APPENDIX. THE MAIN EARLY WARNING SYSTEMS

1. GIEWS, THE FAO VERSION

The Global Information and Early Warning System (GIEWS), created by the FAO in 1975, has been co-managed by the WFP since 1991. GIEWS is concerned with food security at national level. It measures production/imports against consumption/exports, and highlights localized pockets of food insecurity. The emphasis is on the availability of food stocks rather than their accessibility.

Based in Rome, GIEWS relies on information supplied by governments, NGOs and local FAO representatives. It produces regular reports as well as annual assessments. GIEWS data is often used to initiate the first rounds of the aid appeal, before the WFP has carried out assessments used for distribution planning and beneficiary identification. GIEWS is regarded as a “key source of information, and is often described as the gold standard” (Buchanan-Smith & Davies, 1995). Donors tend to put great faith in it, given its status as a critic of local EWS.

Because GIEWS data relies on secondary information sources within local governments, it is generally no better in quality than the data provided by national EWS. However, international agencies have more faith in the FAO’s system; its compilations of quantitative production estimates are regarded as authoritative and as essential for comparing situations in various parts of the world. The information emanating from GIEWS, expressed simply and directly, facilitates negotiation and decision-making (Buchanan-Smith & Davies, 1995; Darcy & Hofman 2003).

2. NATIONAL AND REGIONAL EWS

• USAID is the only donor organization with its own EWS, the Famine early warning system (FEWS). Established in 1985, FEWS exists primarily to provide USAID decision-makers with information. FEWS, with its extensive use of satellite images and computer modeling, is seen as a ‘high-tech’ system. It publishes regular information bulletins but makes no recommendations, leaving decisions to the parent organization. FEWS information is widely used in Washington by USAID, and sometimes by the US Congress (Buchanan-Smith & Davies, 1995; Darcy & Hofman, 2003). USAID downplays the importance of United Nations assessments, which are felt to lack credibility because of their association with recipient governments. However, FEWS warnings rely just as much on secondhand data, which is simply analysed and then presented in a concise, highly visual form (Buchanan-Smith & Davies, 1995).

• The European Union has no EWS dedicated to famine response, but relies on the European Food Security Network (RESAL), which consists of a network of NGOs, research institutes and other organizations. The EU also funds remote sensing projects (SCOT, PUMA).

• The 1990s saw the emergence of regional EWS designed to coordinate national early warning systems. The Southern African Development Community SADC has 14 member countries(SADC) links the EWS operating in southern Africa. The Permanent Inter-State Committee for Drought Control in the Sahel CILSS has nine member countries: Burkina Faso, Cape Verde, Gambia, Guinea-Bissau, Chad, Mali, Niger, Mauritania and Senegal. (CILSS) does the same for the Sahel region, while the Intergovernmental Authority on Development  IGAD has seven members: Kenya, Uganda, Somalia, Sudan, Ethiopia, Eritrea and Djibouti.(IGAD) coordinates EWS in east Africa and the Horn of Africa. CILSS has established an Integrated Early Warning System (IEWS) to support national EWS. Despite these efforts, national systems have encountered serious difficulties: the EWS in Burkina Faso and Chad no longer function, and some countries have no system at all. Niger’s EWS was restructured following the crises in 2005 (RPCA, 2006).

3. SOCIO-ECONOMIC APPROACHES

In the 1990s, the NGO Save the Children Fund (SCF) introduced a methodological approach known as the Household Economy Survey, which combines traditional nutritional surveys with data collected directly from target groups. Another NGO, CARE, developed a similar method using fieldwork that focuses on the security of household living standards. These methods, based on semi-structured interviews with groups of farmers, are known as rapid rural appraisal (RRA). They combine qualitative and quantitative data drawn from local and government sources (Ministry of Agriculture, etc.).  As a first step, a family’s general food needs are estimated and a balance sheet, complete with income and expenditure, is drawn up. A set of calculations is made for every socioeconomic group (‘wealthy, average, poor, very poor’), the groups being defined beforehand through interviews with ‘community representatives’. Food deficits for each socio-economic group are then estimated, taking into account harvests (data supplied by the Ministry of Agriculture), local cereal prices and the ‘coping mechanisms’ of each group. Beneficiary numbers can then be established. An approach of this kind requires the prior development (according to the same methodology) of a baseline or reference situation that helps to assess developments. The baseline recasts traditional administrative boundaries as agroecological zones which are defined according to the activities and incomes of the peasant population.Such methods contrast markedly with traditional approaches, for they privilege time spent with peasants over office meetings, stress the importance of information acquired at grassroots level during the decision-making process, acknowledge the role of economic factors in the concept of food security, analyse data at the level of the family unit, and redraw administrative boundaries for final projections.

Nevertheless, the data is still secondary in that it comes from government sources (crops and yields) or from ‘representative’ socio-economic groups which may resort to omissions and ‘strategies’ when interviewed. This method is notable for its independence with regard to local authorities (whose data is put into perspective), and for its challenge to the standard balance sheet assessment models (see below). Developed in tandem with EWS, socio-economic approaches are beginning to be acknowledged by governmental authorities and agencies. After a period without exerting any real influence on the evolution of early warning systems (Buchanan-Smith & Davies, 1995), this methodology is now recognized and applied.