Local famines, global food insecurity
by Roger Persichino, Action Against Hunger-USA July 2004

In early 2004, the Food and Agricultural Organisation (FAO)’s Food Outlook report estimated that 38 countries were in need of exceptional food assistance, the same number as in 2003, about 15% more than in 2001 and 2002, and the highest since 1984. This list is noteworthy less for what it includes (the majority of affected countries are in Africa) than for what it leaves out, in particular China and India where, by the FAO’s own assessment, in the State of Food Insecurity 2003, about 70% of the world’s undernourished people live. In other words, the FAO’s list says more about state capacity and/or recourse to international assistance than it does about actual food needs. It thus tells only one part of the story. To tell the full story, we need to take a closer look at the local dynamics of famine and the political and economic processes that influence it. This report argues that a focus on countries, rather than regions or specific areas, is a significant weakness. Famines are local, never national, phenomena.

General geographic patterns

The FAO’s publications yield consistent and useful data: since 1996, Food Outlook has listed countries requiring ‘exceptional food assistance’; since 2000, it has also given the main reasons leading to such food insecurity. Exceptional food assistance is defined by evaluating those food requirements that will not be met by national production and government intervention, for instance a strategic grain reserve and/or government purchase. The list of countries affected is therefore not indicative of famines as such, but rather of vulnerabilities that, especially when recurring over time, point to famine processes at work.

Overall, 53 countries have required exceptional assistance over the last six years: 29 in Africa (more than half the continent’s states), 14 in Asia, eight in Latin America (75% from Central America and the Caribbean) and two in Europe (Serbia and Russia – the northern Caucasus). The trend is steady overall for all continents except Africa, where requirements for food assistance have risen noticeably.

Twenty-five of these countries have required exceptional food assistance every year since 1998. This indicates sustained high levels of food insecurity.

Screen Shot 2012-10-18 at 3.35.23 PM

Screen Shot 2012-10-18 at 3.35.14 PM

The causes of famines

There is a wide range of views on what actually causes famines. Since 2000, FAO has given a country-by-country indication of posited reasons why food assistance is required. These fall broadly into four categories: natural disasters; economic issues (an international coffee crisis in Latin America, economic policy in Zimbabwe and Tajikistan); civil strife; and other causes (refugees in Tanzania and Guinea, IDPs/returnees in Serbia and Montenegro, returnees in Angola). In several instances, more than one cause is given, for instance drought and internal displacement in Ethiopia and Eritrea. The data allows for a crude presentation of famine causes since 2000, as estimated by FAO.

There are two clear features of famine patterns. First, Asian and South American food insecurity is mostly related to economic issues (dependency on monoculture in Latin America; the nature of economic policies in Haiti, Tajikistan and Mongolia) and regular natural disasters or unreliable weather. Second, African and European food insecurity appears to be essentially related to conflict and a high caseload of vulnerable populations. There are caveats to this analysis, of course; it does not, for example, capture the primary importance of economic issues in recurring food requirements in Madagascar. But it does indicate that two factors usually combine to determine structural vulnerabilities and chronically high food insecurity. The combination is, however, strikingly different between Africa and Europe on the one hand, and Asia and Latin America on the other.

In India, the Bharatiya Janata Party (BJP)’s unexpected electoral defeat in May 2004 was in part explained by the discontent of a rural population affected by drought and high levels of indebtedness. This suggests that rural voters feared that the BJP’s policies were threatening the social compact that has prevented large-scale lethal famines in India for the past half-century. This would be in line with the analysis here, which suggests that natural phenomena and economic policy issues are of primary concern in Asia.

Screen Shot 2012-10-18 at 3.35.04 PM

Famines, entitlements and emerging political complexes

FAO’s assessment is indicative of needs, not of actual assistance delivered. In that sense, it does not show those countries where government intervention could single-handedly address food insecurity, but has not done so.

A further limitation is that the FAO does its analysis on a country basis. This does not take into account the highly variable dispersion of food insecurity within countries, a point well-illustrated by early-warning systems such as FEWS and the FAO’s own GIEWS. Similarly, Save the Children’s food economy assessments differentiate between regions or food economy zones.

The point that famines are local phenomena is not new, but it is increasingly lost in aid agencies’ communications campaigns. Also lost is the fact that assistance requirements factor in government capacity to redistribute aid within their own countries. This capacity is calculated on the basis of national production, but does not incorporate political and/or logistical constraints (e.g. local conflicts; the state of the road network or the truck fleet), nor does it evaluate a government’s willingness to redistribute this aid (the issue of access).

