Living memory of famine in South Sudan: using local knowledge to inform famine early warning

Soon after armed conflict began in South Sudan in late 2013, humanitarians began raising questions about the risk of famine This research utilised the IPC definition of Famine, including the theoretical view of Famine as a process in line with the IPC analytical framework. Catastrophe and Famine (IPC Phase 5) are often the result of an accumulation of one or multiple shocks that disrupt livelihoods and strain coping capacity to the point of collapse over time, leading to distress migration and human disease outbreaks among severely food insecure populations. See the 2016 IPC Guidelines on key parameters for famine classification for the IPC definitions of Catastrophe and Famine (www.ipcinfo.org/fileadmin/user_upload/ipcinfo/docs/IPC_Famine_Guidelines_Nov16.pdf). should the crisis endure, especially if compounded by natural disaster. Only 15 years prior South Sudan experienced a particularly virulent man-made famine, worsened by poor rainfall, while under the watch of the largest coordinated humanitarian operation of its time, Operation Lifeline Sudan. The need to avoid the early warning mistakes of the past was clear, yet how to achieve this was not.

In late 2015, the Integrated Food Security Phase Classification (IPC) was used to report that there was insufficient evidence to declare famine in South Sudan’s Unity State, despite available information suggesting famine may have been ongoing at the time. In February 2017, the IPC Key Messages described nation-wide food insecurity as ‘unprecedented’ and classified two counties of central Unity State as ‘in famine’. By January 2018, more than half of the population, over five million individuals, were reportedly food insecure – a 40% increase over the previous year, when famine had been declared.

Humanitarian access remains severely constrained in many of the most at-risk areas of South Sudan. Given this difficult operating environment, and with an eye on the 2018 lean season – the period of the year when hunger is typically most severe – REACH REACH is joint initiative of IMPACT Initiatives, ACTED and UNOSAT. REACH builds on the expertise and experience of the three organisations and strengthens effective and evidence-based humanitarian action through the development of information products and provision of independent assessments and analysis in support of aid coordination mechanisms. sought to contribute to a better understanding of famine risk in South Sudan.

REACH began designing its research by reconsidering the Southern Sudan Vulnerability Study The Southern Sudan vulnerability study, published by Save the Children with funding from USAID/OFDA, can be accessed through the Sudan Open Archive managed by the Rift Valley Institute (www.sudanarchive.net). and similar work in Darfur Placing a strong emphasis on local perspectives, these include de Waal, A. (1995) Famine that kills: Darfur, Sudan. New York: Oxford University Press; and Young, H. and Jaspars, S. (1995) Nutrition matters: people, food and famine. London: Intermediate Technology Publications. – all conducted before, during, or after famine periods. While these studies all emphasise the need to use local knowledge in humanitarian analysis, their findings have been largely ignored by newcomers to South Sudan as the preface to the 1998 Southern Sudan Vulnerability Study predicted.

As an anthropologist and an aid worker myself, I can understand why so few people have read any of the rich material that has been written about southern Sudan. However the disadvantage is that we are often re-inventing the wheel – commissioning studies, like the present one, to write reports that will not be read by the next set of practitioners, who will commission their own reports, feeling they are the first to tread uncharted territory. Harragin, S. and Chol, C.C. (1998) Southern Sudan vulnerability study. Kenya: Save the Children, p. ii (www.southsudanpeaceportal.com/repository/southern-sudan-vulnerability-study/).

To avoid this, REACH chose to use a qualitative approach to explore recent and historical experiences of severe hunger and famine and the shocks perceived to have caused them, as told by South Sudanese in their own words. As several populations across South Sudan have experienced famine within one or two generations, particular emphasis was given to the use of local definitions and comparisons between locally named historical periods and the present crisis. Living memory of famine and extreme hunger was explored through 36 focus group discussions with almost 200 participants, and findings from 22 of South Sudan’s 78 counties were combined with a review of secondary data. REACH continues to use the 10-state system for humanitarian purposes in line with the humanitarian community in South Sudan.

South Sudanese views of famine

While the research produced insights into a range of specific shocks affecting South Sudanese communities and their resilience to hunger and famine, the following three findings also apply to a wide range of other contexts.

1. From snapshots to narratives of famine

Focus group participants consistently reported that unusually severe natural shocks, such as flooding or a large-scale outbreak of crop pests, were the proximate causes of historical famines in assessed areas. However, respondents noted that these events followed or took place alongside extended periods of increasing vulnerability tied to armed conflict affecting access to food sources, cattle and agricultural tools, and common ways of coping with food insecurity. Natural disasters and disease outbreaks then tipped the balance from worsening food insecurity into mass mortality, a chain of events that culminated in a situation described as ‘famine’. Ultimately, the circumstances participants described as famine were the result of the cumulative impact of a series of shocks occurring over several years, calling into question the temporal range of contemporary early warning in South Sudan.

