TIME FOR CHANGE:
23-24 October 1998
Prevention and Preparedness: Mitigating the Effects of Natural Disasters
This paper was written by Robin Jackson of the
Table of Contents
WFP AND DISASTER MITIGATION
PREPAREDNESS AND PREVENTION ACTIVITIES
Prevention in Practice
Prevention Activities in a Development Context
Prevention in Relief and Recovery
RAPID RESPONSE - A LINK BETWEEN PREPAREDNESS AND PREVENTION
Natural disasters are a major contributor to global food insecurity, particularly in areas prone to drought, flooding and agricultural pest outbreaks. In addition to the loss of life and shelter, the very basis of peoples livelihoods is often undermined, leading to a gradual decline in self-sufficiency and food security. As a result it is important to institutionalize disaster mitigation1 activities in order to buffer the impact of crises on the poorest, and reinforce their capacity to emerge from a crisis with their livelihood system intact. This was recognized by the World Food Summit (1996), which recommended support for disaster prevention and preparedness as one of the priority areas.
Evaluations of humanitarian aid for the African famines in the mid-eighties emphasized that short-term relief interventions act only to save lives, and do little to promote longer-term recovery and improve peoples livelihood options (Anderson and Woodrow, 1991). The two conclusions that emerged were that greater priority should be given to preventive as opposed to curative measures, and that a strong link needed to be forged between relief and development efforts. Development interventions in zones of recurrent disasters that specifically address a populations productive capacities have been identified as an essential component of disaster prevention.
Current thinking within the United Nations follows this reasoning, and the responsibility for a system-wide strategic approach to disaster mitigation has been transferred from the Office for the Coordination of Humanitarian Affairs (OCHA), formerly the Department of Humanitarian Affairs (DHA), to UNDP. It was felt that disaster mitigation activities relate primarily to development assistance. The importance Member States attach to strengthening the capacity of the United Nations system in disaster reduction has been stressed in recent documents and in the formulation of future system-wide options for disaster reduction in the concluding phase of the International Decade for Natural Disaster Reduction (IDNDR). The role of advocating and facilitating system-wide policy-making and cooperation in disaster reduction is highlighted in the United Nations new disposition of responsibilities.
WFPs development programming can be used as an integral part of disaster mitigation strategies by linking activities to the requirements of disaster preparedness and prevention in the relevant development and rehabilitation settings. WFPs mandate, which includes both emergency relief and development assistance, places it in a unique position to approach strategically the issues relevant to disaster mitigation. Appropriate investments in protecting livelihoods of the poor is one way of preventing future crises. Supporting development activities, particularly those in areas where shocks to food systems occur on a regular basis, is a means of supporting more effective emergency interventions, or prevent a crisis from developing.
This paper reviews preparedness and prevention strategies and activities both within WFP and among donors. First, it provides a systematic overview of current practices and thinking from the literature on disaster mitigation. Second, WFPs current prevention activities and programming are reviewed. Lastly, the paper discusses the issue of how to improve the integration of prevention activities in WFP programming so that the complementarity between emergency and development activities is strengthened.
It is important to note that this discussion focuses on prevention and preparedness for periodic natural disasters that cause recurrent shocks to household food security: primarily drought, floods and locust/pest outbreaks. The concentration on natural disasters rather than complex humanitarian crisis is considered appropriate as WFPs direct role in the prevention of conflict is limited. Nonetheless, WFPs role in emergency preparedness (e.g., contingency planning and logistical arrangements) is applicable to food aid needs arising from conflicts.
The rise in emergencies worldwide requires that more attention be focused on disaster preparedness and prevention. However, the shift of funding priorities from development to humanitarian relief during the nineties seems to have negatively affected disaster mitigation efforts. As stated clearly by WFP and confirmed by the literature, "relief assistance alone does not strengthen the capacity of poor people to cope with the next emergency" (WFP, 1994).
According to several recent reports, the extent and severity of human and economic damage caused by natural disasters have actually increased. Both the number of deaths and the economic losses attributable to natural disasters are increasing (OECD/DAC, 1994; IDNDR, 1996). Between 250 and 300 million people are affected annually by natural disasters (IFRC, 1996). People in developing countries are particularly exposed to these effects, owing to their high rates of poverty and to inadequate national capabilities for disaster response and prevention. Often the poorest segments of the population live in areas prone to disasters.
There is universal agreement that natural disaster prevention and preparedness activities in poor, disaster-prone countries are important. However, funding of activities relating to complex or man-made emergencies is increasingly dominating humanitarian budget priorities. This is occurring despite the predominance of natural disasters in causing casualties and damage (SIDA, 1996). The conflict-related emergency focus has diverted attention away from natural disasters, particularly those that recur and are limited to a particular geographical area. As a result, populations living in these areas are becoming increasingly poor and unable cope with shocks to their livelihood systems.2
Additionally, the change of focus has meant that little has been done in areas recovering from civil conflict to promote the prevention of natural disasters. Civil unrest and migration act to destroy a populations capacity to ensure its food security. Thus in the wake of political emergencies, even the smallest disturbance, for example in weather patterns, can produce a natural disaster. This is well illustrated by the present situation in Sudan. The reduced productive capacity of people caught in the midst of a politically based emergency dramatically increases their vulnerability to natural disasters, and limits their ability to recover to their previous level after the effects of the natural phenomenon have disappeared.
Although there is much discussion about the importance of mitigation and prevention, little has been done to focus on strategic ways to prevent the impacts of natural disasters from having devastating effects on the most vulnerable. A long-term perspective is missing as well as the idea that real prevention means early intervention, before the alarms of the early warning systems sound. Currently, prevention activities are more than likely to be short-term investments rather than long-term support to improving peoples capacities to respond to crises. Agencies and NGOs are struggling to define specific interventions that are truly preventive and sustainable, and are trying to find ways to use more development-oriented interventions in situations where in the past relief has been employed.
A review of donor practices and the literature reveals that there are no universally accepted definitions of the terms disaster, mitigation, prevention and preparedness. Moreover, the distinction between terms is often blurred. In one situation an activity may be considered to be an act of preparedness, and in another it is prevention. For this reason, many of the definitions found in the literature are vague and all-encompassing.
There appears to be a confusion in the terminology between emergency and development when activities are implemented very early on in a potential crisis situation to mitigate negative effects. Usually it is not relief but development interventions that are implemented at this stage. However, these are frequently financed with emergency funding. It is no accident that many preparedness and prevention units are housed within humanitarian aid departments.
