Research Fellow, HPG
Researched, written and published by the Humanitarian Policy Group at ODI Paul Harvey Research Fellow, HPG Executive summary This discussion paper examines the use of cash and vouchers to provide people with assistance in emergency situations. The first product of an ongoing research project by the Humanitarian Policy group (HPG), it is based on a critical review of existing published and grey literature, initial discussions with aid agency staff and a survey of project documentation from recent and ongoing cash- and voucher-based responses.
Proponents of cash- and voucher-based approaches argue that they can be more cost-effective and timely, allow recipients greater choice and dignity, and have beneficial knock-on effects on local economic activity. Sceptics fear that they are often impractical because they incur additional risks of insecurity and corruption, and argue that cash may be more difficult to target than commodities. Even where these approaches are feasible, there are concerns that women may be excluded, that cash may be misused by recipients and that it may have negative effects on local economies and could fuel conflicts. Others feel that cash- or voucher-based responses sound interesting, but that in practice commodities are what is available, and what relief agencies have the skills and experience to deliver.
Crudely put, our literature review has thrown up two main findings. The first is that cash and voucher approaches remain largely under-utilised in the humanitarian sector. The second finding is that there is a growing amount of experience with cash and voucher approaches, and that the absolute dominance of commodity-based approaches is beginning to erode. This growing experience is mirrored in the development sector, where various types of cash transfer have been used in the area of social protection and safety nets. Of course, in the West cash has long been used for both long-term welfare payments and in response to emergencies, both through the insurance system and as part of relief action.
Literature and theory
What is being covered?
A confusing array of interventions fall under the general heading of 'cash and voucher responses'. This paper focuses on cash grants, cash for work and voucher programmes where the cash or voucher is given to individual households, not to communities or governments. Interventions such as monetisation, microfinance, insurance, budget support and waivers for school fees are beyond the scope of this study. Cash and vouchers have usually been considered as alternatives to food aid, but other types of intervention can also be supported through the provision of cash. Cash or vouchers should be considered as alternatives to agricultural inputs, shelter and non-food items, as well as an alternative to food aid.
Humanitarian relief is delivered in a huge range of contexts, from natural disasters to wars, in rich developed countries and poor developing ones. This study attempts to address the question of where cash and vouchers are suitable in the full range of emergency contexts. Sometimes it is assumed that cash may be possible in relatively well-developed countries with banking systems, but not in less developed contexts, or that cash can be used in peaceful contexts but not in complex emergencies. The experience reviewed here challenges these assumptions, suggesting that cash or vouchers can be possibilities even amid state collapse or conflict, or where there is no banking system. Clearly, however, some of these factors make implementing a cash- or voucher-based response more difficult.
The economics of cash
The literature in this area often focuses on the advantages and disadvantages of cash. The main arguments are summarised in the table below. There is a need to emphasise that both columns summarise the hypothetical or possible pros and cons; as we will see in subsequent sections, some of the theoretical fears about the drawbacks of cash have not been borne out in practice. The existing evidence, for example, suggests that people rarely use cash for anti-social purposes, and that women are not necessarily particularly disadvantaged by the use of cash rather than in-kind approaches.
Vouchers are often used when cash is not seen as possible or appropriate. This may be due to donor constraints, to a desire to ensure that a particular type of good or commodity is purchased by the recipients, because of security fears about the use of cash or because of market weaknesses. Vouchers can be exchanged to purchase commodities from traders, at distribution outlets, markets or special relief shops.
By far the largest experience to date with voucher programmes is in the provision of seeds and other forms of agricultural inputs. Vouchers have also been used in other contexts. A large ICRC programme in the occupied Palestinian territories in 2002-2003, for example, distributed vouchers that could be exchanged for food and other basic goods. Other instances include the UK government's response to the Montserrat volcano eruption in 1996 and Save the Children's voucher programme in northern Iraq in the early 1990s. Voucher programmes may require more planning and preparation than the distribution of cash; agreements need to be reached with local traders, for example, and 'seed fairs' at which vouchers can be exchanged take time to set up. If vouchers are not providing goods that people see as priorities then a parallel market may well develop, with vouchers being traded for cash at a discounted price.
Evaluations comparing vouchers and commodity approaches have been broadly positive, emphasising that they give people more choice and can have positive effects on local markets. Where voucher approaches have been compared to cash,
however, questions have been raised about whether the additional administrative burden that managing a voucher programme imposes for the implementing agency is worthwhile. Donor constraints and reservations about cash seem to play an important role in discouraging agencies from switching from vouchers to cash, even where this might be appropriate. There may, however, be situations in which voucher approaches are more appropriate than cash: when cash raises particular security difficulties which vouchers would not, where there is a need to restrict support to a particular type of commodity or where markets have been weakened and need revitalising.
