (such as food, cash and inputs); public works programmes (food for work and cash for work); education and training (particularly around prevention, nutrition and life skills for orphans); and financial resources (micro-credit, savings and insurance).

· Singling out the HIV/AIDS epidemic as a special and unique kind of crisis can be useful for directing resources and political attention towards dealing with the impacts of the epidemic. However, actual activities focusing on HIV/AIDS mitigation and coping should be part of larger programmes (for example those dealing with chronic illness or food security).

· Except in very specific circumstances, social protection mechanisms should target vulnerable people in order to reduce risks, some of which are the result of HIV/AIDS and some of which have other sources, rather than people affected by HIV/AIDS specifically/only.

· Support should be targeted to households and not just individuals, because of the problems that emerge when an AIDS patient dies and because, since it is generally orphans left behind, household recovery options are severely hampered.

· Direct targeting of HIV/AIDS orphans, as opposed to other orphans, raises equity and social justice problems and is, in many cases, inappropriate. HIV/AIDS orphans should be supported alongside other orphans who have similar needs, for example, with alternative curriculum and training at school to help them take on adult roles and responsibilities.

· Food for work (FFW) and cash for work (CFW) programmes can be appropriate for HIV-positive but asymptomatic people, but these should be in parallel with other transfers, notably food and cash, for households that are labour constrained through morbidity or mortality effects. Running FFW and CFW programmes in parallel with food and cash transfers is important in preventing children, especially orphans, from being forced into labour markets.

· Innovations in microfinance to support HIV/AIDS-affected and other vulnerable households should be encouraged, accompanied by a careful consideration of the embedded inequalities in communities that may result in exclusion of HIV/AIDS-affected households.

· Various institutions have a role to play in contributing to or implementing safety nets. Outside HIV/AIDS-affected households and communities, other stakeholders, notably NGOs, governments and donors should scale up community safety nets without generating a ‘crowding out’ effect. Partnerships among NGOs, governments and donors are crucial in this respect.

· Better coordination is required among NGOs, governments and donors and could be provided through a National AIDS Authority with a multi-sectoral mandate. However, actual programmes and projects should be mainstreamed into sectoral activities, in part to prevent HIV/AIDS exceptionalism.

· Social protection interventions should be designed around impact rather than prevalence rates, and donors, governments and NGOs should ensure an appropriate balance between prevention, care and recovery activities, whatever the prevalence rates.

· Donors and governments should acknowledge the policy choices that are made between fixed-life projects that promote people’s livelihoods through economic growth, and recurrent expenditure on social protection for households that cannot contribute to, and are unlikely to benefit from, economic growth. They should recognise that the HIV/AIDS epidemic will create a long-term welfare bill and find ways of supporting this.

1 Introduction: why focus on social protection for HIV/AIDS?

The aim of this paper is to explore the implications of HIV/AIDS for social protection policies, programmes and instruments. With increasing numbers of people infected with the virus, and spiralling morbidity and mortality effects, there is an urgent need to think about the ways in which different kinds of social protection interventions might be used to help people reduce, mitigate or cope with the impacts of HIV/AIDS. How far can existing forms of social protection absorb the impacts of the epidemic? To what extent are new instruments or a new social protection strategy required?

This paper begins by exploring what is already known and understood about HIV/AIDS and about social protection. It then develops an analytical framework, based on the World Bank Social Risk Management framework, through which to consider the kinds of social protection instruments that might enable people to reduce, mitigate or cope with the risks and vulnerability associated with HIV/AIDS.

In order to explore these issues in greater details, the remainder of the paper will focus on three main (1–3) and three subsidiary (4–6) objectives:

1. Exploring the most effective social protection instruments for addressing HIV/AIDS affected households and ways in which social protection policy and mechanisms can be made flexible enough to respond to the rapidly changing and unpredictable nature of impacts.

2. Identifying the most effective means of providing social protection to families and communities that support orphans, including particular measures that are likely to benefit the increasing number of elderly people looking after orphans.

3. Identifying which institutions (CBOs, other NGOs, the private sector, public services or agencies like social action funds) are currently providing social protection for people affected by HIV/AIDS, exploring which are the most effective channels for social protection provision, and identifying challenges and policy trade-offs faced by these institutions.

4. Assessing whether there is a strong, or special, case for promoting social protection measures for HIV/AIDS-affected households.

5. Reviewing the issues and lessons learned around targeting HIV/AIDS-affected households and considering evidence of targeting and stigmatisation.

6. Exploring the gender dimensions of providing social protection for HIV/AIDS-affected households to identify whether some measures are better at reaching female-headed households than others.

2 What do we know about HIV/AIDS?

2.1 Household impacts

The impacts of HIV/AIDS are manifold and have different permutations in different social, economic, political and geographical contexts. This, and the fact that as the epidemic continues its impacts can be unpredictable, means that there is a need for more empirical research to improve the evidence base for policy-making.

In terms of agriculture, households that are affected by HIV/AIDS face decreasing asset status over time and become less able to produce enough, either for subsistence or for income generation. Their declining capacity to produce crops results from a number of factors. Farming households affected by HIV/AIDS experience labour shortages, either through loss of labour (when people become unable to work through illness and, ultimately, die) or through the displacement of labour (as household members look after those who are sick in the household rather than working in their fields). Agricultural productivity also falls, because of a lack of investment. Money that would otherwise be spent on fertilisers and other inputs is allocated towards paying for medicines and funerals. Finally, when people die from AIDS, the local knowledge and skills that are crucial for successful agricultural production are not passed down to the next generation (Barnett and Blaikie, 1992; Barnett et al., 1995; Gillespie and Loevinsohn, 2003).

