The Horn of Africa has been affected by a variety of climatic events which have led to several major droughts in the last 15 years. The international aid system has tried to adjust to these events and improve its capacity to respond. The most recent episode of severe drought took place in 2016-17. This led to a huge international relief effort, and despite very high malnutrition rates, enormous loss of livestock assets and a great deal of displacement, mass mortality was avoided across the region, and famine conditions did not ensue.
In 2018, the Directorate-General for European Civil Protection and Humanitarian Aid Operations (ECHO)1 contracted the INSPIRE consortium to assess how the response by the international humanitarian community as a whole, and ECHO in particular, had changed since the previous droughts in the Horn of Africa (HoA), in 2011-12. The review aimed to analyse the factors that determined the performance of the response, to document improved practices and the challenges in their implementation, and to identify lessons, so that improvements can be made to ECHO’s operational procedures, structures and policies.
Contextual changes and specific crisis dynamics
Although the crises in 2011-12 and 2016-17 were labelled droughts, it would be misleading to think of them simply as the consequences of rain failure. The interaction between rain shortage, mobility, political crises and conflicts has been widely documented. Throughout the region, the worst indicators were not necessarily in the areas with the greatest anomalies in rainfall, but rather in the areas that had been affected by conflict, or were marginalized and under-developed. In pastoral or agro-pastoral areas, one failed rainy season should only create hardship rather than famine. On the other hand, if several consecutive bad seasons are combined with structural poverty, political marginalisation or conflict, the population may be unable to adopt adaptive population movements and coping mechanisms. Climate change is increasingly affecting the area, with changing rainfall patterns (both in terms of time and geography) and the modification of the vegetation cover. As a result, periods between bad years are becoming shorter and recovery is more difficult.
Levels of resources allocated
Between 2011-12 and 2016-17, there was no significant difference in the amount mobilised for the global international response for Somalia, while the amount was somewhat higher for Ethiopia. For Kenya, very little was mobilised in both cases by ECHO, while resources coming from EU funded development programmes were allocated to the drought response. It seems that the main question is not how much money is mobilised but how it is used.
Understanding weather information better and using it more efficiently
Although weather forecasting is continually improving and becoming more accessible, it is still not playing a significant role in determining how resources should be used. Governments, who have access to the information, are still rarely willing to allocate resources on the basis of forecasts, when they have so many other pressing demands for resources. Unfortunately, the general unwillingness to act on the basis of forecasts is extended, to a certain degree, to preparedness measures. The implications of the unwillingness to allocate resources until indicators of suffering (e.g. GAM) are rising are well recognised: given the time taken to translate funding decisions to assistance on the ground, response will always be late.
Timeliness of the response
There was a significant improvement in terms of alertness and geographic coverage between 2011 and 2016-17, with some variation between the three countries. The quicker triggering of the response in Somalia in 2016, and even more in 2017, was possible due to robust early warning information which was provided more quickly through different formal and informal channels, the presence of more actors in the field and the engagement of key donors who were determined to avoid a repetition of the 2011 famine. In Kenya and Ethiopia, the improvement in timeliness was more limited as most of the early warning systems and response mechanisms are government-led systems which too often react slowly. The humanitarian response was delayed by the same factors in 2015 and in 2016 as it had been in 2011-12, namely: limited willingness to respond to meteorological forecasts, even for heightened preparedness; an insistence on waiting for ‘official’ early warning (despite the well-known inability of these systems to be timely, see above); an unwillingness to respond based on an analysis of the inevitable trajectory of livelihoods and humanitarian indicators, waiting instead until such indicators (especially child malnutrition) were already critical; an unwillingness to divert development resources to scale up support where critically needed (for livelihoods, water etc.) in the absence of a Government-recognised emergency; slow bureaucratic processes, exacerbated by the centralisation of decision-making; and lack of preparedness by operational agencies, leading to long delays between the decision to act and actually reaching people in need.