Cash transfers in the development literature
Within the development literature, there is a growing body of experience and documentation of successful examples of cash transfers, particularly in Central America. A key feature of these programmes is that they were accompanied by systematic efforts to measure their effectiveness and impact. This is perhaps one of the key lessons for the humanitarian sector. If cash and voucher approaches are to be accepted, there is probably a need for similarly systematic attempts to assess their impact. The results of evaluations of these cash-transfer programmes have been broadly positive, showing measurable improvements to school enrolment rates, reductions in rates of child labour and improvements in child health and nutrition and in levels of consumption.
Renewed interest in long-term welfare safety nets and social protection may also provide opportunities for reinvigorating the debate around linking relief and development. There is an emerging consensus within the development community around the need to pay greater attention to the basic welfare requirements of populations living in difficult environments. There might be opportunities both for welfare safety nets to be expanded during periods of crisis to help people to deal with shocks, and for cash transfers that begin as emergency interventions to be developed into longer-term social protection programmes.
Assessment and appropriateness
It often seems that aid agencies are reluctant to consider cash because of concerns about its appropriateness, because agency policies or staff skills preclude it or because funding for cash or voucher approaches is not available. Getting cash and vouchers onto the humanitarian agenda and into the humanitarian toolbox would mean moving away from resource-driven assessments. As a first step, it would be encouraging to see agencies explicitly considering a range of intervention options as part of the assessment process.
Issues around the appropriateness of cash divide fairly neatly into two categories: practical questions around its feasibility, and economic questions around the ability of local markets to respond. In order to make judgements about the economics of cash and voucher responses, agencies therefore need to improve their capacity to assess local markets. The tools to do this already exist; the challenge is getting these tools into manuals and standard assessment checklists, and making market analysis a routine part of the assessment process.
Impact and cost-effectiveness
The existing documentation of cash- and voucher-based responses shows that they are overwhelmingly successful in terms of their impact. People spend the money they are given sensibly, cash projects have not generally resulted in sustained price rises and women have been able to participate, and have a say in how cash is spent. Cash responses have usually been found to be more cost-effective than commodity-based alternatives. The body of experience that these conclusions are drawn from is still small and there is a need for caution. There is still only limited evidence about the likelihood of inflationary impacts if cash and voucher projects were to be implemented on a larger scale. There is also only limited evidence about their feasibility in complex emergencies. What experience there is, however, strongly suggests, however, a case for the further development of cash- and voucher-based approaches and for piloting their application on a larger scale.
One of the central reasons for caution in adopting cash- or voucher-based responses relate to insecurity and corruption. In many of the contexts in which humanitarian agencies work, there are clear concerns about putting cash into conflicts and predatory political economies. However, evidence from existing projects suggests that ways can be found to deliver and distribute cash safely even in conflict environments; indeed, in some situations cash has been less prone to diversion than alternatives. The tendency to assume that cash is a priori more vulnerable to looting or diversion perhaps needs to be examined. Cash is both highly portable and not necessarily as visible as large-scale commodity distributions. This suggests that, in some circumstances, cash could be safer for both staff and recipients.
Innovative ways have been found to minimise the risks of insecurity and corruption, and evaluations have found little evidence of insecurity or corruption relating to cash-based approaches. Since much of this evidence is context-specific, one of the generic lessons is probably the unsurprising point that there is a need for a locally nuanced understanding of particular security risks. For example, in Afghanistan and Somalia it has been possible to use the local hawala (moneytransfer) system to distribute cash. In Ethiopia, Save the Children take out insurance coverage against losses in transporting cash to projects in areas where there are no banks. Recent technological advances may provide further options for minimising corruption in all types of distributions, including cash. UNHCR's use of iris-recognition technology in repatriation from Afghanistan to Pakistan is thought to have greatly reduced the risk of people claiming multiple repatriation grants by moving to and fro over the border. In Bam, Iran, the government has simply set up bank accounts for all recipients and transfers cash directly into them.
There is a risk that cash, particularly in the form of grants, could be more difficult to target and more prone to diversion. Cash distributions also bring with them the particular risk that those not targeted could be worse off if prices rise following a cash distribution. The cash and voucher experiences reviewed for this paper, however, did not suggest that targeting cash or vouchers was significantly more difficult than targeting commodity-based approaches. Targeting was sometimes problematic, but no more so than is usually the case.
The project experience collated for this review suggests that the operational challenges faced in implementing cash and voucher programmes in emergencies are surmountable. Justifiable fears of insecurity and the risk of diversion and corruption can be overcome and recipients can be successfully targeted. Proponents of cash and voucher-based responses also argue that cash can be an intrinsically more dignified way to provide assistance. Recipients of cash tend to prefer it to alternatives because of the greater flexibility and choice that it provides.
Institutional and organisational barriers
Given the arguments in favour of cash-based responses, why have agencies remained so reluctant to use them? The way in which the architecture of the humanitarian system is currently structured seems to inhibit consideration of cash and voucher responses. In the UN system, in particular the consolidated appeals, the almost complete absence of cash or voucher-based approaches suggests that cash is not being seen as an option in part because the dominant operational agency is mandated to provide food. There is also a wider debate about the dominance of food aid in current humanitarian responses and the extent to which this is due to the continued tying of aid to food surpluses in donor countries.