HIV/AIDS also has implications for food security. Urban households where people are sick and unable to work have reduced entitlements to food. Rural households oriented towards subsistence production struggle to produce enough food and have no surplus labour to supply larger commercial farms or to move into off-farm labour markets. The ‘New Variant Famine’ Thesis posits that a different type of famine is emerging, driven not by drought or conflict but by the effects of HIV/AIDS as it increases the vulnerability of households to shocks and risk (de Waal 2003; WHO 2003; FANTA 2001).

Linked to food security, there is evidence, albeit incomplete, that HIV/AIDS also has implications for nutrition. Beyond limiting households’ capacities either to grow or buy enough food to meet their nutritional requirements, HIV-negative people with poor diets are more susceptible to infection, HIV-positive people with poor diets develop AIDS more quickly, and people with AIDS have increased nutritional requirements (Gillespie and Haddad, 2002). In addition, anti-retrovirals (ARVs) must be combined with a good diet in order to be most effective and to avoid side-effects (de Waal, 2003). However, knowledge about the precise impacts of different aspects of nutrition on different PLWHA is patchy, and WHO have identified significant knowledge gaps where more research is required (Table 1). There are operational implications resulting from these knowledge gaps, particularly as to what food and nutrition support programmes should do differently because of HIV/AIDS. There are strong arguments for increasing ration sizes, but less clarity over fortification and micronutrients, given the mixed outcomes from supplements identified in Table 1.

It is important also to highlight the social development impacts of HIV/AIDS. The impacts of HIV/AIDS reflect inequitable gender relations (Baylies, 2002): Women are more likely to be infected (both because of physiology and because they are less able to protect themselves through abstinence or condom use), and they take on greater burdens of caring for the chronically ill. Men tend to die before women, increasing the number of female-headed households. There are many published references to the susceptibility of widows to property-grabbing by in-laws (Baylies, 2002; FAO, 2003; FAO, 2004a), though elsewhere the evidence on this and distress sales has been dismissed as anecdotal (Aliber et al., 2004).

This points to a common problem in our understanding of the impacts of HIV/AIDS: evidence used in policy-making has often been anecdotal rather than empirical, and findings from significant empirical studies have been assumed to hold weight in other geographical, social and economic contexts.

Increasing numbers of orphaned children represent another outcome of the HIV/AIDS epidemic (Box 4). In 2001, there were 13.4 million AIDS orphans. This figure is expected to reach 25 million by 2010 (UNICEF, 2002). Many orphans and vulnerable drop out of school because there is no money to pay for school fees, uniforms and books, and because the opportunity cost of lost labour in agriculture or in domestic work, including caring for the sick, is high. This applies particularly to girls. Orphans taken in by adults are most often cared for by the elderly, who are without the resources and income-generating capacity to feed and clothe themselves, never mind additional children. Child-headed households are at greatest risk of utter destitution (Levine, 2001).

1.1 What do we know about HIV/AIDS at the macro-level?

Whilst our understanding of the micro-level impacts of HIV/AIDS has progressed, at a macro-level our knowledge is patchy and, in general, qualitative. Quantitative data about the prevalence of HIV and of AIDS, and effects on life expectancy and death rates, tend to be unreliable and based on samples from ante-natal clinics. Furthermore, there are millions of people in the developing world who do not know their HIV status. Similarly, it is very difficult to aggregate up the impacts on households and individuals to understand how HIV/AIDS affects economic growth at national or regional level (Anderson et al., 2004). There have been some attempts to estimate impacts on economic growth and GDP. Modelling by Robalino et al. (2002) estimates that in the Middle East and North Africa, a region which has seen relatively low HIV prevalence rates, average GDP losses resulting from HIV/AIDS for the period 2000–25 could approximate 35% of current GDP. In different economic sectors the impacts are different,; they are in general greater in sectors where male workers live away from their families (for example transportation, construction and power generation) (Bollinger and Stover, 1999). The Population Division in the Economic and Social Affairs Department of the UN Secretariat demonstrates the varying findings of research, arguing that ‘In many of the highly-affected countries, studies have been undertaken to model the impact of HIV/AIDS on economic growth. In some cases, estimates of the economic impact of HIV/AIDS have been ‘small’. In other cases, annual reductions of 2–4 percentage points of gross domestic product per year have been found.’ (UN Secretariat, 2003: xiv–xv). The actual impact may be worse than estimated because calculations of the impact on GDP do not take into account the damage caused by lower investments in human capital, particularly children’s education.

In spite of the difficulties associated with understanding the macro-level, particularly macroeconomic effects of HIV/AIDS, an important outcome of research into the impacts of HIV/AIDS is the increasing acknowledgement by governments, donors and civil society that HIV/AIDS is eroding the hard-won development progress of the last few decades. Whilst UNDP reports increases in global life expectancy through the 1990s, this has fallen in countries with high prevalence rates, including Thailand, Botswana, Malawi, South Africa, Zimbabwe and Zambia (UNDP, 1996; 1997; 2000). Child mortality is on the increase too. AVERT (an international HIV and AIDS charity) estimates that under-five child mortality rates will more than double in countries such as Botswana, Kenya and Zimbabwe by the year 2010 (http://www.avert.org/children.htm), with implications for the MDGs. Whilst the identification of the first AIDS cases in many countries was met with denial, either about the causes of the disease itself or about the seriousness of the epidemic (Population Foundation of India, 2003; Barnett and Whiteside, 2002), this view has been replaced in most countries by recognition of the increasingly vulnerable livelihoods of those living with HIV/AIDS and of the fragile state of the broader economies in which they live.

