Issue 27 - Article 5

Famine (again) in Ethiopia?

July 13, 2004
Sue Lautze and Angela Raven-Roberts, Tufts University

In 2002, Stephen Devereux, an economist at the UK-based Institute of Development Studies (IDS) stood on the dusty plain outside Gode in Ethiopia’s Somali Region, and argued that the Ethiopian government and the international community had failed to prevent another famine. Devereux cited a US Centers for Disease Control study led by Peter Salama that demonstrated a concentration of excess mortality that coincided with the onset of relief food assistance to Gode in 2000. The IDS critique stung many in the humanitarian community in Addis Ababa, including USAID/Ethiopia staff responsible for food and humanitarian assistance.

As the warning signs of another crisis in Ethiopia intensified in 2002 and 2003, the USAID mission in Addis was keen to avoid another round of academic post-mortems. In the midst of the crisis, USAID engaged a group of academics, including specialists from Clark University, the Food Economy Group and the Feinstein International Famine Center at Tufts University. A number of these academics had experience in Ethiopia predating the famine of 1984–85. USAID’s first (and repeated) question was: ‘Is this a famine?’.

When pressed on the issue by an IRIN reporter in Addis in July 2003, the Tufts team stated: ‘You have a widespread livelihood crisis leading to emergency levels of malnutrition, morbidity, mortality, with alarming implications for destitution. That for us is a famine’. However, to call the 2002–2004 crisis in Ethiopia a famine had different implications for different actors. Famine, like genocide, seems to imply a particular level of moral responsibility above ordinary disasters, albeit without genocide’s legal obligations. As noted by Alex de Waal and a number of other writers, declaring that a famine exists is a political exercise. This article examines some of the issues that arose for analysts in their engagements with USAID and others on the question of famine in Ethiopia.

‘No Famine on My Watch’

In the interim between the two most recent crises, there were two important political developments in the US. First, following the 11 September attacks in the US, Ethiopia’s importance to US national security interests increased. USAID’s aggressive food aid response to the government’s 2003 appeal for food aid accordingly raised questions about political motives. Some in the donor community thought that USAID was overreacting; as one British government official put it, Ethiopia was experiencing ‘just another normal bad year’. Second, President George W. Bush challenged his administration to ensure that famines were avoided during his tenure, a policy known as ‘No Famine on My Watch’; declaring the existence of a famine could be seen as a political shortcoming and, therefore, a political vulnerability.

On the other hand, elevating the crisis to the level of famine was a tool for galvanising US public opinion, for advocating for a new food security strategy by the Ethiopian government, and for encouraging a higher level of engagement by humanitarian and development institutions, including the rest of USAID. The compromise eventually reached within USAID was to refer to ‘localised famines’. This implied that the situation was serious, but not unmanageable. The qualifying concept of ‘local’ makes possible unsubstantiated claims of success: famines that did not expand to a national scale.

Declaring a famine was also a complicated question for the Ethiopian government. Famines have contributed to the downfall of Ethiopian regimes. The government’s Disaster Prevention and Preparedness Commission (DPPC) is charged with preventing famines of the 1984–85 type: that is, those types of famine that make regimes fall. In both the 1999–2000 and the 2002–2004 crises, the DPPC fulfilled its mandate well; neither crisis was allowed to develop into the type of famine that threatened the political order.

Some humanitarian practitioners gauge their successes, in part, according to ‘famines averted’; declaring the existence of a famine is seen as a mark against the humanitarian effort. Both the World Food Programme and Addis-based European Union staff, for example, reacted strongly against the Salama article. Development actors resisted the use of the word famine out of a concern that emphasising humanitarian responses would come at the cost of development efforts.