This raises a further set of questions around the responsibilities of different actors. Do donors have a responsibility to provide the most appropriate resources for meeting the needs identified in emergencies? National governments in the countries affected by disaster also have responsibilities in this respect. In 2002, the Afghan government tried to shift from food aid deliveries to a cashbased, labour-intensive public works programme. Aid agencies also have a responsibility to make a case to donors for providing appropriate resources. Outside of the UN system, there seem to be fewer barriers to considering cash and voucher responses, and NGOs and the Red Cross have led the way in their increasing use.
This leads us into the question of whether aid agencies currently have the skills and expertise to implement cash and voucher approaches. The answer to this at the moment is that these skills are in short supply. Existing guidelines and manuals often do not include cash. However, a growing number of practitioners are developing experience with cash- and voucherbased responses, and this will be taken into new assignments and organisations. Manuals are also starting to be developed.
Finally, there is the wider issue of the attitudes and assumptions that humanitarian aid practitioners have towards the people that they are trying to help. Cash seems to inspire a reluctance on the part of aid agencies that goes beyond practical fears about security or its inflationary effects. There is a sense in which cash is threatening. Partly, this is about a loss of control; giving people money involves a transfer of choice from the agency to the affected population. The widespread assumption that people will misuse cash, for example, hints at the feelings of superiority which sometimes underpin relations with the people agencies label as 'beneficiaries', a term which itself suggests the passive receipt of assistance. These questions are rarely openly acknowledged or discussed, but they nonetheless play an important part in shaping the way in which humanitarians relate to the people they seek to help.
A strong body of evidence is beginning to emerge to the effect that providing people with cash or vouchers in a wide range of emergency situations works. It is possible to target and distribute cash safely, people overwhelmingly spend money on basic essentials, and cash provides a stimulus to local economies and is often more cost-effective than commoditybased alternatives. The evidence also suggests that, in some situations, cash may complement commodities. Cash should not necessarily therefore be seen as a replacement for other forms of aid, but as an additional instrument.
There is therefore a strong case for investing further in the rigorous evaluation and documentation of cash and voucherbased responses, in order to be able to make better-informed judgements about their impact. There is also a need for humanitarian practitioners to develop the skills and capacities they need to implement cash and voucher interventions, and for the development of a body of practice and guidelines.
The recommendations below provide a starting-point for thinking about how greater use of cash and vouchers in emergency response could be taken forward.
- Assessment capacity should be developed. This should be either independent or robust enough to stand up to external analysis, made public and less driven by considerations of resources.
- Aid agencies should be able to consider cash or vouchers as alternatives to commodity-based approaches as part of the assessment process.
- Investment is needed in the skills and capacity to assess markets at local, national and regional levels.
Evaluation, learning and expertise
- Further investment is needed in rigorous evaluation and documentation of cash- and voucher-based responses, in order to be able to make a clear case about their impact and effectiveness, and when and where they are appropriate.
- Investment is needed in further learning and training to equip those involved in assessments and programme management to assess the possible appropriateness of cash and voucher responses, and to implement them where appropriate.
- A documented body of practice and practical guidelines on cash and voucher responses should be developed for staff involved in emergency responses.
- There needs to be a greater willingness to examine attitudes of paternalism and superiority on the part of aid practitioners, and to overcome these at individual and organisational levels.
- As part of reform to the UN system, consideration should be given to where responsibility for implementing cash-based responses to food insecurity should lie, to enable cash and voucher responses to be included in the consolidated appeal process.
- Aid should be untied, and donors should endeavour to provide the resources identified as most appropriate, including cash.
Links with social protection
- There is a need for investigation into ways to link emergency response more closely with emerging social protection systems, which increasingly have a cash-based component.
These recommendations have potentially far reaching consequences for the ways in which humanitarian relief in emergencies is managed and delivered. It implies the likely expansion of cash based programming, probably at the expense of in-kind mechanisms in some contexts. This would require the development of additional skills within aid agencies to assess when and where cash based responses are appropriate and implement them where cash is appropriate. Donors will also need to develop the skills and capacity to make informed decisions about whether or not to fund cash based responses.
More fundamentally, it suggests a need to examine the current architecture of humanitarian responses and in particular the mandates of the main operational UN agencies. Food aid currently dominates the international relief response to emergencies and this paper argues that both agencies and donors need to re-examine existing food aid policies and take a hard look at the appropriateness of food aid in many contexts. Although cash will be complementary to food aid in some circumstances, in others it is likely to be an alternative to it, and this suggests a tighter and narrower role for food aid in the response to emergencies. 2005 will be a key year for the future of food aid, with a new Food Aid Convention and a new WTO Agriculture Agreement being negotiated, enabling these issues to be raised. Finally, there is a need to explore further possible linkages between emergency response and social protection and to examine what this means for the use of cash and in-kind mechanisms as well the implications it might have for relative involvement of international, national and local actors in responding to emergencies.