For the purposes of this paper the following definitions are used and are based on documents relating to natural disasters published by UNDP/OCHA, WFP and the Organization for Economic Co-operation and Development (OECD)/Development Assistance Committee (DAC).
A disaster is a situation where impacts occur on populations that are vulnerable to those impacts: "events which give rise to casualties and/or damage or loss of property, infrastructure, essential services or livelihoods on a scale which is beyond the normal capacity of the affected community to cope with unaided" (UNDP/UNDRO, 1991).
Most experts distinguish between disasters caused by man (e.g., armed conflict, pollution, natural resource degradation) and those caused by nature (hazards arising from droughts, floods, cyclones and geological events). Another distinction is made between disasters that are sudden-onset (where there is little warning or lead time, such as earthquakes and flash floods) and slow-onset (drought, locusts and animal disease epidemics). However, in practice, distinctions between different types of disasters are difficult to make. For example, flooding will often accompany drought. Moreover, natural resource degradation is common in drought-prone areas, with the result that the two characteristics reinforce each other.
The term "complex emergencies" has recently been introduced to describe a humanitarian crisis "where there is a total or considerable breakdown of authority resulting from conflict" and where the international response exceeds the mandates and capacities of any single agency or United Nations Country Programme (IASC, 1994). These different categories often overlap. Mans actions can influence the occurrence and severity of natural hazards, for example, through natural resource degradation, while natural factors (crop failures, water and resource scarcities) can fuel conflict. The DAC Guidelines on Disaster Mitigation (OECD/DAC, 1994) note that while considerable similarities exist in methods to plan and provide humanitarian relief for both types of emergencies, prevention approaches for them differ considerably.
Mitigation includes both prevention and preparedness: "all measures taken to reduce the damage, disruption and casualties" of a hazard. Thus, anything other than pure relief during the acute phase of an emergency can broadly be considered to be mitigation.3 In relation to crop failures resulting from natural disaster, mitigation has been defined as:
identifying areas and sectors of the population most vulnerable to food insecurity;
preparing contingency plans to improve response time;
promoting sustainable livelihoods through support to local production systems; and
introducing viable income-earning opportunities (Frankenberger and Shaw, 1993).
Disaster preparedness is synonymous with "readiness" or measures that allow societies and aid agencies to respond rapidly to emergency needs. This includes:
early warning systems (EWS);
hazard and/or vulnerability assessments;
contingency planning exercises;
planning for enhanced logistical capabilities; and
disaster training and awareness.
As regards drought, substantial improvements have been made in the detection and mapping of the vulnerability both of the local environment (e.g., caused by inappropriate land use or cropping practices) and of different population groups (WFP, 1996a; IFRC, 1997).
Prevention (although the term "disaster reduction" is sometimes preferred) includes structural and non-structural measures to impede the occurrence of a hazard or to reduce the resulting damage and casualties. These include:
using higher or more rigorous engineering and construction standards to build structures (e.g., dams, roads and railways) that will resist severe hazards;
building structures (e.g., flood control dams, dikes and levees; transport infrastructure) specifically aimed at disaster reduction;
undertaking natural resource/environmental management projects (forestry, soil and water conservation) that seek to reduce hazard risks; and
taking measures to improve household food security and local coping strategies, including safety-net instruments.
Based upon a review of the literature and discussions with donors, six broad conclusions or lessons have emerged from the past 20 years of development and emergency experiences.
Lesson 1: Institutional divisions between humanitarian and development programmes result in the neglect of prevention activities.
A review of United Nations agencies and donors revealed that most aid agencies continue to maintain separate programmes for relief and development, or else have a mandate in only one of these areas (UNDP, UNHCR, the World Bank). The institutional separation of relief and development means that relief planning tends not to consider longer-term development objectives, and strategic country development programming may not adequately address natural disaster risks and vulnerabilities. Emergency operations responding to recurrent natural disasters are often carried out in isolation of the countrys ongoing development programme. Despite discussion of the relief-to-development continuum, "the continuum is seldom adhered to in practice" (Finland, 1996). Similar concerns have been raised over the years by both multilateral and bilateral agencies (USAID, 1996b; WFP, 1994; WFP, 1998c). WFP is trying to address this issue in its recovery operations. Designing strategies to incorporate developmental approaches in relief and recovery from both natural and man-made emergencies is one of the main tenets of WFPs new approach to recovery (WFP, 1998c).
In most cases, discrete programmes on disaster preparedness and prevention are funded through the humanitarian relief channels (e.g., ECHO, USAID, Canada and Finland), making it more difficult to take a strategic development approach to prevention measures. Moreover, there appears to be a bias towards preparedness to meet the need for improved, more rapid response in emergencies rather than longer-term prevention or very early interventions in a potential crisis situation. For example, much effort has gone into capacity-building, training, maintenance of infrastructure for food delivery, and development of early warning systems and vulnerability assessments. Substantial progress has been made by UNHCR, UNICEF, WFP and FAO in this area. However, concerns remain that existing early warning methods are insufficient in terms of mobilizing support to prevent major declines in food security.
The incorporation of analyses from hazard and vulnerability assessments and contingency planning exercises into regional or country strategy and project identification is the key to linking preparedness and prevention. Whether the relationship between being prepared and prevention mechanisms is operative and successful is still unclear and does not appear to have been scrutinized in the literature and policy documents. For example, the experience with natural disaster contingency planning has not yet been evaluated in terms of its contribution to countries development strategies. To some extent, this is a result of the fact that agencies have only now begun to adopt contingency planning. In some agencies (CIDA, SIDA, USAID and UNDP), progress appears to have been made on this front: disaster preparedness experts are given explicit responsibility for guiding development strategies.
Lesson 2: If development activities in disaster-prone areas are to have an effect in reducing the impact of disasters on the poor, then they need to be strategically designed to achieve that objective.
|It is a commonly held view in the donor community...that development reduces the
risk of, and vulnerability to, disaster whether natural or man-made....The more the aid,
the lower the risk of disaster....Conventional development aid should be more explicit
about the extent to which it is disaster preventive.