Whilst the significance of HIV/AIDS for poverty reduction and development is now taken seriously, exactly what to do about the epidemic has become the central question for policy-makers. At the heart of debates is the issue of whether the HIV/AIDS epidemic is particularly different from other shocks or other epidemics and whether it warrants special or specific attention. There are significant differences of opinion on this issue (Baylies, 2002). Currah and Whaites (2003) argue not for special focus on HIV/AIDS per se, but a focus on countries affected by HIV/AIDS. However, given that development funding is limited, there are concerns that maintaining a special focus on HIV/AIDS can divert attention away from the other causes of people’s poverty. In Zambia, the food shortages in southern parts of the country in 2002–03 were not so much the result of HIV/AIDS but of the reduction in veterinary services under structural adjustment which led to outbreaks of corridor disease and a decrease in the amount of draught animals available for ploughing. Another example is the potential diversion of healthcare workers and resources away from malaria, typhoid, cholera and other diseases into HIV/AIDS. It is also argued elsewhere that the priority should be tackling vulnerability, whether it is caused by HIV/AIDS or not.

2 What do we know about social protection?

In parallel with the growing literature on the impacts of HIV/AIDS there has been a return to concerns with options for supporting households that are in danger of becoming destitute. Social protection is defined here as per Shepherd (2004), as a range of processes, policies and interventions to enable people to reduce, mitigate, cope with and recover from risk in order that they become less insecure and can participate in economic growth. Thus, rather than focusing solely on safety nets (for coping), there has been a shift towards identifying potential linkages between the protection of people’s livelihoods and the promotion of livelihoods through economic growth (Devereux, 2001; Farrington et al., 2004). This new conceptual content in social protection focuses on ‘how public actions designed to help people manage risk and adversity may contribute to larger policy objectives of economic growth and poverty reduction’ (Conway and Norton, 2002: 533). The scale and seriousness of HIV/AIDS in some countries, particularly in sub-Saharan Africa, poses a challenge to some of this new thinking on social protection and its contributions to economic growth. Vulnerability caused by HIV/AIDS has led to demands for social protection measures to assist HIV/AIDS-affected households. However, there is also recognition that people who are infected with or affected by HIV/AIDS may be unable to contribute to economic production. It may also be useful to consider the benefits of social protection within a broader context, that of the social and long-term economic value in providing support for the children of those affected by HIV/AIDS, who might otherwise become destitute. There are good arguments for investments in young people, given the evidence that certain conditions of childhood poverty lead to the transmission of poverty over lifecourses and to future generations (Harper et al., 2003).

In terms of approach, this paper also seeks to strike a balance between assessing what social protection options will be useful and appropriate given current resources, and what might be possible given a greater allocation of resources. The growing prevalence and impact of HIV/AIDS could provide new opportunities to increase support for social protection and there is some evidence of additional resource to fight the epidemic. However, given the growing capacity constraints, it is important to also think about what is possible given current or declining resources.[i]

The following section of the paper develops an analytical framework to explore appropriate social protection interventions for people living with HIV/AIDS. Drawing on evidence mainly from sub-Saharan Africa and from Asia, potential interventions or tools are assessed through a matrix that identifies the advantages and disadvantages of different tools, their feasibility (resource constraints, levels of prevalence and capacities of communities, governments or donors) and their desirability (for beneficiaries and for politicians). The final section of the paper makes recommendations for developing an appropriate and effective policy on HIV/AIDS and social protection.

1 Assessing social protection interventions for HIV/AIDS

Before considering various different social protection interventions, it is important to establish an analytical framework for exploring the interventions. This involves clarifying what we mean by various HIV/AIDS-related terms and showing what the implications of these differences are for social protection.

1.1 Social protection for whom? Individuals, households and communities

First, we need recognise that the social protection needs of people who are HIV-positive but asymptomatic are different from those of people who have AIDS and increasingly debilitating diseases. Not all HIV-positive people are unproductive and unable to work, so interventions should not be restricted to handouts. Because it is often difficult to identify (and therefore target) those who are either HIV-negative or HIV-positive but asymptomatic or those who have AIDS, in assessing interventions we need to remain mindful of the differential impacts on each of these groups.

Secondly, we need to think about people who do not have AIDS but live in households with people who are HIV or AIDS-infected and/or in households where people have died from AIDS. The household impacts of morbidity are different to those of mortality.

Thirdly, beyond households, we need also to include those in the broader community who are indirectly affected by the disease through disruptions to local labour supply and community-based safety nets.

The framework, then, involves distinguishing between the different stages of the disease and the impact at different levels, from individuals, households, communities and more broadly.

1.2 What can social protection do?

Theoretically, social protection interventions can include a massive range of activities (from cash transfers to price support for consumer foods or producer staples to inflationary controls) and it is not possible to consider all of them here. The preoccupations of different agencies reflect this broad range of possible interventions and their particular sectoral priorities (Box 2).

Much of the HIV/AIDS literature divides the response to HIV/AIDS into prevention, treatment/care and mitigation. Social protection instruments can contribute under each of these headings respectively by, for example, reducing the risk of infection through condom distribution or improved nutrition, through protection against deterioration in health using ART, or through the development of labour-saving technologies. In this paper, a different but overlapping approach from the Social Risk Management (SRM) framework (Holzmann and Jørgensen 1999) is used, and instruments are assessed according to whether they reduce, mitigate or help people cope with risk.[ii] Interventions that are dealt with in depth in this paper are divided into transfers, public works, education and training, financial services, and care and treatment. Two vulnerable groups are singled out for particular attention – orphans (Box 4) and the elderly (Box 3).

Box 2: Selected agency activities in social protection and HIV/AIDS[i]

ADB: Social protection is not explicitly linked to HIV/AIDS but focuses on: 1) labour market policies and programmes designed to promote employment, the efficient operation of labour markets, and the protection of workers; 2) Social insurance programmes to cushion the risks associated with unemployment, ill health, disability, work-related injury and old age; 3) social assistance and welfare service programmes for the most vulnerable groups with no other means of adequate support; 4) micro and area-based schemes to address vulnerability at the community level, including micro-insurance, agricultural insurance, social funds, and programmes to manage natural disasters; and 5) child protection to ensure the healthy and productive development of children.