The media are interested in the answer to the ‘is there a famine’ question essentially because of the definition of ‘news’: vulnerability, poverty, suffering, disease and a certain level of death are viewed as the status quo for Ethiopia. To be news, these tragedies must be elevated to extraordinary levels, levels which the press and the public have come to associate with the term ‘famine’. There has been what Mark Bradbury has called ‘an accommodation of suffering’ over time: the numbers of people affected by recent crises dwarf those affected in 1984–85. However, the 2002–2004 crisis largely failed to produce the ‘famine image’ the media and public associate with Ethiopia.

Is it a famine?

Although most famines are characterised by widespread distress migration, this is neither necessary nor sufficient for the definition of famine. The absence of distress migration in a crisis may be because migration to camps is prohibited, or because there are no camps, as in North Korea throughout the 1990s. Alternatively, deeply vulnerable populations may be able to remain in their home communities, for example in the absence of forced displacement or due to the timely provision of relief food. By mid-2003, this seemed to be one possible explanation for the absence of large-scale distress migration in Ethiopia. In 2003, USAID provided 999,650 tons of food aid.

The Tufts team argued that the nature of the food aid response (coupled with increasing restrictions on internal migration) induced vulnerable populations to remain in their home communities. The humanitarian community has long believed that preventing distress migration is a highly desirable outcome, but one rarely achieved. The team concluded that the Ethiopian government and the humanitarian community were ill-prepared to manage the resulting dispersion of vulnerability. The team argued the following line of logic:

  • Vulnerability to malnutrition, morbidity and mortality increased sharply because of the nature of the multiple collapses of livelihood systems.
  • Against these threats, food aid alone was insufficient and, in some cases, inappropriate.
  • The hungry, the sick and the dying then had to rely on existing institutions for assistance; these were either grossly inadequate or already overtaxed.
  • The humanitarian community proved once again that is not adept at fixing systems and institutions in the midst of crisis.
  • Consequently, instead of concentrations of famine victims, as in 1984–85, there was a diffusion of famine victims.

In the absence of large-scale migration, it is much easier for the Ethiopian government, donors, UN agencies and NGOs to declare once again that famine was averted, or minimised as a local affair. The lack of classic famine images from IDP camps facilitates further disengagement by the media and Western publics, even as large numbers of vulnerable people face destitution, malnutrition, morbidity and mortality.

The capacity of institutions to manage crises remains compromised by policy reforms, including the Ethiopian government’s too rapid and inadequately supported decentralisation process. While decentralisation may in future enable more localised management of crises, the timing of the decentralisation process was unfortunate; the distribution of famine suffering in Ethiopia neatly corresponds to the areas of greatest administrative weakness, such as Afar and Somali Region. The international humanitarian community has been unable to step into the breach. While this can be interpreted as a lack of commitment, it is also due to the fact that the humanitarian community is overwhelmed with disasters throughout the world.

Neither declared nor averted

Many now hasten to declare the crisis over. In Ethiopia, famines are neither declared nor averted, but are rather declared to be averted. Instead of engaging in potentially divisive discussions on the existence or otherwise of famine in Ethiopia, a far better strategy would be to describe the recent confluence of interlocking vulnerabilities as a terrible, slow-onset crisis of livelihoods. There remain great needs for immediate, transition and longer-term responses, both to save lives and to protect and promote livelihood systems among disparate livelihood groups. Consensus is still needed on the long-term processes that induced the tragedies characterising Ethiopia in order to inform more coherent short-, medium- and long-term humanitarian, development and political strategies to address the structural vulnerabilities inherent in a range of Ethiopian communities and production systems. These processes entail far more than droughts; they include local, regional and international political, social and economic dynamics as well as other man-made and natural causes.

In moving forward, it is important to challenge at least three key assumptions that dominate the current discourse about crises in Ethiopia:

  1. That emergencies ‘suddenly emerge’, and therefore should be treated as rapid-onset crises, rather than slow-onset crises.
  2. That crises are ‘food crises’ as opposed to ‘livelihoods crises’.
  3. That the people most at risk of malnutrition, morbidity and mortality are the same people identified as being ‘chronically vulnerable’.

