About three years ago, the Famine Early Warning System Network (FEWS NET) looked for places where it could eliminate country coverage in order to reduce costs. We looked at the Sahel, and its successes in creating a regional West African/Sahelian market, and in substantially increasing regional grain production in response to new market incentives, and wondered if famine had been beaten there. Last year, pictures of stick-thin babies and claims of millions starving seemed a devastating response, proving that the problem of famine in the Sahel still exists.
On the surface, Nigers crisis of 2005 was no different from others in Africa: a drought, locusts, extremely poor people, a food crisis, malnutrition and disease, followed by visible community and regional famine. However, there were features of this crisis that led to a very public debate about whether it was a famine at all. For some, this may have been seen as an inappropriate debate about semantics. For others, though, it was indeed a semantic debate, but one which was trying to find the right words to describe and appropriately deal with a very different type of crisis.
Famine arrives, devastates the community, is (poorly) resolved and goes away. But Nigers deadly famine-like crisis is driven by decades of chronic, extreme poverty that has undermined the ability of individual households in the community to participate effectively in the economy that surrounds them. It is permanent, and insidiously attacks individual infants and children in a large and scattered number of households, all the time. Children are acutely malnourished and die in numbers that are far beyond emergency levels each year, irrespective of good or bad harvests, drought or good rains or food prices. This is not a famine, nor a chronic nutritional crisis. It is something much worse.
What happened in Niger?
Late in the main 2004 rainy season, locusts invaded the agro-pastoral and pastoral zones of Niger, causing substantial but very localised, damage to grasslands and crops. Shortly afterwards, the rains came to an abrupt halt in many of these areas. Because this was before the millet and sorghum crops had finished filling their grains, there was considerable loss of yield (estimated at twice the losses caused by the locusts). Early warnings were given, generally suggesting the likelihood of severe local food insecurity and a need for food aid.
Cereal prices began rising in January 2005, leaving millions unable to buy the quantities of grain they needed. By June, as media stories and photos of malnourished and starving children became more common, the food security crisis seemed to be clearly regarded internationally as a famine. Yet for many Nigeriens, and a good number of the food security community working there, the case for a famine remained remarkably difficult to accept. Why?
The crisis we expected
In reality, there were at least two simultaneous crises, the causes and impacts of which were not adequately distinguished. The first occurred in the agro-pastoral and pastoral zones, commonly considered the most vulnerable in the country to food insecurity. These areas rely on rain-fed grain production, the sale of livestock to buy grain and loans of money and grain from merchants. People in these areas have some of the lowest incomes and fewest assets of anyone in Niger, and experience malnutrition rates almost constantly above the 10% global acute malnutrition (GAM) rate that is a threshold for an emergency. Most of the food aid needs assessments carried out in 2004 and 2005 by the Niger government and its partners focused on these areas. The governments subsidised sales of food security stocks and food aid distributions were principally carried out in these areas. Almost certainly, malnutrition and infant mortality increased, as they do every dry season. But data are insufficient to quantify the increase; for many observers, calling this crisis a famine, meaning in some sense an abnormal and deadly food-related crisis, was never justified.
and the crisis we didnt
It is curious and significant that the most public and documented evidence of the famine emergency in Niger the admission of thousands of very malnourished children to nutritional centres did not come from the pastoral and agro-pastoral zones. The children found in Médecins Sans Frontières (MSF) and other nutritional rehabilitation centres in Maradi and Zinder were from these two urban centres and from farming areas along the border with Nigeria, well to the south of the pastoral and agro-pastoral zones.
By almost any measure, both of these areas are among the wealthiest in Niger: they are the countrys breadbasket, the source of most of its agricultural surpluses and commercial crop production and its commercial hub. They also enjoy close trade links with Nigeria. Both were largely untouched by locust damage in 2004, and did not suffer the early termination of rains that affected areas to the north. The cereal prices that these districts paid on the market were at least several percentage points below prices in the crisis districts to the north. Relatively wealthy, spared the locusts, little affected by the drought, paying less for food why did desperate famine conditions emerge here in 2005?
Curiously enough, despite their apparent wealth, these areas have some of the highest chronic rates of malnutrition in the country. A Food and Agriculture Organisation (FAO) nutritional profile for Niger, published in 1998, stated: The anthropometric indicators for children clearly show that the Department of Maradi (20% wasting, 43% stunting) is most affected, followed by Zinder, and numerous factors [including] diarrhoeal disease, high rates of infant and child mortality (>350/1,000), very young first-delivery mothers, and above all, feeding habits such as the early weaning of newborns 45 days after birth being fed on water, herbal teas and cows milk, partially explain the contradictions observed between nutritional status and food consumption, especially in Maradi. Most of these explanatory factors do not relate to the consumption of food, and food aid will not eliminate the causes of the areas malnourishment.
The vast majority of cases of severe malnutrition were infants aged between six and 24 months, who are very vulnerable to weaning-related illnesses, and it was reportedly uncommon to see more than one child from any single household in these camps. There were almost no cases of severe acute malnourishment of adults. Taken together, all of these features are more consistent with a chronic poverty-related condition, where non-food factors are more important in creating the nutritional crisis, than a generalised lack of food in the household, or starvation. Yet the case was forcefully made by numerous actors, including MSF, the media and UN organisations, that a famine was occurring in these areas, and the pictures and the nutritional centre admissions were there to prove it. On the other hand, as Nigerien government officials and others noted at the time, we have often seen famine, and it is not this.