The claim that development assistance mitigates vulnerability to emergencies by focusing on poverty alleviation is only partially true. Conventional development aid may not recognize the erosive effects of repeated shocks to vulnerable households and what measures are needed to reduce their uncertainty and increase their resilience (Ross, Maxwell and Buchanan-Smith, 1994). Moreover, formulation of medium-term economic objectives has generally failed to consider the potential impact of drought-related shocks (Benson and Clay, 1997). Benson and Clay have shown that the risk of drought had not been explicitly considered in structural adjustment programmes in Africa. In over one third of African countries with International Monetary Fund (IMF) stabilization programmes, drought or other natural disasters have an important and perhaps dominant influence on the outcome of these programmes.
Additionally, many development activities are not undertaken in disaster-prone areas as these are marginal and sub-marginal zones with little agricultural potential. Nearly 50 percent of the developing worlds poor live on low-potential, marginal lands (WFP, 1998a). Marginal areas are more vulnerable to environmental degradation (low-lying floodplains, semi-arid and arid zones), and peoples livelihoods are potentially most at risk in the face of a disaster. A much greater share of development assistance is devoted to areas with higher agricultural potential, and more developed transportation and rural infrastructure, including markets. The reason for this is the assumption that investing in areas with relatively high agricultural potential will yield a higher return than the same investment in marginal areas.
Development activities that are undertaken in disaster-prone areas are usually not specifically linked to preventing disasters or improving the capacity of households to cope with shocks. Many development projects in marginal areas have general poverty alleviation objectives. These might or might not help households cope with repeated disasters. More often than not each time a shock or disaster occurs, any development gains accrued are wiped out and communities find themselves starting all over again (UNDP/DHA, 1994).
People living on marginal lands have fewer resources to draw on. Food shortages can mean that populations employ strategies that meet immediate food needs, often at the expense of their natural resource base. Disaster prevention activities that are part of a natural resource management strategy can help to reduce vulnerabilities and hence contribute to preventing future crises. However, this is only possible if it is done with an explicit prevention activity objective in mind.
There exists little or no specific data on development funding devoted to the prevention of natural disasters. Mitigation activities often tend to be small components of broader rural development and poverty alleviation projects. As such, their effectiveness is difficult to assess. Additionally, activities that are explicitly preventive in their objectives are often treated as supplementary activities rather than as part of the core interventions of a given programme. Inherent in this approach is the notion that these activities are optional, to be carried out as time and resources permit (WFP, 1998b).
Although much headway has been made over the last decade by donors, NGOs and national governments in the areas of early warning, satellite imagery, vulnerability mapping and food security information systems, most development activities remain haphazard and piece-meal in terms of their contribution to disaster prevention measures. There appears to have been more work carried out and resources spent on "being ready", than on actual issues of prevention. Development strategies are not conceived with prevention objectives in mind. Many information systems are used as alarm signals for emergency measures, rather than as means to enable interventions to take place early enough to prevent the adoption of damaging coping strategies.
Lesson 3: Progress in reducing the impacts of recurrent natural disasters depends on site-specific, targeted support to local capacities.
|Ultimately, there is no single set of policies or mitigation and relief measures that can be implemented anywhere and at any time to combat the impacts of drought. (Benson and Clay, 1997)|
As has been rightly pointed out, "the practical challenge is not merely to anticipate a famine before it occurs, a daunting task in its own right, but to locate it spatially and socially so as to intervene on behalf of the people who need it most" (Field, 1993). Drought impacts different members of a community differently and prevention needs to take this into account. The vulnerability and responses of disaster-affected groups to diverse shocks are variable and depend on their local resource base, livelihood structures, coping strategies, cultural values and traditions.
To ensure that prevention activities are appropriate and sustainable requires in-depth knowledge of the coping strategies employed by different communities. One problem often cited is that indigenous coping mechanisms may be so time and location specific and so difficult to identify and understand as to be of little use in guiding relief and development programmes (UNDP/DHA, 1994). Another is that practical tools have not yet been developed (e.g., methodologies for capacity assessments, a working definition of capacity-building) to help identify suitable interventions. An overall problem that relates to the development of methodology and other tools is the lack of support for up-front investments for both preparedness and prevention.
Coping strategies are complex and are influenced by many factors. However, if coping strategies and their impact on future production and food security are unknown, it is impossible to implement prevention interventions that address the needs of those most at risk. One means of identifying and understanding coping strategies is by involving the communities in the identification, design and implementation of activities whose specific objective is prevention. Participatory methods have been and are currently used worldwide to identify household responses to shocks and to understand the nature of those shocks vis-ŕ-vis household members. Participatory development offers an effective means to identify ways in which to support vulnerable populations livelihoods and their productive capacity. It enables communities and development professionals to design, implement and monitor measures that prevent irrevocable erosion of the subsistence base of poor households.
Lesson 4: The overriding goal for preventive measures is the preservation of livelihoods.
|....the objective of providing food aid might not be in response to an increase in the number of deaths, but rather to save assets and preserve future livelihoods by providing families with the option of retaining livestock they would otherwise have to slaughterfor food; a timely intervention designed to avert rather than respond to an emergency. (Save the Children/WFP, 1998)|
The literature on mitigation of natural disasters underscores the need to avoid destitution rather than starvation. Rather than waiting for a population to show signs of acute crisis, as in the case of actual starvation, preventive measures are taken before the crisis stage is actually reached. Thus, the choice of saving lives, or saving livelihoods, is fundamentally about the timing of when assistance is provided. This translates into two objectives:
to obtain earlier and more targeted warning of slowly unfolding threats to food security by monitoring local peoples coping strategies; and
to design measures that prevent irrevocable erosion of the subsistence base of poor households (Walker, 1989; Field, 1993).
Coping strategies are used by poor people to maintain their livelihoods in the face of natural disaster and other stress factors.4 They fall roughly into three main categories and correspond to different stages of a potentially famine-producing event: a) precautionary or self-insurance strategies; b) crisis-coping strategies; and c) destitution/distress migration (Corbett, 1988). In reality, these stages are not delineated clearly, and often households will move back and forth between the different phases and use the coping mechanisms of different phases simultaneously. Insurance mechanisms or strategies are those that have been developed in response to exposure to the non-acute risk (for example, populations that are exposed yearly to seasonal food shortages). They include changes in cropping/pasturing practices, gathering of wild food, small reductions in food intake, sale of surplus animals, borrowing money or food from relatives (kinship relationships play an important role), obtaining credit from merchants and diversification of income sources (e.g., day labour). None of these measures threaten the subsistence base of the farmer, or significantly limit future options.
