The Missing Link

Introduction

By the early 1990s, most drought-triggered famines and food crises in Africa were not hard to predict. This is testimony to the success of famine early-warning systems (EWS), many of which were set up during the 1980s. In 1984–5, failure to prevent famine in the Sahel and Horn of Africa was widely attributed to a lack of early warning (EW). Since then, better prediction has been a major policy concern, for both donor agencies and national governments. Greatest attention has been paid to the establishment of purpose-built EWS. In the Sahel and Horn of Africa alone, more than eight new EWS were set up between 1985 and 1990. Considerable progress has been made, in improving data-collection methods, developing indicators, and making use of sophisticated information technology. By 1990 more information on the likelihood of famine was available to donors and governments than ever before.

Famine prevention, however, has remained an elusive goal. Better prediction has not led to corresponding improvements on the response side. There is clearly a missing link between the provision of EW information and the use of that information to trigger a timely preventive response. Anecdotal evidence abounds as to why EW information has not been influential. The EWS was not sufficiently vociferous; the information was inappropriate, late or untrustworthy; donors were ill-disposed to help a particular government; adequate resources were not available; institutional and logistical obstacles overwhelmed good intentions; the domestic political will to react was lacking, and so on. Intuitively such explanations make sense, but they do not amount to a systematic analysis of what happens to EW information once it enters the decision-making process and how it is used by policy-makers. How and when do they receive the information, and what kind of data are most influential and why? Most important, what are the missing links between prediction and prevention?

That is the subject of this book. It is based on an analysis of early warning and relief responses in 1990 and 1991 in five African countries: Ethiopia, Sudan, Chad, Mali and Kenya focusing on Turkana District in the north; and in a number of donor agencies. Our approach is to begin with the decision-making processes which determine the response to a threatened crisis. From there, it is possible to ascertain the relative importance of information and of other factors in triggering or inhibiting the response.

The agricultural year 1990–91 was a year of drought across much of the Sahel and Horn of Africa, with early indications that food crises would be widespread. In some regions, it was the combination of drought and conflict that threatened to cause the most severe cases of famine and suffering. Elsewhere, the drought was not necessarily as bad as in the mid-1980s. Nevertheless, conditions deteriorated to the point where relief aid was necessary, to protect livelihoods and often lives as well. A number of EWS in the region were being put to the test for the first time. Could they fulfil their ultimate objective of triggering an adequate and timely response to prevent acute food insecurity and/or famine developing? The results have been mixed. In some countries the EWS were remarkably ineffective. In others, they had much greater influence.

The reasons why EW information is, or is not, used fall into four broad categories: first, reasons to do with the EWS itself, and the information provided; second, reasons to do with the institutional context within which the EWS sits, and the institutional links to decision-makers; third, reasons to do with the broader political environment; and fourth, logistical obstacles to launching a timely and adequate response. Much has been written about the technical aspects of EW and numerous evaluations of different systems have been carried out. Most have focused on the internal workings of the EWS: the scope of indicators, accuracy of the data and timeliness of the warnings. These relate to the first category of reasons, concerned with the performance of the EWS itself. But few have looked at how EWS fit into the wider context. From the analysis presented in this book, the second and third categories of reasons emerge as the most important explanations of whether EW information is used, and of variations in performance between the different case-studies.

Most countries in the Sahel and Horn of Africa rely on the international relief system to provide resources to run relief operations in times of food crisis. National governments rarely have adequate resources or capacity to respond. Donor agencies and non-governmental organizations (NGOs), the key actors within the international relief system, have been particularly influential in developing famine prediction for the Sahel and Horn of Africa, both in funding and operational roles. But much less attention has been paid to developing the response side of the equation. There are two key reasons for the persistent failure to translate EW into timely response, illustrated in the case-studies in this book. First, the international relief system responds to famine once it is under way but is ill-equipped to respond to genuinely early warning, to intervene in time to prevent it. Second, it is not the severity of the crisis, but relations between international donors and national governments which tends to be the single most important determinant of the timing and scale of the international response. Thus, in the case of the southern African drought in 1992–3, as well as national capacity to respond, the desire of the donors to keep structural adjustment programmes on track and a determination to avoid further political unrest in the region combined to initiate a timely response by the international relief system (SADC, 1993). The same factors did not conspire to trigger a timely response to food crisis in the Sahel and Horn in 1991.

Understanding famine

A common theme running through this book is that a gap exists between theory and practice in famine prevention. Our understanding of famine as 'outsiders' has improved in leaps and bounds during the last 10 to 20 years. This should improve our ability to prevent it with appropriate policies and interventions. Instead, conventional and often inappropriate relief responses persist. Explanations of the causality of famine have shifted from a preoccupation with supply-side factors in the 1970s, towards a recognition of the pre-eminence of access or entitlement to food in the 1980s. Famine is no longer solely — or indeed even primarily — attributed to food availability decline (FAD), but increasingly to food entitlement decline (FED). Thus 'starvation is the characteristic of some people not having enough food to eat. It is not the characteristic of there not being enough food to eat' (Sen, 1981:1). This key...