Evaluation of Dafur Early Warning and Food Information System

North Dafur where SCUK operated until December 2004

Summary of evaluation¹

Save the Children UK (SCUK) recently commissioned an evaluation of the Darfur Food Information System (DFIS), which was established in North Darfur in 1993. This evaluation was carried out in November and December 2005 and covered the third phase of the DFIS from 2002 to 2004. By that time, the system had expanded to monitor the food security status of displaced Dinka in South Darfur and more recently had expanded into West Darfur. During most of DFIS's lifetime, its principal role has been to provide early warning of the impact of drought, the main threat to food security in Darfur. In 2003, armed conflict broke out in the three states of Darfur on an unprecedented scale and with ferocious intensity. Thus, DFIS had to adapt from being a drought-oriented early warning system (EWS) to one where conflict was the principal threat to food insecurity, triggering the displacement of over a million people. How DFIS performed in adapting to this very different context was one of the key focus areas of the evaluation. The evaluation also reviewed the DFIS's effectiveness in providing early warning of droughtrelated food insecurity before 2003 and in triggering timely response.

After the tragic death of SCUK staff, the organisation decided to rapidly pull out of all the Dafur States on 19th December 2004. As a result, SCUK had to discontinue DFIS without a proper hand over to other actors. This evaluation also considers the legacy of DFIS since SCUK withdrew from Darfur.

Performance of DFIS prior to the conflict - 2002 to early 2003

Since 1996, DFIS has been based on the Household Economy Approach (HEA) methodology. The starting point was to divide rural North Darfur into six 'food economy zones (FEZ)' and to carry out baseline profiles in each. DFIS has been run entirely by national SCUK staff during its lifetime, ensuring a strong sense of ownership of the system and of its methodology by the SCUK Darfur team.

It is unfortunate that some of the baseline profiles that underpinned DFIS were written up but not published up until 2004 - at least four years after they had all been completed - thus missing opportunities to influence and inform. Nevertheless, the baseline profiles provide very valuable information on livelihoods in North Darfur and were essential to the annual village and household survey carried out in October each year. This is when HEA really came into its own, producing a clear plan for food aid needs in each FEZ, as well as recommendations for other non-food emergency interventions. There is every indication that this annual assessment was carried out effectively, professionally and in a timely manner during the pre-conflict period. The collaboration between international non-governmental organisations (INGOs), the World Food Programme (WFP) and government in the process was exemplary, under the auspices of the Food and Water Emergency Committee of North Darfur. In 2003, SCUK planned a nutrition survey encompassing causal analysis of one of the FEZs, but the field work was hampered by insecurity.

Market monitoring was done weekly throughout the year as an ongoing indicator of food security status. Nutrition surveys usefully complemented the regular food security monitoring by confirming the existence (or otherwise) of a food crisis in a FEZ of particular concern. However, the underlying causes of malnutrition are still poorly understood for some FEZs in North Darfur.

The publication of periodic DFIS bulletins throughout the year was the main means of communicating EW messages between annual assessments. These bulletins brought together analysis of all the different indicators and were particularly useful for trend analysis. However, presentation of the bulletins was not very compelling. Recommendations in earlier evaluations to improve the bulletins do not seem to have been effectively implemented.

SCUK's early warnings of annual food aid needs in North Darfur were rarely heeded and did not trigger a timely response from donor governments. A very small proportion of estimated needs were delivered by June/July - the critical months - in 2001, 2002 and 2003. The political environment was simply not conducive to a timely response.

Between 2001 and 2003, DFIS's close relationship with state government really paid off in terms of how the authorities, and the Wali in particular, could use the DFIS analysis for their own lobbying and advocacy purposes with federal government and with international donors. Collaboration with federal government at a technical level was also strong. DFIS was seen to provide accurate and realistic estimates of need partly because of its methodology. But federal government's overall willingness to respond to food crises and food insecurity in Darfur was limited.