FAO: Social protection activities include prevention and mitigation at a range of levels from: 1) grassroots (including voucher systems for improving access to farm inputs; improving nutrition; securing the asset base, especially land and other assets that improve agricultural productivity; and strengthening resilience through promotion of labour-saving technologies, introducing farmer life schools for OVCs; creating field-level methodologies for recording and sharing indigenous and agrobiodiversity knowledge); 2) through national policy environments and institutions (including guidelines for incorporating HIV/AIDS considerations into food security and livelihood projects; developing new assessment indicators; research to better understand impacts); and 3) global level (international advocacy drawing attention to inter-linkages between HIV/AIDS, food security, nutrition and the role of the agricultural sector in mitigation).

IFAD: IFAD’s poverty alleviation strategy focuses on the economic empowerment and development of the rural poor through organisational and institutional development; and through the facilitation of access to resources and their efficient use. Poverty is viewed as a driving force of HIV/AIDS and, simultaneously, HIV/AIDS increases the depth and extent of rural poverty. There are five main areas of IFAD’s response to the HIV epidemic: 1) HIV/AIDS information, education and communication programmes for HIV prevention and AIDS mitigation among IFAD target groups; 2) poverty alleviation and livelihood security programmes adapted to the conditions created by HIV/AIDS, including income-generating programmes, microfinance projects and adult literacy programmes; 3) food security and nutrition-related innovations of adaptation of existing practices; 4) socio-economic safety nets, with special emphasis on support to orphans and households fostering orphans; and 5) integrated HIV/AIDS workplace programmes for IFAD-supported projects.

ILO: HIV/AIDS is a workplace issue given that most of the world’s people who are infected with HIV are of working age. The focus of social protection activities in ILO is on basic worker rights (working against AIDS-related discrimination), establishing alternative workplace arrangements for workers with HIV/AIDS and for their carers (especially women), protecting OVCs from child labour, especially sexual exploitation, provision of practical guidance to employers and workers’ organisations for prevention of infection and promotion of behaviour change. The emerging challenge for ILO is incorporating decentralised systems of social protection for informal workers.

World Bank: 1) Labour Market Interventions: Helping governments’ and individuals’ skill enhancement programmes, improvements in the functioning of labour markets, and the development of active and passive labour market programmes. 2) Pensions Helping: governments take care of their older and aging populations through the creation of or improvements in private pension provision, mandatory savings and public old-age income support schemes. 3) Social Safety Nets: programmes designed to provide targeted income support and access to basic social services to the poorest population groups, and/or those needing assistance after economic downturns, natural disasters, or other events that pose major risks. 4) Social Funds Agencies: that channel grant funding to small-scale projects to help poor communities design and implement their own projects. 5) Lending to governments for social protection.

WFP: Focus on food and nutrition in the fight against HIV/AIDS includes: food aid to prolong the lives of people suffering with HIV and AIDS; free WFP school lunches and take home rations; food for healing; and HIV prevention and AIDS awareness campaigns.

UNICEF: Programmes concentrate on 1) Preventing mother-to-child transmission of HIV; 2) providing education, vocational training and psycho-social counselling to children orphaned by HIV/AIDS; 3) ensuring that young people are informed about HIV/AIDS prevention; 4) working with governments to make HIV/AIDS education part of the standard school curriculum; 5) strengthening families and community capacity to protect children through healthcare services and farming assistance; and 6) organising communications programmes to prevent the spread of HIV/AIDS.

Sources: ADB (2003); FAO (2004b); ILO Factsheet; ILO (2002); IFAD (2001); UNICEF 2002 WFP (2004).

1.1 Transfers for HIV/AIDS infected and affected people

The main advantage of transfers is that they can safeguard existing productive assets which otherwise might be drawn down to meet basic households needs. Drawing down on productive assets is a coping strategy that can lead households into a vicious cycle of impoverishment.

1.1.1 Food aid

The most common type of transfer in many of the countries with high HIV prevalence rates is that of food. (Food can, of course, be used in many other ways beyond transfers – for example, see later section on food for work.) In some circumstances, food may be the most appropriate form of transfer, particularly in remote areas where there is limited supply of food and where cash transfers are likely to cause inflation because of increased food demand without an accompanying increase in food supply (Devereux, 2002). Food distribution can also addresses nutrition issues – in Southern Africa between 2002 and 2004, estimates of calorific requirements and protein requirements were increased to take account of the greater nutritional needs of people with AIDS. Targeting additional food to individuals within households is unlikely to be effective, as any additional food supplies are likely to be shared amongst household members. Furthermore, adult household members (who are much more likely to have HIV or AIDS and therefore to require more food), especially women, often give additional food to their children. FANTA (2002) identify various examples of situations where food aid may not be appropriate, but all but one of these examples (stigma) is about food aid in general rather than food aid as a risk-coping intervention against the impacts of HIV/AIDS explicitly. (And indeed, the stigma issue is common to many other interventions, not food aid alone, and will be returned to later.)

What is significant about the food transfer example is that the development of the New Variant Famine thesis (at the same time as the Southern Africa humanitarian appeal of 2002–03) has led to the re-emergence of debates about the appropriateness of food distribution as a long-term developmental response (in this context as a risk-coping response for HIV/AIDS-affected/infected people).

There is a long debate and a great deal of literature about the possible negative impacts of food aid, particularly when delivered over the long term (Clay and Stokke, 1991). Negative impacts may include disincentives to local production and impacts on local markets. Another commonly cited concern is of food aid creating dependency. However, HIV/AIDS may reinforce the need for a shift in perspective in which a degree of reliance on long-term welfare is accepted for the most destitute. As de Waal argues (2003: 21): ‘We must face the distinct possibility that we can no longer talk about food aid and other forms of welfare assistance as short-term measures until ‘normal’ development is “resumed”’.