Crisis strategies are those that are developed in response to a severe or unexpected threat to food security. During this stage, animals needed for subsistence are sold, agricultural tools and land are sold, and migration to seek employment occurs. Family members eat less and severe hunger becomes routine. These second stage strategies, characterized by asset stripping, undermine the households means of survival; their appearance marks the true beginnings of famine. If nothing is done to prevent asset stripping, the third stage in coping begins with destitution leading to distress migration in large numbers. This can be accompanied by severe malnutrition and possibly death.
In terms of improving famine response, waiting until the final stage is too late. Interventions must be made between stages 1 and 2, when real stress on the household begins, in order to prevent situations from escalating beyond control. Knowledge of these factors has been incorporated into many risk and vulnerability assessment systems. Despite the general adequacy of such tools for predicting many natural disasters, decisions to intervene are often based upon malnutrition surveys, crop assessments or changes in terms of trade. These decisions are tardy for mobilization of aid, given the lag time in deliveries of supplies (Buchanan-Smith et al., 1993). Moreover, many agencies are questioning the wisdom of using indicators such as malnutrition to launch interventions. If the goal is to avoid destitution, then intervention at an earlier stage is required; nutritional status is not an optimal early warning indicator as its decline comes quite late in the overall sequence of events.
Lesson 5: Livelihood systems are increasingly sensitive to, and less able to recuperate from, vulnerability to recurrent shocks.
|Barriers to re-entry into more secure livelihood systems lock poor people in a constant state of structural vulnerability....Buffers against periods of stress no longer exist....The scene is therefore set for widespread crisis for the next cycle of drought. (Davies, 1996)|
In sub-Saharan Africa, agricultural, agro-pastoral and transhumant livelihood systems have become less able to provide food and income surpluses in good years to cushion against food shortages in dry years (Frankenberger, 1993; Davies, 1996). This decline is based on several factors, including increasing population pressure, exhaustion of soil and other natural resources, greater reliance on markets and a cash economy, increasing frequency of drought and abnormal weather patterns, and unsustainable state policies that negatively affect agricultural production including livestock.
There are two dimensions to vulnerability in cases of successive natural disasters. The first is increased sensitivity, or rather severity of impact, that a shock to the food system would have on a population. The second is reduced resilience, or the declining capacity to recover after a shock (Davies, 1996). Evidence from the Sahel indicates that successive production failures have eliminated the traditional buffers against periods of stress. This has made it harder and harder for populations to adopt secure and sustainable livelihood strategies. Not only have they become more sensitive to the shocks on their livelihood systems, but they are less and less able to return to a pre-shock level. There is little room to manoeuvre in the current livelihood systems; it is a delicate balancing act, with short-term food needs competing against longer-term livelihood protection. Livelihoods have become more diversified as a risk reduction strategy, but diversification is becoming less effective.
Increasing structural vulnerability in areas prone to disaster needs to be addressed directly through interventions that are preventive in nature. These interventions can reduce the impact of natural disasters on peoples livelihood systems, thereby increasing their ability to engage in sustainable livelihood strategies and increasing their capacity to face future shocks to the food system.
Lesson 6: It is less expensive in the long run to invest in mitigation activities than in emergency operations.
It is broadly claimed that support to disaster prevention is justified when considering the enormous cost of natural disasters to affected countries, which is equivalent to all Official Development Assistance (ODA) (Lindahl, 1996; USAID, 1995). Although few studies have investigated the economy-wide impacts of drought in sub-Saharan Africa, or floods in Asia, a recent preliminary study for the World Bank (Benson and Clay, 1997) identified "large, but highly differentiated" economic impacts owing to drought shocks.
Several examples support the argument that long-term development activities that are focused on populations in disaster-prone areas can prevent crises at a lower cost than humanitarian interventions. A continuing subsidy for projects supporting local food production in Mali was deemed cheaper than the cost of supplying famine relief in dry years (Soule et al., 1991). The cost of livestock restocking in Turkana, Kenya, though high, compared favourably with providing free food aid for four years. Food aid was found to halve destitution rates, while restocking virtually eliminated poverty (Kilby, 1993).
WFP analyses suggest that the cost of delivering food in emergency operations is substantially higher than in development projects. A case study on Ethiopias food-based safety nets also suggests that programme costs for these interventions were lower than the costs of comparable emergency operations (WFP, 1998b).
Lastly, the International Food Policy Research Institute (IFPRI) is currently investigating the relative merits of investing in development and relief for USAID. The study simulates the outcome of investing the money spent on drought relief in 1994-95 in agricultural development activities during the recovery period after the earlier drought. Initial findings suggest that households would be no worse off in drought years with this intervention and would have been significantly better off in non-drought years.
Although much has been written on the importance of disaster preparedness and prevention, little has actually been done to measure the cost-effectiveness of preventive interventions. However, practical experience in many disaster-prone areas supports the conclusion that early interventions are more cost-effective than waiting for a crisis to develop. Emergency operations that rely on relief interventions and which are designed as life-saving measures are indeed costly, though necessary. The questions of when to intervene and what are the best, most appropriate activities to achieve real cost-effectiveness are difficult to answer and are key in any mitigation strategy.
WFP AND DISASTER MITIGATION
WFP is involved in disaster mitigation activities in both its development programme and its emergency operations. In fact, the funding window for development is called "Development/Rehabilitation/Disaster Preparedness" and includes "preventive components such as disaster preparedness and mitigation" (WFP, 1996b). The following discussion outlines the principal programming measures WFP uses to carry out its mitigation activities. It attempts to identify WFP's current approach to mitigation for recurrent natural disasters and to highlight areas where WFP could strengthen its interventions. Although much of the discussion below concerns prevention activities, it is recognized that preparedness and prevention are closely intertwined and cannot be easily separated. The following information is based on a review of all Country Strategy Outlines (CSOs) and/or Country Programmes (CPs) approved as of June 1998 covering 38 countries, selected emergency operations (EMOPs), and evaluations of relevant projects.
WFP's current definitions of preparedness and prevention are as follows (WFP, 1991):
Preparedness: "a variety of activities, including the use of vulnerability and risk mapping, information systems and organizational arrangements to improve response capability, and assurance of logistical capabilities to deliver food aid rapidly and efficiently."
Prevention: "reducing the impact of disasters and the likelihood of food scarcities through development projects in disaster-prone areas. For drought or pest-related crop failures, the implementation of integrated, village-level rural development approaches that maintain or improve food production (e.g., improving water supplies, soil fertility and crop storage) is favoured."