The localisation of high food insecurity is significantly related to two distinct issues: wide disparities in entitlements are of primary concern in Latin America and some parts of Asia (Mongolia, North Korea, some regions of China and India); while emerging political complexes have more explanatory power in Africa and, to some extent, in Europe (the Balkans and the Caucasus) and Asia (Afghanistan and Central Asia). Both topics are vast, and can be touched on only briefly here.

Simply put, entitlements, from Amartya Sen’s analysis, deal primarily with demand issues in situations where some food is available, but the capacity to purchase it is low, because of restricted resources for instance, or because logistical problems impede adequate redistribution of food among markets. Rural populations are highly vulnerable to income variations related to crop prices that are set externally. This is especially true in economies that rely heavily on export crops such as cotton, coffee or cocoa, as in Central America or Africa. Cyclical variations in commodity prices lead to recurring falls in both employment and income, in turn leading to decreased capacity to purchase food.

This can be compensated for through welfare programmes and/or redistribution. But in the absence of successful poverty eradication, such measures do not affect the issues that underpin this structural vulnerability to high food insecurity. Furthermore, poverty eradication, even when moderately successful, is unevenly distributed geographically, as shown in India, Brazil and China.

Mark Duffield has defined ‘emerging political complexes’ as areas characterised by warlordism and illiberal economies (e.g. trade in illicit substances, the criminalisation of the economy). These complexes are transnational in essence, and reflect a process of political reconfiguration. The Mano River Union (Liberia, Sierra Leone and Guinea), Somalia, the Caucasus and Afghanistan illustrate the concept, with their reliance on drugs, weapons or primary commodities (and in some cases all of the above) through a political structure essentially disaggregated among warlords wielding local influence.

Duffield’s point is not uncontroversial, and it could be argued that it describes the 1990s more accurately than the present, with its substantial decrease in conflict areas through international intervention and/or peace agreements (in Sierra Leone, Liberia, Afghanistan, the DRC and the Balkans, for example). Nonetheless, criminalised economies and/or local violence substantially persist in the countries given as examples here.

The pervasiveness and resilience of emerging political complexes have a clear impact on food security and, ultimately, on famines. First, reliance on criminal trade reinforces income disparities between those who engage in such trade and those who do not (with the exception of narcotics production in Afghanistan and Colombia). Second, as demonstrated by David Keen in his analysis of the famine in south-western Sudan in the 1980s, such political structures amplify arrangements whereby famine benefits a fraction of the population. Lastly, environments that remain conflictual are an obvious impediment to effective redistribution mechanisms.

Finally, it is worth emphasising that economic and political issues are not separate: entitlement packages and political arrangements are mutually dependent. Most of Central America was in a state of armed conflict throughout the 1980s, and several Indian states remain affected by low-level insurgencies, such as the Naxalite movement or the Assam Liberation Front: either through ‘post-conflict’ or ‘low-intensity conflict’ phases, violence is never far away from the political and economic processes that underpin famine.


A focus on the locality of famines illustrates three critical points. First, the spread of poverty eradication – which should lead to increased entitlements – at country levels will determine salient areas of high food insecurity: the more poverty, the more vulnerability to famines will remain and, conversely, successful poverty eradication campaigns should translate in less vulnerability to entitlement crises. Second, the emergence of peace agreements, especially in Africa (in Liberia, Burundi, the DRC, and potentially in Sudan) at national level should not mask the resilience of conflicts at the local level, tied as they are to emerging political complexes. Third, while most of the analysis in this article implicitly deals with rural populations, increased urbanisation is set to pose new challenges. In Sub-Saharan Africa alone, urban populations are projected to account for 45% of the total in 2020, up from about 20% in 1980. At current levels of employment and services in urban areas, this will inevitably result in the spread of urban slums, with very low access to water and sanitation services and no recourse to subsistence agriculture. While none of this is inevitable, it suggests that the geographical distribution of famines is bound to become increasingly discrete. To address this will require a full acknowledgement of the importance of local issues.

Roger Persichino is desk officer for Action Against Hunger-USA. The views expressed in this article do not represent an agency position.

References and further reading

Stephen Devereux, Famine in the Twentieth Century, IDS Working Paper 105 (Brighton: Institute of Development Studies, 2000).

Mark Duffield, Global Governance and the New Wars: The Merging of Development and Security (London: Zed Books, 2001).

FAO, Food Outlook, 1996–2004, available at www.fao.org/docrep. FAO, The State of Food Insecurity 2003 (Rome: FAO, 2003).

David Keen, The Benefits of Famine: A Political Economy of Famine and Relief in South-western Sudan, 1983–1989 (Princeton, NJ: Princeton University Press, 1994).

Sue Lautze et al., Risk and Vulnerability in Ethiopia: Learning from the Past, Responding to the Present, Preparing for the Future (Washington DC: USAID, 2003).