However, food aid it not always the most appropriate type of transfer:

Perhaps the most fundamental issue is around the appropriateness of food aid as a resource over the long term. There appears to be a common assumption that, because HIV/AIDS exacerbates food insecurity, and because people with HIV/AIDS have additional nutritional requirements, food aid is needed. Few of the agencies interviewed in southern Africa had explicitly considered the ongoing appropriateness of food aid, or alternatives to it, in countries such as Malawi and Zambia, where there were no longer national-level shortages in 2003. WFP in southern Africa sees food aid as a catalyst for non-food activities, and is emphasising regional purchases of food where possible. In the first emergency appeal, WFP bought nearly 400,000 tons of food in southern Africa (WFP, 2003b). The WFP’s second regional emergency appeal makes reference to food as a component in wider service delivery programmes that aim to ‘deploy food, technical advice and advocacy measures to encourage and support government efforts to create or strengthen safety nets that provide minimum protection to populations facing food insecurity and the risks of living in an HIV/AIDS affected environment’ (WFP, 2003b: 8).

Source: Harvey (2004: 33).

The costs and benefits of food aid over the medium to long term need to be compared with other ways of providing long-term welfare and safety nets. It should not be assumed that, even where people with HIV/AIDS are hungry and have inadequate diets, food is the only or best way of addressing their needs.

1.1.2 Cash transfers

With references to emergencies in general, rather than HIV/AIDS specifically, Peppiatt et al. (2001: 1) argue that, in comparison with food distribution, ‘cash is more cost-effective because its transaction costs are lower, it is more easily convertible, allows for greater beneficiary choice and can stimulate local markets. On the other hand, cash can be used in ways not intended by the donor, can contribute to local inflation and poses security risks not normally associated with food aid.’ Where cash is used in ways not intended by the donor, this may indicate misappropriation, but also possibly that the donor has not fully understood people’s needs. Transfers such as food aid offer people no choice (and sometimes include the distribution of inappropriate foodstuffs), whereas cash transfers can be used to buy food, household goods and can be invested in income-generating activities (IGAs). For women, these can be activities (brewing, sewing, etc.) which can be easily combined with household work and caring roles, rather than activities (including agriculture) that force women to travel long distances from their homes. Giving cash rather than food can, in some cases, also enable the avoidance of complex targeting mechanisms. Where interventions are attempting to differentiate between people who require relief, rehabilitation or recovery support, households can tailor their use of the cash transfer to their own specific relief or rehabilitation or recovery needs, without the involvement of donor or government institutions.

However, a growing body of evidence from countries with starkly contrasting HIV prevalence rates demonstrates the important of social pensions paid to elderly people, particularly those who are supporting orphans (Box 4). Pension payments are often used to pay for education costs of grandchildren or to buy food for the rest of the household (IDPM and HelpAge International, 2003). There is a frequent assumption that social pensions for the elderly are unaffordable in all except middle-income countries such as Brazil, South Africa and India. Whilst there are certainly large recurrent costs associated with pensions, these need to be balanced with the growing numbers of orphans and vulnerable children, particularly in Africa. Devereux (2003) argues that the payment of social pensions in some countries is not necessarily unaffordable; rather, it represents a policy choice where politicians have prioritised (fixed-term) investments which are intended to alleviate poverty by driving economic growth. Similarly, Farrington et al. (2003) make strong arguments about the (in)efficiency of food distribution systems in India and suggest that for every rupee of food delivered, there are administration, transportation and storage costs of one to two rupees. They argue that ‘cash transfers paid through certain channels (e.g. the Post Office) for specific purposes such as pensions and allowances are less corruptible than many ‘in kind’ transfers. They may help in reducing under-nutrition and stimulating the local economy by reducing ‘demand deficits’ and merit increased funding’ (p.1). In the context of the growing prevalence of HIV/AIDS in India (which now has the second-largest number of infections after South Africa) and the likelihood of a growing number of orphans being supported by the elderly, pensions could play an important role in mitigating and coping with the effects of HIV/AIDS. There are signs that, amongst governments which have prioritised dealing with HIV/AIDS in, for example, their PRSPs, there has been a rethink about the affordability of long-term cash transfers. In Lesotho, for example, the government has just introduced a pension for people over the age of 70 years. The size of the cash transfer is small – 150 maloti (about US$21) – but can buy a 50 kg bag of maize-meal.

Box 3: Old-age allowances in South Africa and Nepal

With HIV infection rates of 24.5% in 2001, South Africa experiences some of the highest HIV prevalence levels in the world. In comparison, Nepal is in a much earlier stage of the pandemic, with adult prevalence in the 15–49 age group of 0.5%[i] (http://www.unaids.org/nationalresponse/result.asp). However, evidence from both countries demonstrates the importance of transfers to old people.

In Nepal, all people above 75 years old have been entitled to a payment under the Old-age Allowance Program (OAP). Payments of 100 rupees (per month) were first made in 1995 and increased to 150 rupees in 1999. By 2002, there were nearly 200,000 beneficiaries in the programme plus 227,000 receiving helpless widows assistance (for widows between 60 to 75 years) and nearly 4,000 receiving disabled pensions (Irudaya Rajan, 2003). Whilst there has been research on the process through which people apply for pensions, much less is known about the ways in which allowances are utilised and the extent to which they support others in society. Accompanying investments to support old people (for example, the construction of old-age homes by NGOs) suggest that old-age allowances are intended directly to benefit the elderly and not their relatives or orphans.