Why the second crisis?
Why did these relatively wealthy, well-watered southern districts suffer the most in terms of severe acute malnutrition and child mortality? Curiously, one of the principal reasons may have been the areas cash economy and market orientation, key features of its relative wealth. These districts, like those of the ethnically-linked Hausa-speaking areas to their immediate south in Nigeria, are among the most densely populated in Africa. Agricultural plots are relatively small, and becoming smaller due to population pressure. The pressure on land and the presence of the huge Nigerian market form strong incentives to grow commercial crops for export. The long commercial traditions of the Hausa state have also built a large body of traders and merchants willing to provide commercial financing, and informal production and consumption loans.
Many of the poorer farmers cannot compete in this arena, and have had to sell their land and become wage labourers on cash-crop farms. Both farm owners and labourers often have to supplement the cash they receive, mostly at harvest-time, with loans of grain to support food consumption, but there are long periods when cash and food are in short supply in the household. This may help to explain the high chronic levels of malnutrition amongst the children of farmers who, from their cash income alone, appear to be better-off than most.
A loan repayment system characteristic of these areas greatly worsens the impacts of this uneven availability of cash. Traders offer loans of grain that are monetarily valued at the prevailing market rate when the grain is given, and the loans are paid back in kind for the same value. Because loans tend to be taken out at periods of the year when grain is most expensive, and are repaid right after the harvest when grain is cheapest, the farmer may have to give back several times more grain than was originally received. The worst possible combination for these farmers would be a year of extremely high grain prices before the harvest, followed by very low prices afterwards due to good grain production. Even worse would be a string of such years, as has arguably been the case in Niger since the last poor harvest in 2001.
Thus, net disposable income, after loan repayment, has decreased each year in these areas, leading to progressive impoverishment. Add in extremely high food prices, a dependence upon food purchases on the market and a reduction in accessible food for pre-existing chronic conditions of severe malnourishment, and it makes sense that many children in these relatively wealthy areas faced the worst nutritional outcomes, if not mortality, in 2005. What does not make sense is calling this a famine.
Neither of the two food crises in Niger in 2005 displayed the breadth, depth or prevalence of hunger most formal definitions of famine require. Based on the CDC/UNICEF nationwide nutritional survey carried out in October 2005, it is not even certain that the levels of malnutrition during 2005 were far outside the range of chronic rates found every year in Niger. Neither crisis was primarily driven by a complete shortage of food in markets, nor by a complete shortage in individual households. Patterns of affliction in the households where infants and children were solely affected are more reflective of long-standing poor childcare practices, health and sanitation issues and poor water quality. Today, after the 2005 famine has subsided, the numbers of children requiring nutritional assistance remain at emergency levels.
Nigers 2005 crisis is an elegant case study of two different livelihood crises occurring in the same poor country, at the same time, in different locations. Both compel us to more fully appreciate the chronic poverty that underpins many, perhaps most, food emergencies. This is the reality of food and nutrition crises in Africa: the traditional famine-initiating droughts and locust plagues matter much less in creating acute humanitarian crises than we normally assume. What is much more important is the degree of underlying poverty, the state of essential public services and how markets and food access are integrated into livelihoods.
We should all review our assumptions about food crises and incipient famines. Are we measuring the right indicators of food security and the risk of malnutrition? Are we monitoring the right vulnerable groups? Do we have a good enough baseline of malnutrition to clearly capture this important factor of food security assessment in most countries today? Full-country monitoring on a regular basis, of even the lesser vulnerable groups, is clearly more important than we thought. The market needs to be more closely monitored, not only near the livelihoods it serves, but throughout the regional market watershed that feeds it, or is fed by it.
For FEWS NET and other early-warning systems, the most immediate challenge lies in accepting that we can no longer limit our monitoring and analysis to the strict confines of food security and food-related crises. Livelihood emergencies of many different types, in many different places, will produce many of the next food security or famine crises. Indeed, identifying what not to monitor will be among our most difficult tasks.
References and further reading
FEWS NET/Niger reporting: http://www.fews.net/centers/?f=ne.
FEWS NET/Niger livelihood zones: http://www.fews.net/livelihoods/files/ne/national.pdf.
Excerpts from the Niger Comprehensive Food Security Vulnerability Assessment (CFSVA), WFP, Rome, 2005, http://documents.wfp.org/stellent/groups/public/documents/ena/ wfp073376.pdf
P. Howe and S. Devereux, Famine Intensity and Magnitude Scales: A Proposal for an Instrumental Definition of Famine, Disasters, 28(4), 2004, pp. 35372.
T. Reardon, P. Matlon and C. Delgado, Coping with Household-level Food Insecurity in Drought-Affected Areas of Burkina Faso, World Development, 6(9), 1988, pp. 1,0651,074.
H. Young et al., Food Security Assessments in Emergencies: A Livelihoods Approach, HPN Network Paper 36, 2001.
J. von Braun and R. Pandya-Lorch (eds), Income Source of Malnourished People in Rural Areas: Microlevel Information and Policy Implications, IFPRI Working Paper on Commercialization of Agriculture and Nutrition, no. 5, 1991.
P. Webb and A. Harinarayan, A Measure of Uncertainty: The Nature of Vulnerability and Its Relationship to Malnutrition, Disasters, 23(4), 1999, pp. 292305.
Gary Eilerts is Program Manager and Cognizant Technical Officer, Famine Early Warning System Network (FEWS NET), in the US Agency for International Development (USAID). His email address is email@example.com.