WFP activities related to "being prepared'' are undertaken either in the context of a stable development situation or an emergency situation. They are mainly carried out as part of a response to ensure both rapid and effective delivery of food. Moreover, these mechanisms are not confined to natural disasters, but rather to all situations requiring logistical capacity and a rapid response.
Prevention, according to the above definition, has the immediate goal of maintaining or improving food production in the face of a natural disaster. It focuses mainly on the management of the physical environment and the activities necessary to improve the environment's ability to withstand a shock to the food production system. It does not specifically include activities whose objective is to strengthen the social and economic fabric of communities exposed to the shock of natural disasters. Nor does it appear to put emphasis on the systematic inclusion of measures to protect assets and livelihoods.
The underlying assumption of this definition is that a population's food security depends on the ability to produce enough food, rather than the ability to secure adequate access to food, whether through production or by buying it on the market. Thus protection and reinforcement of the physical environment is the optimum choice to reduce hazards. This approach reflects only a partial view of how prevention measures can assist poor households in coping with recurrent natural disasters and illustrates the need to expand the definition.
PREPAREDNESS AND PREVENTION ACTIVITIES
The need to reduce costs and response times has led WFP to develop a "Framework, Preparedness and Response Strategy". The main components include:
Vulnerability Analysis and Mapping (VAM), an integrated system that provides information relevant to food security and contingency planning to improve targeting of both emergency responses and development programming;
the Augmented Logistics Intervention Team for Emergencies (ALITE), which facilitates the exchange of information on logistical capacities and needs and coordinates the relevant aspects of contingency planning. The logistics capacity assessments (LCAs) is part of ALITE and determines potential bottlenecks or inadequacies in existing infrastructure, personnel and equipment for transport, storage and delivery of food;
contingency planning, whereby information from VAM, LCAs and assessments of intervention options is used to develop likely scenarios and establish a process for planning for emergencies; and
diverse funding categories that enable rapid mobilization of limited funds pending approval of EMOPs.
WFPs preparedness activities mainly focus on response mechanisms for emergency situations. These mechanisms fall into two categories: those that have to do with advance knowledge and planning for potential emergencies, and those that relate to inputs for immediate response. Advance knowledge and planning refer to activities such as vulnerability mapping, contingency planning and logistics capacity assessments. Inputs for immediate response include funding, food stocks, human resources, equipment and service packages. Although most activities related to inputs for immediate response are carried out in emergency programming, those that concern advance knowledge and planning are conducted in both development and emergency settings.
Prevention in Practice
WFPs prevention activities include both managing environmental resources and protecting and promoting the livelihoods of poor households. Moreover, WFP's prevention activities are carried out in development, recovery and emergency programming. They can be divided into two categories:
activities that are strategically conceived and planned to contribute to preventing food-security-related disasters before they occur; and
activities that are implemented as a rapid response to prevent potential disasters from escalating or from carrying over into new cropping seasons.
The first type of interventions are primarily funded from the development window, are much more long-term and consist of development activities, principally food for work (FFW). These prevention operations are concerned principally with the management of the physical environment. Most of these activities deal with constructing or restoring rural infrastructure in order to improve a populations ability to weather a hazard, be it drought or floods. Certain natural resource management activities, such as forestry projects, also fall into this category.
The second group of prevention measures are activities that are put in place or undertaken during the early onset of a potential emergency or after the acute crisis stage has abated. These are usually relatively short-term and are frequently financed by emergency funding. They can consist of either development activities, such as food for work, mother and child health (MCH) programmes and vulnerable group feeding, or limited free distribution of food or a combination thereof. Their primary objective is not to manage the physical environment but to support and protect households and their livelihood systems in periods of stress. These interventions are designed so that potential emergencies are contained and the effects of ongoing natural hazards reduced.
Activities implemented in the early phase of a potential emergency come in many forms. Sometimes these activities are an intensification of ongoing longer-term development projects, sometimes they are new projects. Food resources can also be used in support of and within the framework of other ongoing projects.
In some circumstances, activities implemented at the onset of a crisis are called "safety nets". These measures are specifically designed to assist food-insecure households temporarily with direct transfers. The objective of a food-based safety net is to prevent poor people's access to food from temporarily falling below minimum acceptable levels.5 Food-based safety nets are usually institutional arrangements using targeted food-assistance programmes to help populations suffering from shocks, including natural disasters. Populations in disaster-prone areas are thus able to preserve their assets, and hence their livelihoods. Safety-net programmes are generally well integrated with and complementary to other national food security and growth strategies.6
The difference between mitigation operations, either during the early onset of a potential crisis or after an acute crisis has abated, and emergency relief is a fine line. Most practitioners make the distinction between assistance that consists of life-saving and assistance that prevents individuals and communities from sliding into destitution and mass migration, and promotes and preserves livelihoods.
Prevention Activities in a Development Context
The following three categories of activities can be classified as prevention. They are most frequently implemented in development programmes, but are also used in both emergency and recovery contexts. Most of the disaster prevention activities described below are implemented through FFW.
Construction or rehabilitation of infrastructure that directly protects against natural disasters: This category includes small-scale dams and earthworks, coastal embankments, sea dikes, and river training works. It also includes projects to improve small-scale irrigation schemes and rural roads, as the former increase food security and improve the logistical aspects of emergency food delivery. This category of activities is the most common in WFP. In over half of the country programme documents reviewed these measures were mentioned as either basic or supplementary activities.
Soil and water conservation and erosion/desertification control activities: This category includes the planting of trees, windbreaks and shelterbelts to prevent soil erosion, promote dune fixation, fence communal pastures and development of water harvesting techniques.
Village-level food storage facilities or grain banks: Crop losses due to inadequate storage facilities are a major problem in many countries. WFP uses FFW to construct improved storage facilities or to establish grain banks managed by local communities. The grain banks provide low-interest loans for withdrawals in times of pre-harvest scarcity when commodity prices tend to inflate. These activities are the least-used prevention measure. The negative impact of drought on repayments in kind is a generic problem with grain banks in drought-prone regions.
Prevention in Relief and Recovery
WFP intervenes with a variety of activities in the early stages of a potential disaster and in the recovery period. A review of the EMOPs approved to date in 1998 reveals that FFW activities were undertaken in approximately half the emergencies related to natural disasters. Usually, FFW for community infrastructure and agricultural rehabilitation is carried out in combination with free food distribution to the most vulnerable groups. Farmers need seeds to plant their next harvest. In order to keep populations from eating their seeds and from harvesting their crops prematurely after crop failure or a disaster, WFP provides free food aid for a limited time period prior to the harvest. WFP also procures and distributes tools and drought-resistant seeds, in some cases.