Elsewhere, there has been more research on the utilisation of old-age pensions, particularly in middle-income countries in Southern Africa (for example IDPM/HelpAge International, 2003; Barrientos, 2003). In South Africa, it has been demonstrated that non-contributory (or social) pensions are shared within households and can have a substantial impact on poverty, both long and short-term, for both the elderly and their dependents. Barrientos estimates that social pensions in South Africa reduce the poverty headcount by 2.8% (Barrientos, 2003). The burden on elderly people is growing rapidly as the number of AIDS-deaths increases in South Africa. Ferreira et al. (2001) argue that older persons have to take on roles as carers for those who are terminally ill, and carers and providers for the dependents of the terminally ill or those who have already died, whilst Whiteside and Sunter (2000) estimate that, by 2005, there will be nearly one million AIDS orphans in the country. In this context, social pensions to the elderly will become increasingly important and their roles in supporting orphans and the chronically ill should be recognised (Legido-Quigley, 2003). However, it is also important to remember that not all elderly carers are of pensionable age (over 60 for women and over 65 for men in South Africa). Hunter and May demonstrate the growing vulnerability of 50–59 year olds as old age is approached, highlighting the ‘risk of unemployment or retrenchment, rising costs of living, the possibility of loss of assets or constraints to the effective use of assets, the possible reintroduction of reproductive work [especially caring]’ (2003: 2). This highlights the need to explore orphan allowances for carers, in addition to expanding the role of old age pensions.

Cash transfers to/for orphans have been the subject of only very limited research, but should be explored, especially given that not all elderly carers are of pensionable age. Cash transfers to orphan may be inappropriate for orphans unless they have benefited from specific training or support to help them make sensible purchasing and budgeting decisions. It is also possible that people in the surrounding community may take advantage of, mislead and exploit orphans as they attempt to manage household resources. However, it may also be the case that orphan allowances (paid either to orphans or their adopted guardians) encourage other households to take double orphans in. There has been increased demand for child-benefit allowances in South Africa and Thailand; these ‘demands may, however, compete with other claims on the budget, with the risk that the support to families with orphans, foster families, poor relief and so on, may stagnate in the face of mounting needs’ (Cornia and Zagonari, 2002: 11).

Box 4: HIV/AIDS, social protection and orphans

The number of single and double AIDS orphans is set to rise to 25 million by 2010 (UNICEF, 2002). The Children on the Brink Report in 2002 lays out the main issues and problems that require urgent attention (UNAIDS/UNICEF, 2002). The most immediate point is that AIDS threatens children’s lives:

The impacts of AIDS on children are both complex and multi-faceted. Children suffer psychological distress and increasing material hardship due to AIDS. They may be pressed into service to care for ill and dying parents, be required to drop out of school to help with farm or household work, or experience declining access to food and health services. Many are at risk of exclusion, abuse, discrimination and stigma. (UNAIDS/UNICEF, 2002: 4).

There is, however, a danger of HIV/AIDS exceptionalism – a danger that orphans from causes other than HIV/AIDS will be left out of programming.[ii] Similarly, the focus on HIV/AIDS orphans could mask the problems of vulnerable children who are not directly affected by AIDS. There are, as raised elsewhere in this paper, good reasons for investing in children in order to diminish the possibility of the transmission of poverty through the lifecourse and to future generations ( (Harper et al., 2003).

Some of the key instruments that are being developed to support orphans and other vulnerable children (OVCs) are discussed in the main text. Additional evidence is presented in this box.

Cornia and Zagonari (2002) argue for two main priorities in tackling the impact of HIV/AIDS on children. In the long term, ‘the negative impact of HIV/AIDS needs to be counterbalanced by ensuring that enrolments in primary and secondary education are sustained by means of traditional academic planning and through measures … such as curriculum simplification, the waiver of enrolment fees, special provisions for the education and training of a mounting number of orphans’ (p. 26). They also discuss the potential for orphan allowances which have been introduced in various countries.

Income transfers can be targeted directly or indirectly to AIDS-affected children (through orphan allowances, foster care allowances, basic pensions for the elderly – who often are in charge of a number of orphans – as well as to impoverished people sick with HIV and AIDS). Such transfers can be in kind (food, clothing, less fungible than money), in cash (book/school/transport allowances) and exemptions from school and medical fees. Elements of such schemes are in existence in several AIDS-affected countries. In Botswana the government introduced in 2000 a ‘package’ of subsidies in kind for orphan children worth US$60 per child/month. South Africa has instituted a child support grant, a foster care allowance and a care dependency grant for children with severe problems (see Section 6). Thailand has developed a mixed system in which temple and community-based transfers start to be accompanied by interventions targeted at children originating from the central government. Even financially stretched countries such as Zambia have considered a modest transfer system (worth half a million dollars a year) to offset school cost of AIDS orphans (personal communication of UNICEF Zambia). These programmes have to be expanded, better analysed, and evaluated (2002: 29).

The authors note the need to pay particular attention to key design issues, including questions of whether payments should be direct or indirect (i.e. to orphans or to carers); whether the target population should be all children in AIDS-affected families, only AIDS orphans, all orphans, or all OVCs; and whether governments, NGOs and communities have the capacity to distribute the allowances. Finally, they highlight the danger of incentives traps and stigma for children.

1.1.3 Inputs programmes

These have traditionally been focused on agricultural production and, whilst their stated aim is usually about increasing agricultural productivity, they also have important risk-reducing effects. Examples include the distribution of different (drought-resistant) seed varieties to help households extend the length of the cropping season and so smooth income, and distribution of free fertiliser, which reduces the risk associated with spending money on it only to see crops fail. Given that poor households are very risk adverse, inputs are seen as an important risk reduction strategy. However, for households that are severely labour constrained, agricultural inputs may be inappropriate, although options for adapting inputs programmes to provide labour-saving technologies should be explored.

1.1.4 School feeding

Another intervention in which there is renewed interest in the context of HIV/AIDS is that of school feeding. In the past, school feeding schemes have aimed to promote nutrition amongst children, to improve enrolment rates and reduce dropouts. Nutritional impact on children has been questioned (in the sense that feeding schemes have a household substitution rather than child nutrition effect, whereby children eat at school instead of at home); however, there is evidence that school feeding improves enrolments and reduces dropouts, particularly amongst girls (Devereux 2002; Farrington et al., 2004). HIV/AIDS threatens to undermine school enrolments, as children are taken out of school to work in the fields and generate income for the household. The long-term impacts will be a generation of adults who, if they are able to avoid HIV infection, do not have the skills gained through education to lift themselves out of poverty. It is unclear whether school feeding is enough of an incentive to keep OVCs in school, particularly in the case of child-headed households. As a result, school feeding schemes that previously provided a lunchtime meal at school for children, are also including take-home rations.