The negative impact of a consistently late response to food crises in Darfur included distress migration, high malnutrition rates and reduced cultivation because of food shortages in the early 2000s. As is typical for most EWS, DFIS's focus was biased towards information gathering and analysis with much less attention paid to communication and advocacy. Given the hostile political environment that DFIS was operating within, there needed to be a much greater investment in advocacy (and persuasive oral presentations rather than written reports) than was actually the case. This would have required organisational commitment at all levels: in Darfur, Khartoum and London.

Technical support and capacity building of DFIS partners has been a prominent feature of its work. As a result, SCUK has built up a cadre of individuals in North Darfur - in government and in INGOs - with a good understanding of food security and of HEA, and has built a sense of ownership of the DFIS approach. The challenge with government was how to build capacity in a sustainable way when government departments were so poorly resourced in terms of basic infrastructure, such as computers and transport. Also, as government officials became more skilled, many sought work with international agencies. The one gap in DFIS's capacity-building work was with local community based organisations (CBOs) and NGOs, some of which enjoy much greater access to rural areas and local communities during the current conflict than government staff.

Performance of DFIS once conflict became the principal threat to food insecurity, 2003- 2005

During 2003/04, SCUK's programme in Darfur went through a major transformation to adapt to a large-scale and highly politicised, conflictrelated emergency. By all accounts, this was a very painful transition, ridden with tension between incoming international staff and longterm national staff, unclear management structures and the emergence of dysfunctional parallel systems. This seriously inhibited DFIS's performance.

Methodologically, it really struggled to adapt. A major gap was the lack of any assessment of DFIS's capacity to adapt to this escalating conflict environment, nor was consideration given to bringing in external expertise, despite the limited expertise in-country to work in such a politicised conflict environment. Erroneously, the focus of the country programme was on longterm issues of vulnerability and food security in Darfur. But fortuitously this meant that an external consultant was brought in during 2004 to support DFIS to address underlying issues of vulnerability. Adapting to the new conflict environment was not part of the consultant's terms of reference, although this soon became central to her role. This input was critical but was late. For example, not until mid-2004 did the HEA methodology adapt to doing rapid assessments. Yet these were needed from mid-2003. When rapid assessments were carried out, the reports were strong and well-written. Meanwhile the DFIS bulletins were not. They tended to follow the old format so they were neither compelling nor clear in terms of their key message. A much sharper and tighter analysis was needed. In short, the key early warning/monitoring role that SCUK could have played during 2004 was not fully realised.

In order to respect humanitarian principles when conflict broke out, and to protect SCUK's independence, DFIS needed to adapt institutionally to this changed and highly politicised environment. Most notably it needed to distance itself from government, one of the key actors in the conflict. Unfortunately this took a long time to happen, dangerously affecting perceptions of SCUK. Such an adjustment in institutional relationships would never have been easy and could only have been managed by an expatriate presence. It was too much to have asked of national staff who had spent years building those relationships.

Despite these shortcomings, there is evidence that DFIS played a valuable role early on in the conflict, briefing incoming agencies and providing time-series data (e.g. market price data) on demand. Although slow to be written up, its database on livelihoods before the conflict is invaluable for comparative purposes to understand how livelihoods have been affected by the conflict, and is being used as such by a small number of international agencies.

SCUK's withdrawal from Darfur was lamented, without exception, by all agencies interviewed during this evaluation. It has left a gap in information collection and analysis at state-level that has not yet been filled, as most agencies focus on information collection in their own particular geographic area of coverage. Unfortunately the way that SCUK withdrew did not help to bridge this gap. Withdrawal appears to have become a logistics exercise in which strategic decisions about handover to other agencies and even protection of the SCUK resource base, were overlooked. In short, SCUK has not left behind a functioning information system in Darfur.

¹The Darfur Early Warning and Food Information System. Final Evaluation of Phase 111. By Margie Buchanan-Smith