As the IPC is explicitly designed to examine food insecurity as a series of discrete snapshots, often within a timeframe of a year or less, there is a risk that the long and complex story of how a famine developed, or how the threat of famine may be rising, is lost. Famine is not a singular event, but more often the result of continued shocks that occur, individually and in combination, over several years, ending in a collapse of coping strategies and widespread death due to hunger, acute malnutrition, and disease. Focused largely on the end of this process, IPC is useful for declaring an ongoing famine and identifying the start and end points of mortality peaks. Yet, it is not always the best tool for identifying approaching famine.

The cumulative impact of a series of distinct or overlapping shocks over a multi-year timeline appears to be important in predicting the onset of famine – especially from a South Sudanese perspective. While taking a longer view is not new in studies into the nature and causes of famines, stringing IPC updates together into a holistic narrative of famine risk is not common practice in contemporary early warning.

2. Zero-sum coping: choosing between hunger, security and health

As the food insecurity of communities and households increases, it is expected that a selection of coping strategies of varying severity will be used to bridge gaps in food consumption. In early warning analysis, it is equally important to understand why some strategies have been used and others have not.

Participants described how the strategies used to cope with hunger, human disease, and physical insecurity were often zero-sum, with the use of one strategy limiting or preventing the use of others. The competing demands of these issues on labour, assets, and time often force stark trade-offs, where communities and households must decide which risks are most acceptable. In South Sudan, this may mean being forced to hide from armed groups in wetlands at the expense of increased exposure to diseases like cholera and malaria; abandoning far-off agricultural land in favour of more defensible but less productive land; and increasing access to food by moving more household members to fishing and cattle camps, despite heightened disease risks to vulnerable groups like the elderly and children.

3. Disease: cause and consequence of food insecurity

Death in famine is rarely the result of outright starvation, but typically occurs through the interaction of severe hunger, acute malnutrition and disease. In many cases, a collapse in a population’s ability to cope and subsequent migration in search of food greatly increases the risk of human disease outbreaks among extremely food insecure populations. Focus group participants strongly challenged the idea that human disease outbreaks only occur when food insecurity is already severe, describing disease outbreaks as both a cause and effect of severe food insecurity.

Human disease outbreaks were most commonly reported as causing large shortfalls in labour availability as community members became sick, spent time caring for the sick, died, or fled the area of an outbreak. Outbreaks during labour-intensive periods, such as planting or harvest seasons, had a more severe impact on food security.

Reviewing famine history in South Sudan: a case of famine in northern Unity State 1987 – 1992

The research noted a locally identified period of famine in what is now northern Unity State which started in the late 1980s and, because of a series of additional shocks, lasted until 1992. Yet this series of events, spanning a period of roughly five years, was not generally acknowledged as famine elsewhere. While the end result of these episodes would have likely matched the contemporary humanitarian definition of famine used today in IPC, it is worth considering these historical events as a series of IPC snapshots.

Throughout the 1980s and 1990s, agro-pastoralism in the area was repeatedly disrupted by armed conflict and large numbers of cattle were lost to widespread and recurrent raiding. Termed Ruon Murhaleen, this was the longest named time period identified in this research. In 1987 the area suffered a dry spell followed by Ruon Rahk, a severe crop pest outbreak. The following year, widespread harvest failure led to a period of severe hunger known as Ruon Nyakuojok, named for the green leaves that supposedly sustained many members of the affected population. Between 1988 and 1992, communities in northern Unity State described a kala-azar epidemic – a parasitic infection otherwise known as visceral leishmaniasis – leading to a substantial number of deaths, particularly in food-insecure areas. The spread of the disease was amplified by households fleeing the outbreak and searching for food. See Seaman, J., Mercer, A.J. and Sondorp, E. (1996) ‘The epidemic of visceral leishmaniasis in Western Upper Nile, Southern Sudan: course and impact from 1984 to 1994’, International Journal of Epidemiology, 25(4): 862–871 for an overall excess mortality estimate.

A snapshot of food insecurity in 1987 may have noted the dry spell and imminent crop failure, but not the steady loss of assets to raiding in the years prior. Similarly, a snapshot in 1989 may have struggled to connect the rapid spread of kala-azar with the displacement caused by food insecurity and conflict in earlier years.

Defining famine in a humanitarian crisis

When the IPC was used to declare famine in parts of Unity State in 2017, it was unclear if serious efforts had been made to ask the affected population whether they also considered the situation to be famine. Similarly, it is not clear to what extent local perspectives had been taken into account when evidence was deemed insufficient to declare a famine in December 2015 in the same areas.