1.2 Public works programmes

Public works programmes focus largely on food for work (FFW) and cash for work (CFW) but can also include inputs for work. Programmes provide payment in cash or kind (food and transfers) in return for labour. The benefits of these programmes, in general and in comparison with other forms of transfers, have been discussed elsewhere (see, for example, Devereux, 2002; Gebhre-Medhin and Swinton, 2001) Central to the appropriateness of these programmes to HIV/AIDS-infected and affected households is the labour constraint that many households affected by HIV/AIDS face. Amongst donors and NGOs there is no agreement about whether CFW and FFW are appropriate for households affected by HIV/AIDS. One side of the argument is that both are inappropriate for people at risk of developing AIDS. Elsewhere it is suggested that asymptomatic HIV-positive people can participate in FFW or CFW schemes, and that perspectives (for example FAO’s ‘vulnerable but viable’ classification) wrongly assume that households with chronic illness are not viable. Harvey (2004: 35) argues that there is ‘an urgent need for better and more explicit monitoring and evaluation of the labour constraints relating to HIV/AIDS to see whether they really are restricting effective participation in agricultural input programmes’, and the same is true of participation in CFW and FFW. Furthermore, the Zimbabwe Red Cross argues that ‘if it is assumed that people HIV/AIDS are unable to benefit from input programmes without careful assessment, there is a clear risk that they could be further stigmatised’ (in Harvey 2004: 35). It is clear that inputs for work schemes should be designed to include labour-saving inputs that will be appropriate for the household, should illness set in.

In the case of FFW, self-targeting is frequently used to ensure that only poor households participate. However, self-targeting is usually achieved either by offering non-preferred foods (for example, yellow maize in Eastern and Southern Africa, or broken rice in Asia) or by paying low wage rates (i.e. an amount of food per day that is below the market wage). This could be counterproductive for people who are HIV-positive, since it is only with a full diet that people delay the onset of AIDS. Self-targeting processes that force households to reduce consumption because of low wages, or undertake physically demanding work without an increase in calorific intake, are likely to induce a more rapid onset of AIDS.

For households where one or more people are chronically ill, and other household members are under increasing pressure as carers, FFW and CFW are less appropriate.[iii] FFW and CFW programmes must also be designed alongside support for orphans and vulnerable children in order to avoid encouraging children to take on responsibilities of work associated with adulthood. Perhaps the single exception to this is food that children take home from working in school gardens.

1.3 Education and training interventions

These interventions are mostly risk-reducing and risk-mitigating activities. The previous section on school feeding highlighted the danger that the next generation of adults may not have the education, knowledge or skills to pursue a sustainable livelihood. The problem of OVCs drives the need for changes to the curriculum in schools to provide more appropriate skills and knowledge (especially ‘life skills’). There are examples, in Malawi and elsewhere, of schools specifically for orphans and, whilst these may have more appropriate curricula specially designed for orphans, there are both ethical questions regarding the separation of orphans from other children and questions about the danger of encouraging stigma and exclusion.

Other education and training interventions include the transfer of knowledge about local agrobiodiversity and skills in indigenous agricultural systems, and the dissemination of knowledge about the HIV virus and AIDS to prevent infection, encourage better nutrition, and prevent stigma.

1.4 Interventions focusing around financial services and financial capital

Various donors and NGOs have highlighted the important role that community-based financial services can play, particularly in ex-ante risk mitigation for households affected by HIV/AIDS. Interventions can be divided into three main groups: savings, micro-credit and insurance. In all cases, HIV/AIDS presents challenges to the sustainability of these activities and, whilst they may be appropriate for adults within the community, these types of activities have less direct relevance for OVCs. Financial services may be more appropriate in Asia (where prevalence of HIV and AIDS is mostly lower) than in sub-Saharan Africa (which, on the whole has high prevalence rates) because they are ex ante rather than ex post mechanisms.

1.4.1 Savings

Drawing down on savings is an important risk-mitigating strategy on the part of households affected by HIV/AIDS. In the case of social risk management, it has been argued that ‘financial saving as well as the accumulation of other assets that can be sold at fair market prices is perhaps the most important asset management instrument used to address income variability’ (Holzmann and Jørgensen 1999: 1015). In the case of HIV/AIDS, though, the disposal of savings is recognised as one of the first coping strategies that households draw on in times of stress. Encouraging savings ex ante is one way to help households prepare in advance for the effects of AIDS-related poverty. Since households are usually unaware of their HIV-status, encouraging savings needs to take the form of ex ante preparation for various kinds of poverty or vulnerability, not solely HIV/AIDS.

For poor people (particularly women), formal saving accounts with banks are often not possible, either because people do not have the required documentation to open accounts or because they live in areas that are too remote. Instead, savings can be held in other forms, including cash, jewellery and livestock. The multiple functions of livestock, including their importance in saving, have been documented by Cousins (1999). Given the unpredictability of the impacts of HIV/AIDS (so that households need to draw down on savings quickly and often in relatively small quantities at a time to pay for medicines) and the risks associated with large stock units (especially cattle and buffalo), smaller stock units, for example goats or chickens, can be a very important and flexible source of savings for poor households. An exploration of ways of supporting more flexible types of savings is worthwhile but must include broader impact assessment, for example of environmental impacts.