WFP funds rehabilitation and reconstruction efforts that effectively prevent emergencies from spilling over into the next harvest year. This is particularly the case with longer-term emergency operations such as those in Somalia, Sudan and the Great Lakes region. FFW in the Great Lakes region is used to increase food production by improving rural infrastructure, terracing arable land, draining swamps and rehabilitating fish ponds. Seed protection and reforestation activities are also planned. In Somalia, where farmers have no seeds to plant the next harvest and flooding has damaged irrigation and drainage canals and access roads, WFP plans to meet the vast need for rehabilitation through short-term FFW interventions. These two-month operations will focus on clearing cropland, improving agricultural productivity, and repairing roads, canals, river embankments and communal storage facilities.
Supplementary feeding through MCH programmes is also used to channel food specifically to malnourished children at times of exceptional food shortfalls, together with the expansion of ongoing food-for-work and school feeding projects. For example, in Malawi, a unique arrangement exists whereby the European Union and USAID are willing to provide additional food as needed to increase core development activities in support of expandable food safety nets in times of sudden food shortages. An indicative budget of 2.5 million dollars a year, to cover food, transport and handling costs is set aside for these activities. The required expansions are determined by an annual assessment, monitoring and appeal processes.
WFP has many FFW projects that are related to the construction and restoration of rural infrastructure. What gives these projects their preventive characteristics is that they were planned with this objective in mind. For example, small dams and roads must be built to withstand specific disaster-related stresses. Moreover, projects must be located in areas which are subject to repeated natural disasters, and they must be designed to incorporate activities that reduce stress on highly vulnerable groups during the onset of disasters or after the acute period of a disaster has passed.
It is critical to incorporate specific policy objectives on disaster prevention, preparedness and mitigation in certain country programming exercises. This involves reviewing past experience with natural disasters (including the strengths and weaknesses of the disaster response) as well as evaluating the likely type, intensity and effects of future events on vulnerable populations and agricultural production. Such information should be an integral part of project identification in country programming (UNDP/DHA, 1994).
Disaster mitigation activities are often part of broader development interventions and strategies in WFP country programming. The focus of many Country Programmes tends to be on small-scale rural development activities and basic infrastructure, which represent the means to achieve disaster mitigation. Although WFP acknowledges the important relationship between natural disasters and food insecurity, disaster mitigation is too infrequently identified as a specific programme objective. For example, although drought recurs every three or four years in Mauritania, and the basic activity of the Country Programme is designated as small-scale rural development, mitigation or prevention are not mentioned as specific objectives of this project.
Strategically linking the phenomenon of recurrent natural disasters to country programming of development activities needs to be strengthened. The statement that development projects planned for drought-prone areas in India are "inherently drought and emergency mitigating" encapsulates the assumptions made in most country programming exercises (WFP, 1997). Because a development project is located in a disaster-prone area, it must have a mitigating effect when a hazard strikes. Although simply "being there" does indeed have a positive effect if a natural disaster occurs, the diversity of the situations faced by poor households in disaster-prone areas, and the dynamic nature of natural disasters calls for the incorporation of specific mitigation objectives in order to improve the effectiveness of prevention efforts.
Too often operations are planned without reviewing past experience, including the effects of previous disaster responses. Moreover, more analysis is needed to assess the intensity and effects of disasters on vulnerable communities. Key questions to be answered are: a) how early to intervene; b) when early intervention is appropriate; and c) what are the most effective types of interventions or activities, given the situation.
In countries prone to recurring natural disasters, further work needs to be done in order to link development activities and emergency interventions effectively. A step has been taken in this direction with the new Protracted Relief and Recovery guidelines, which emphasize strategic planning for prevention in recovery situations: "Because crises resulting from natural phenomena and resource degradation affect both immediate food consumption and longer-term food production, longer-term approaches are required; responses which include prevention, preparedness and early warning are required" (WFP, 1998c). WFP needs to reflect on how to develop a framework to reinforce the complementarity between emergency and development interventions in terms of reducing the impact of recurrent natural hazards. Such linkages could strengthen WFP's capacity to help those affected by recurring natural disasters.
RAPID RESPONSEA LINK BETWEEN PREPAREDNESS AND PREVENTION
WFP has many mechanisms that come under the heading of "being prepared" and as an institution excels in rapid response. Many of these concern logistics, especially the transportation, storage and delivery of commodities. Some of the more common methods of getting resources in place rapidly are to borrow from the national reserve for immediate distribution pending arrival of commodities. In some countries WFP has a special agreement with the government to use national, regional and local food stocks or reserves. Food stocks can also be pre-positioned. In other countries, national food funds exist, though they are not very common. These funds "constitute a strategic food reserve to respond efficiently . . . to immediate needs for food resources in cases of disaster" (WFP, 1998d).
To ensure rapid response in countries where there are ongoing development activities, WFP often uses the resources from these activities as a springboard for early intervention. "Borrowing" commodities from development projects and in some cases PROs7 is a common means of handling smaller-scale operations at the early onset of a potential crisis. This is the fastest way for WFP to "kickstart" an operation, as the commodities are often already stored nearby or en route. Borrowing can occur within a country or between different countries, and is mainly a measure to ensure a rapid response, thereby fulfilling a key element of early intervention.
Another means of intervening quickly when a potential emergency strikes is to expand existing development projects. Existing FFW or vulnerable group feeding operations may be scaled up to accommodate additional food deliveries in times of need. The main advantage of this mechanism is that by using existing development projects, institutional and logistical infrastructure is already in place, making the delivery of additional resources relatively easy. Scaling up an existing project can take place in two ways:
the same area as the original project is targeted but rations are increased, so that the project is intensified; and
the number of people covered by the project is increased and hence the geographic coverage is expanded.
The existence of ongoing projects often influences decisions on what type of intervention should be carried out. This signals the importance of designing Country Programmes which explicitly link development activities to prevention goals, in both development and emergency operations. More advance planning needs to be carried out to identify suitable areas, target populations and appropriate activities for expansion. The success of mitigation strategies and operations depends on matching the type of work planned with the emergency event, and the vulnerability of the population.