This last point may be counterintuitive. Much more work is needed to determine exactly who is both ‘chronically vulnerable’ and demonstrated the greatest vulnerabilities in recent crises. It is likely that these investigations will show one population that is well entrenched in a system of annual food aid receipts (the ‘chronically vulnerable’ or the ‘chronically eligible’), and a very different population that is most at risk during the current emergency (for example, those on the margins of administrative reach). There is at present an unhelpful assumption in the development community that these two populations are the same; this is leading to intervention strategies that assume a neat and linear (and doubtful) relationship between identified chronic and acute vulnerability in Ethiopia. These assumptions underpin the government’s Food Security Strategy, which includes the resettlement of chronically vulnerable populations to often distant and unprepared sites.

The second step is to engage with the institutions and systems that vulnerable communities rely on for survival. The definition of vulnerability needs to be expanded beyond a focus on targeted individuals and groups to consider institutions, policies and processes. These systems include informal kinship networks, state-managed social service systems and market networks. In addition to material inputs, a common denominator in each of these systems is the role of cash, income opportunities and debt. Health, water and food security services are commercially available in many of the affected communities; vulnerable populations need to be assisted in accessing these services. The state is functioning in Ethiopia, even in the midst of decentralisation. Direct support to state budgets to manage social services should be considered, while a plan for revitalising former state farms is urgently needed. Similarly, new partnerships with states should be pursued for the creation of emergency, cash-based public works programmes. The extremely poor condition of Ethiopia’s public health system needs to be a focus for major donor input and aggressive UN leadership, underpinned by broad consensus on the implementation of minimum goals and achievements.

The third step is to critically engage with the government of Ethiopia about the relationship between its policies and acute vulnerability. A rational policy of environmental management is critical. Resettlement may be one partial solution to Ethiopia’s food security challenges, but to pursue resettlement in the midst of crisis – and at its present rate – is to impose additional burdens on communities, service providers and the international community. Other policies appear to have been unquestionably endorsed by the international community, despite the vulnerabilities associated with them: the risk of malaria associated with water harvesting, the promotion of maize at the expense of farmer choice about crop diversity, the lack of marketing strategies for bumper harvests. The list goes on.

Donors need to work with the Ethiopian government, the UN and NGOs to tackle the underlying causes that lead to the kinds of acute crises that prompt academics to wonder if it is time to use the F-word. Consensus is needed among major actors on a comprehensive and complementary action plan that provides mitigation, preparedness and recovery activities for vulnerable communities, while also providing accelerated input to those sectors and areas with the capacity to absorb real investment and development.

Sue Lautzeis Director of the Livelihoods Initiatives Program at the Feinstein International Famine Center, Tufts University, Medford, MA. Angela Raven-Roberts is the Center’s Director of Academic and Training Programs. Their email addresses are and

References and further reading

Mark Bradbury, ‘Normalising the Crisis in Africa’, Journal of Humanitarian Assistance, 1998,

Alex De Waal, Famine Crimes: Politics and the Disaster Relief Industry in Africa (Bloomington, IN: Indiana University Press, 1998).

Stephen Devereux, ‘Learning the Lessons? Famine in Ethiopia 1999–2000’ Institute of Development Studies, University of Sussex, 2002 (video).

Paul Howe, ‘Reconsidering “Famine”’, IDS Bulletin, vol. 33, no. 4, October 2002, pp. 19–27.

Sue Lautze et al., Risk and Vulnerability in Ethiopia: Learning from the Past, Responding to the Present, Preparing for the Future (Boston, MA: Feinstein International Famine Center, Tufts University and Inter-University Initiative on Humanitarian Studies and Field Practice, 2003).

Peter Salama et al., ‘Malnutrition, Measles, Mortality and the Humanitarian Response During a Famine in Ethiopia’, Journal of the American Medical Association, vol. 286, no. 5, August 2001, pp. 563–71.


Comments are available for logged in members only.