Private and community-based informal savings clubs are important throughout Asia and Africa, and offer savings opportunities to poor people who cannot access bank accounts or who face high transaction costs with formal banking. Savings clubs can be flexible and enable households to draw down small amounts of money to pay for medicines. Many donors are considering ways to support community-based savings as a risk mitigation strategy for households affected by HIV/AIDS. However, there is evidence from South Africa that certain kinds of social capital may contribute to, rather than reduce, HIV-infection. ‘Amongst members of stokvels (voluntary savings clubs accompanied by social festivities) however, young men were more likely to be HIV-positive, women of all ages were more likely to have a casual partner, and both young men and young women were more likely to drink alcohol than non-members’ (Campbell and Williams, 2000).

In exploring options for mitigating the impact of HIV/AIDS through savings, it is important to remember that, for households that are already poor, ex ante savings are not possible. For the poorest households, other options should be explored. Similarly, many countries with higher prevalence rates have already passed the window of opportunity where savings might provide one form of ex ante preparation for HIV/AIDS. In countries with lower prevalence rates, savings may be more appropriate.

1.4.2 Micro-credit

Parker (2000) argues that there is a need to heighten awareness about the impacts of HIV/AIDS on microfinance, given that, as the disease progresses, HIV/AIDS-affected clients of microfinance institutions are likely to need access to a wider range of financial services. She argues that, ‘even in its most basic form, access to microfinance services gives households a way to both prepare for and cope with crisis’ (2000: 2). However, households are likely to have a reduced ability to make repayments on loans.

Baylies (2002) is critical of the implied assumption in work on microfinance and HIV/AIDS that the sustainability of HIV/AIDS-affected households will depend, in part, on their fuller integration into the market economy. She suggests that ‘The appropriateness of this must be questioned’ since ‘micro-credit has clear limits where high levels of morbidity and mortality undermine the economic arena within which the logic of microfinance schemes is nested’ (p. 625). Poor people, including those affected by HIV/AIDS, are often risk adverse: they are unwilling and unable to take out loans that allow them to access more (risky) remunerative markets or activities and tend to be limited to subsistence activities.[iv] ‘They may also be more susceptible to debt, so that the provision of micro-finance can become a further burden rather than a means of recovery’ (p. 625). Microfinance programmes need to be more sensitive to the changing demography of rural poverty and the needs of old people and orphans if they are to be useful to HIV/AIDS-affected households.

1.4.3 Insurance

Formal contributory insurance schemes with large commercial companies are out of the reach of most of the rural poor. However, there are various informal and semi-formal mechanisms of insurance that people draw on, and concerns about the sustainability of these in communities affected by HIV/AIDS. Burial societies in South Africa are becoming less viable under the pressure of AIDS-related deaths and the failure of households to make regular contributions. However, there are also arguments that community-based informal insurance mechanisms may be more adaptable and flexible, and thus able to accommodate the changing circumstances of households. In Ethiopia, members of community groups to which people pay subscriptions to meet mourning and funeral costs are being trained for HIV/AIDS-related work (UNOCHA, 2004). Barnett and Blaikie (1992) argue that modifications in customary practice regarding funerals are one community-based response to the HIV/AIDS epidemic. Holzmann and Jørgensen (1999) argue that traditional structures combine insurance functions with other activities and the insurance depends on the trust that arises from other functions. Thus ‘while insurance mechanisms provide insurance, they are guided more by a principle of balanced reciprocity’ (1999: 1015). In the context of stigma, discrimination and growing vulnerability amongst many households in the community, it is easy to see how informal insurance mechanisms and reciprocity can break down.

Table 2 Impacts and appropriateness of various interventions
Types of intervention		Impact on and appropriateness for HIV/AIDS-infected and affected households
Transfers
Cash	Social pensions paid to the elderly can be particularly appropriate because of fungibility and the passing of benefits to other household members. Evidence shows that social pensions in South Africa are often used to pay for children’s schooling and are not as expensive as is sometimes assumed Enable households to buy medicines so that they are less likely to adopt coping strategies that are ultimately destructive (i.e. drawing down on productive household assets in an unsustainable way) Require transparency, accountability and financial and administrative capacity on the part of governments, otherwise are subject to elite capture Amongst donors there may be reluctance to commit resources to recurrent welfare budgets, though the HIV/AIDS pandemic is contributing to a rethink of perspectives Child-headed households may not have the capacity to make good decisions about expenditure, though orphan allowances paid to households may encourage and strengthen community-based care of orphans
Food and nutrition	Viable long-term safety net for households that are severely labour constrained and cannot participate in social protection programmes that have a labour constraint Less viable for households that are not labour constrained because of danger of creating dependency Donors, because of own grain surpluses, are willing to commit large quantities of food Costly, particularly where there is poor transport infrastructure (for example, sub-Saharan Africa)
Inputs programmes	Significantly cheaper than importing food aid Provide seeds and fertilisers to households but are inappropriate for households that are severely labour constrained Could be adapted to provide labour saving technologies to households
School feeding	School feeding can encourage enrolment and reduce drop outs but unlikely to present enough of an incentive to severely labour constrained households, particularly child headed households Take home rations can support OVCs and their households
Public Works Programmes
Cash for work and Food for work	Can be self-targeting, for example when inferior staple foods or lower wages are paid that richer households will not work for Appropriate for HIV-positive but asymptomatic people, but only if they have a rich, healthy diet. Since this is unlikely, FFW and CFW can be counterproductive Inappropriate for labour-constrained households, i.e. those containing people with AIDS and OVCs
Education and training
HIV/AIDS	Training (built in as an element of other programmes) about methods of infection and protection can help to ensure that HIV-negative people remain negative
Nutrition	Through an improved understanding of dietary requirements (supplemented with micro-nutrition programmes), it is possible to reduce the likelihood of infection, lengthen the period of asymptomatic infection and reduce the severity of AIDS-related diseases
Life skills for orphans	Changes in education curricula can leave orphans more equipped to deal with the challenges normally associated with adult life (for example, sanitation, cleanliness and hygiene) Passing on agrobiodiversity and indigenous knowledge will equip orphans with greater skills to grow food to feed themselves
Financial services