As important as rapid response is, in situations where there is a large degree of predictability, such as in natural disasters, the planning of not only how to get food there quickly, but also the best use of this resource in protecting and supporting livelihoods is of fundamental importance. As part of its rapid response mechanisms, WFP needs to incorporate more analysis of how resources are to be used to best ensure that vulnerable populations withstand disasters and that livelihoods are maintained.
A key element in the effectiveness of interventions designed to prevent potential disasters from escalating or continuing into new cropping seasons is timing. To meet their objectives these interventions need to occur early enough to protect hard-won development gains from being destroyed, and to prevent serious suffering. Early intervention is particularly important in areas of recurrent natural disasters where reduced coping capacities among vulnerable groups can exacerbate the impact of even a moderate hazard. Moreover, appropriate interventions to aid recovery can help populations already severely handicapped by previous emergencies cope better with small and moderate hazards.
Appropriate timing needs to be combined with appropriate measures. Protecting a household's livelihood requires interventions that are specifically designed for that purposesingle-solution prescriptions are inadequate. Moreover, the dynamic nature of recurrent disasters means that different measures should be undertaken at different stages. For example, an area that has experienced moderate shocks to its food security for two or three consecutive years may require a different type of intervention than an area that has undergone a more severe shock, but only for one year.
The necessity of developing a strategy to incorporate developmental approaches in relief and recovery, for both natural and man-made emergencies, was addressed in the policy document From Crisis to Recovery (WFP, 1998c). Based on this policy, country recovery strategies will be prepared (where appropriate) which describe suitable recovery interventions and funding proposals in the context of a country programme or framework.
These developments illustrate that improvements are being made by WFP in mitigation and in promoting longer-term development objectives in relief and recovery. More progress is still needed in analysing the risks and dimensions of food shortages caused by natural disasters. The challenge is to determine which measures truly prevent inexorable declines in households livelihoods from recurrent shocks. Emergency responses to recurrent drought and flooding, especially when they occur in localized pockets, are often ad hoc, and depend on ongoing development activities as a means to secure adequate commodities, or to use existing delivery systems. Emergency and development interventions concerning recurrent natural disasters need to be strategically linked with complementary objectives. Analysing the needs of populations in disaster-prone areas, together with past performance and the effects of disaster response, is imperative if mitigation and not just a rapid response is to be carried out.
While development funding can be extended somewhat as a rapid response to a worsening situation, the room for manoeuvre is relatively small. With recurrent natural disasters WFP is faced with a fluid situation and a relatively inflexible funding mechanism. As WFP development programming is intended to include prevention activities, the funding should reflect the flexibility needed to respond quickly to recurrent natural disasters. At present, if a rapid, short-term response is needed to reduce the effects of a disaster, emergency financing is usually used, even though the character of many of these interventions is developmental.
Lastly, the issue of when an intervention is early enough needs to be addressed. How early should an intervention be in order to have maximum impact on reducing the effects of a potential disaster? Furthermore, questions of timing are inextricably linked with targeting. Poor, food-insecure households located in disaster-prone areas are often highly vulnerable to shocks such as drought, pests and floods. How their vulnerability is measured, and hence what signals need to be monitored to ensure that the intervention has a prevention effect are critical in determining when and how to intervene. Many of these issues are site specific and depend on the type of hazard, the depth of vulnerability of the population and the effects of the disaster on the ability of households to secure adequate food for their members.
Disaster mitigation needs to become a more explicit objective in the country planning/programming process and accordingly basic activities need to be designed with a greater emphasis on attaining this objective. Development activities in disaster-prone areas should be reinforced through a more focused prevention approach.
The Country Programme approach needs to be used to link strategically development activities in relief, recovery and development programmes to prevention and preparedness goals in the case of natural disasters. More emphasis needs to be placed on developing complementarities between activities in disaster-prone areas, in order to protect and promote livelihoods.
A number of tools are currently used for prevention of food insecurity caused by natural phenomena but guidance is needed on which tools are most appropriate, at what time and for which groups of a community or parts of a region.
The incorporation of information and analysis derived from VAM and contingency planning exercises in Country Programme Documents can be strengthened. VAM provides a valuable source of information on the origins of communities food insecurity, and has the potential to shed light on different coping strategies and household responses to recurrent hazards. The use of VAM in disaster management and prevention is therefore crucial.
Analysing the reasons behind populations food insecurities and the ways in which they respond to recurrent shocks allows WFP to determine appropriate interventions. This analysis also needs to incorporate whether food-assisted programmes are sufficient or whether WFP needs to partner with other agencies and organizations to carry out prevention objectives fully.
The question of how to establish when early intervention is required should be addressed. The use of VAM and other early warning information systems will help to identify appropriate timing and suitable interventions to protect livelihoods. But a policy of early intervention, and what that entails needs to be established.
1 For the purposes of this paper, the term 'mitigation' includes both preparedness and prevention activities. 2 Livelihoods consist of a range of on-farm and off-farm activities, which together provide a variety of procurement strategies for a households food and cash needs. A households total resources are based on productive assets and endowments, and also on its legal, political and social position within the society. Livelihood systems imply a concept of sustainable food security, where the benefits of today are balanced with the benefits of tomorrow (WFP, 1998e). 3 As an example of differing definitions, Oxfam views mitigation as any action to reduce vulnerability, whether implemented before, during or after an emergency. FINNIDA defines mitigation as efforts to support coping strategies, preparedness as the ability to anticipate future emergencies, and prevention as social/political measures to avoid the outbreak of conflict. 4 For an in-depth discussion of coping strategies and livelihood systems, see "Food Security, Livelihoods and Food Aid Interventions", WFP, 1998. 5 For a detailed discussion of food-based safety nets, see "Food-based Safety Nets and WFP", Rome , 1998. 6 It is important to note that safety nets are not only used in the case of natural disasters. They are designed to protect vulnerable populations from a variety of shocks, including economic (sharp increases in the price of agricultural inputs or declines in agricultural output prices; job losses, etc.) and structural adjustment shocks (higher food prices and greater food price volatility). (WFP, 1998b). 7 In the future, the mechanism will be the protracted relief and recovery operation (PRRO).
Adams, M., 1986. Merging Relief and Development: The Case of Turkana. Development Policy Review 4:314-324.
Anderson, M., and Woodrow, P., 1989. Rising from the Ashes: Development Strategies in Times of Disaster. Westview Press, Boulder.