It is worth noting that while the wording may vary across dialects, the Nuer phrase buoth mi diit e long roughly translates as ‘famine’, and is clearly distinct from the term for hunger associated with a typical lean season, buoth. During fieldwork conducted by one of the authors in Leer and Mayendit Counties in August 2018, discussions with Dok and Hak Nuer communities revealed that the word da̱k was used to refer to exceptional periods of forced displacement, violence by armed groups, and severe hunger, such as occurred in 2014 and mid-2018. This word was used to refer to situations in which many deaths occurred and where the causes of death were linked to violence, hunger and disease – rather than famine alone. These periods of da̱k, especially the one in mid-2018, were commonly described as much worse than the severe hunger experienced in 1987 during Ruon Nyakuojok. More research is needed to better understand local perceptions and terminology in order for it to be used effectively within the IPC analytical framework.

In addition to these findings, three lessons from the research process stand out:

Living memory as the pre-crisis baseline

Humanitarian programming in South Sudan benefits from a wealth of conventional pre-crisis information generated by the humanitarian and academic communities, including nationwide surveys and updated livelihood zone profiles compiled by the Famine Early Warning Systems Network (FEWNET). Operation Lifeline Sudan also left behind a trove of analysis when it wound down in 2005. The online Sudan Open Archive managed by the Rift Valley Institute houses much of this information (www.sudanarchive.net).

Although often overlooked, local knowledge also offers deep baseline information for use in historical comparisons. With reference periods stretching back as far as living memory and oral tradition allow, local people’s historical accounts allow comparisons of shocks and their impact across space and time, even across entire livelihood zones.

Local terminology

The definitions used by focus group participants to describe core concepts such as famine, hunger deaths and destitution were often fully in line with widely accepted external definitions used by humanitarians, such as those in the IPC Guidelines on key parameters for famine classification (2016). Importantly, participants used the term famine sparingly, describing situations as severe, but not severe enough to constitute ‘famine’.

Where local and external definitions match, the systematic usage of local terms may improve the quality of data from humanitarian assessments, ranging from rapid assessments to nationwide household surveys. Local definitions may also be applicable over wide geographic areas, such as livelihood zones, allowing for greater comparability.

Reconsidering magnitude and intensity

Participants compared historical famine experiences to the current crisis according to perceptions of magnitude and intensity, similar to earlier conceptualisations of famine adopted by the IPC. The magnitude of shocks was often expressed in both geographic and social terms, with participants noting how far over a given physical area a shock extended and which social units were affected, ranging from nearby settlements to larger sub-tribes occupying roughly a county. The social magnitude of shocks remains central to how communities view their options for coping through social and familial networks. Discussions of intensity of a shock’s impact largely focused on total deaths from all causes, including violence, disease, malnutrition and hunger. Questionnaires for quantitative and qualitative data collection could potentially accommodate comparisons of both magnitude and intensity between historical events and the present to gauge the severity of a situation after a new shock or a series of shocks.

From improved early warning to early action?

Greater and more regular incorporation of local knowledge and experiences into existing data streams and analysis would help humanitarian actors improve the timeliness and accuracy of early warning. However, even the most contextualised and accurate early warning is of no use if it does not translate into prevention and response.

From the Sahelian famine of the 1970s through the Somalia famine of 2011, early warning analysts have noted the frequent lack of effective reaction to their reports until famine mortality has already sharply risen. The nature of this ‘missing link’ remains beyond the scope of the present research. See Buchanan-Smith, M., and Davies, S. (1995) Famine early warning and response the missing link. Intermediate technology. London: ITDG Publishing. Practitioners seeking to incorporate local knowledge and experiences into famine early warning analysis, and working to link early warning mechanisms to early action and response, must remember that their analysis will only be acted upon if key decision-making bodies within a given humanitarian response understand and accept this analysis as a legitimate indication of famine risk or occurrence.

Achieving this is often a long and arduous process, as evidenced by Oxfam’s experience of working to establish a new early warning dashboard for Somalia. See Oxfam (2017) From early warning to early action in Somalia: what can we learn to support early action to mitigate humanitarian crises? Oxford: Oxfam. Today, the lessons learned from the process in Somalia are informing similar efforts in South Sudan. At its simplest, even if significant time and effort is spent on listening to the people most affected by the crisis, it accomplishes nothing if no one listens back.

This article was adapted from the full REACH report South Sudan – now the forest is blocked: shocks and access to food.

Chris Newton was previously the Food Security and Livelihoods Assessment Officer with REACH in South Sudan. He is currently with the World Food Programme. Katie Rickard is the Country Coordinator for REACH in South Sudan. REACH facilitates the development of information tools and products that enhance the capacity of aid actors to make evidence-based decisions in emergency, recovery and development contexts. All REACH activities are conducted through inter-agency aid coordination mechanisms. For more information, please contact the in-country office at southsudan@reach-initiative.org or the global office at geneva@reach-initiative.org.