Anderson, M., and Woodrow, P., 1991. Reducing Vulnerability to Drought and Famine: Developmental Approaches to Relief. Disasters 15: 43-54.
Benson, C., and Clay, E., 1997. The Impact of Drought on Sub-Saharan Economies - A Preliminary Investigation. Technical Paper No. 401. World Bank, Washington, DC.
Buchanan-Smith, M., Davies, S., and Petty, C., 1993. The Use and Abuse of Early Warning Information. In A Special Issue on Disaster Mitigation, Arid Lands Newsletter 34: 16-21.
Corbett, J., 1988. Famine and Household Coping Strategies. World Development 16(9):1099-1112.
Davies, S., 1996. Adaptable Livelihoods: Coping with Food Insecurity in the Malian Sahel. Science, Technology & Development 14(1):144-156.
Frankenberger, T.R., 1993. Promoting Sustainable Livelihoods in Areas Prone to Recurrent Drought. In A Special Issue on Disaster Mitigation, Arid Lands Newsletter 34: 30-35.
Frankenberger, T.R., and Shaw, D.J., 1993. Mitigation in the Emergency-Development Continuum . In A Special Issue on Disaster Mitigation, Arid Lands Newsletter 34: 2-5.
Field, J.O., 1993. Beyond Relief - Toward Improved Management of Famine. In Field, J.O., (ed.), The Challenge of Famine - Recent Experience, Lessons Learned. Kumarian Press, West Hartford, CT.
Finland, Ministry for Foreign Affairs, 1996. Thematic Evaluation of the Finnish Humanitarian Assistance (Evaluation report by COWIconsult). Helsinki.
Finland, Ministry for Foreign Affairs, 1993. Finlands Development Co-operation in the 1990s. Helsinki.
IASC, 1994. Working Paper on the Definition of Complex Emergencies. Proceedings of meeting held in Geneva, 9 December, 1994.
IDNDR, 1996. Reducing Exposure to Natural Disasters: The International Decade for Natural Disaster Reduction. Presentation by Jeggle, T. at the Symposium "Embrace the Future" held in Phoenix, AZ, 5 - 8 May, 1996.
IFRC, 1996. World Disasters Report 1996. Geneva.
IFRC, 1997. World Disasters Report 1997. Geneva.
Kilby, P., 1993. Emergency Relief Programmes for Pastoral Communities. Development in Practice 3: 92-102.
Kloos, H., and Lindtjorn, B., 1994. Malnutrition and Mortality During Recent Famines in Ethiopia: Implications for Food Aid and Rehabilitation. Disasters 18: 130-139.
Kreimer, A., 1991. Lessons on Emergency Recovery from Past Experience. Land Use Policy 8: 310-323.
Lindahl, C., 1996. Developmental Relief? An Issues Paper and an Annotated Bibliography on Linking Relief and Development. SIDA Studies in Evaluation 96/3. Stockholm.
Morgan, R., 1986. From Drought Relief to Post-disaster Recovery: The Case of Botswana. Disasters 10: 30-34.
OECD/DAC, 1994. Disaster Mitigation Guidelines. Paris.
Oxfam, 1995. The Oxfam Handbook of Development and Relief. Oxford (3 Vols.).
Ross, J., Maxwell, S., and Buchanan-Smith, B., 1994. Linking Relief and Development. IDS Discussion paper 344, Brighton, Sussex.
Save the Children/WFP, 1998. An Introduction to the Food Economies of Southern Sudan. Nairobi.
Schilderman, T., 1993. Disasters and Development. A Case Study From Peru. Journal of International Development 5: 415-423.
SIDA, 1996. Annual Report: Swedish Humanitarian Assistance. Stockholm.
Siegel, S., and Witham, P., 1991. UNDP Coordination of Disaster and Development Planning. Land Use Policy 8: 297-309.
Soule, M., Nelson, G. and Due, J., 1991. Is There an Alternative to Famine Relief? An Example from the Sahel. Food Policy 16: 235-244.
UNICEF, 1997. UNICEF 1997 Annual Report. New York.
UNICEF, 1998. The State of the Worlds Children 1998. New York.
UNDP/DHA, 1995. Module on Disaster and the Environment. Disaster Management Training Programme. Geneva.
UNDP/DHA, 1994. Modules on Disaster Preparedness, Disaster Mitigation, Disasters and Development, and Drought and Famine. Disaster Management Training Programme. Geneva.
USAID, 1995. Humanitarian Response and Post-crisis Transition. Washington, DC.
USAID, 1996a. Linking Relief and Development in the Greater Horn of Africa - USAID Constraints and Recommendations. Memorandum. Washington, DC.
USAID, 1996b. Evaluation of USAID/BHRs Grant to the World Food Programme for Disaster Mitigation. Washington, DC.
UNDP/UNDRO, 1991. Disaster Management Manual. Geneva.
Villumstad, S., and Hendrie, B., 1993. New Policy Directions in Disaster Preparedness and Response in Ethiopia. Disasters 17: 122-133.
Walker, P., 1989. Famine Early Warning Systems: Victims and Destitution. Earthscan, London.
Wickramanayake, E., 1995. Rehabilitation Planning for Flood Affected Areas of Thailand: Experience from Phipun District. Disasters 19: 348-355.
WFP, 1991. Food Aid in Emergencies. Rome.
WFP, 1994. Annual Report 1994 - Linking Relief and Development. Rome.
WFP, 1996a. WFP in Emergencies - Framework, Preparedness and Response Strategy. Rome.
WFP, 1996b. Implementation of WFP Resource Policies and Long-Term Financinga Guide for Donors. Rome.
WFP, 1997. Country ProgrammeIndia (1997-2001). WFP/EB.1/97/5/Add.1. Rome.
WFP, 1998a. Improving Food Security in Marginal, Low-Potential Areas. Rome.
WFP, 1998b. Food-Based Safety Nets and WFP. Rome.
WFP, 1998c. From Crisis to Recovery. WFP/EB.A/98/4-A. Rome.
WFP, 1998d. Country Strategy Outline - Ecuador (1999-2003). WFP/EB.2/98/4/5. Rome.
WFP, 1998e. Food Security, Livelihoods and Food Aid Interventions. Rome.
WFP, 1998f. Annual Report of the Executive Director, "Recovery From Crisis". WFP/EB.A/98/3-A. Rome.
WFP, 1997. An Initial Framework for Drought Contingency Planning for the World Food Programme. Southern Africa Regional Office, Maputo.