Natural disasters amid complex political emergencies
by Margie Buchanan-Smith, independent consultant, and Ian Christoplos, ODI Research Associate July 2004

There is a sharp operational and conceptual distinction between natural disasters and complex political emergencies (CPEs). Yet this is often inappropriate: natural disasters are rarely truly ‘natural’, while many areas suffering from complex political emergencies are also subject to periodic natural hazards. In the last five years, at least 140 ‘natural’ disasters have occurred in countries experiencing complex political emergencies.

Operating at the interface between natural disasters and complex political emergencies raises some difficult challenges. How can a humanitarian emergency caused by a combination of conflict and natural hazard best be assessed and portrayed? How can a more holistic and flexible approach to early warning and analysis, that is sensitive to both natural disasters and to conflict, be promoted? Whose responsibility is preparedness, especially if there is no functioning state? These and other issues were debated at a one-day seminar hosted by the British Red Cross Society (BRCS) in November 2003, entitled ‘Natural Disasters and Complex Political Emergencies’. This article summarises the key conclusions that emerged.

The interface between natural disasters and CPEs In advance of the seminar, four case studies were prepared of situations where a natural disaster coincided with a CPE, or at least with conditions of political instability:

  • the Bahr El Ghazal famine in South Sudan in 1997 to 1998;
  • the drought in Afghanistan between 1999 and 2001;
  • the combination of natural disaster and conflict in Colombia (the Eje Cafetero (‘Coffee Belt’) earthquake in January 1999 and recurrent flooding in Atrato Medio); and
  • the earthquake in Gujarat in India in 2001, followed by communal rioting in Ahmedabad in 2002.

In all of these situations, it was usually the indirect effects of conflict that increased the impact of the natural disaster. Where conflict has had a long-term impoverishing effect, the asset base has been weakened, substantially reducing the viability of local coping strategies and therefore local resilience to a natural disaster. By the time the drought hit Afghanistan, most of the population had suffered years of conflict. In Bahr El Ghazal, three years of drought and many years of conflict-related impoverishment had left the population in a state of high vulnerability by the end of 1997. Rebel attacks in early 1998 caused massive displacement from towns into rural areas already facing acute food insecurity, thus triggering a famine marked by some of the highest malnutrition and mortality rates ever recorded. In Colombia, the increased risk of natural hazards has been linked to the weakening effect of conflict on local institutions. The lack of institutional capacity to enforce land use regulations has allowed the unchecked exploitation of forests, increasing the risk of floods and landslides.

Vulnerability assessments – a key role to play

A greater commitment to broad-based vulnerability assessments is the key to strengthening the ability to respond in an appropriate and timely fashion to compound emergencies. Assessments constitute an opportunity to raise attention to both the risks of natural disaster, and the risks associated with violent conflict and/or political marginalisation. Yet in conflict areas there is a tendency to overlook the risk of natural disaster, and how local capacity to cope and respond may have been weakened by years of conflict.

In a more peaceful environment, a vulnerability assessment focused on natural hazards should also capture the underlying social and political factors that make some groups particularly vulnerable. Yet this kind of holistic approach is rare, either because the emergency is narrowly defined, or because of a failure to look beyond assumptions about the direct destructive impact of a climatic or seismic event. After the Gujarat earthquake, many agencies were slow to pick up on the underlying patterns of discrimination and unequal power relations in society. As a result, the poorest and most marginalised groups often received the least by way of relief and rehabilitation resources. Exploitative social structures and power relations were simply reproduced, with even more devastating consequences as limited relief and rehabilitation resources were captured by the better-off, unless positive and sometimes controversial steps were taken.

South Sudan offers an example of good practice. There has been long-term investment in information gathering and analysis, particularly through the food economy work of Save the Children-UK (SC-UK) and WFP. In 1998, SC-UK commissioned a seminal vulnerability study, providing insight into the causes of vulnerability amongst the Dinka and how these were changing. This report has been significant in helping aid workers rethink their intervention strategies, particularly their approach to targeting.

The challenge of early warning

Understanding these longer-term dynamics is critical for early warning, so that when a crisis is imminent the early warning system can communicate why ‘this year is different’, in other words why an acute crisis is impending amidst an extended period of chronic distress. This can be particularly challenging when there are multiple and complex causes of the crisis.

In Bahr El Ghazal, the long-term investment in information-gathering paid off, at least in terms of understanding what was happening. More than nine months before the famine, SC-UK issued a statement that parts of South Sudan were heading for the worst year since Operation Lifeline Sudan (OLS) had begun in 1989. But the impact of the long-term erosion of livelihoods and coping strategies was hard to convey. Donor governments, looking for simple, concise messages, were not well-disposed to receiving and acting upon these early warnings of famine. Significant resources were only made available when there was hard evidence of widespread starvation. Where conventional early warning had failed, shocking media coverage provided the trigger, but by then in response to full-blown famine.

So what can an early warning system do to convince sceptical decision-makers (particularly donors but also NGOs) to respond, in time? First, when trying to persuade a sceptical donor audience to respond, NGOs and other agencies must be coordinated and coherent in their early warning messages – a particular challenge in a complex crisis where the causal linkages may be contested and hard to predict. Second, agencies which have done the early warning analysis must be prepared to engage in effective and sustained advocacy in delivering the early warning message, even more so in a hostile political environment where donors are reluctant to respond.

Conditions required for preparedness

Preparedness to implement mitigating interventions requires a conducive funding environment, adequate skills and clarity about who is leading preparedness initiatives. Afghanistan offers an example of how difficult it may be to guarantee all of these. Prior to 11 September 2001, preparedness was reasonably strong amongst international aid agencies in terms of skills and experience. But as long as the Taliban regime was in power, it was difficult for aid agencies to access donor funding to tackle the underlying issues that made people so vulnerable to the drought. By the end of 2000, the scale of the drought was close to overwhelming all the agencies involved. After the Taliban government fell, the funding environment changed, and there was a huge influx of donor resources. At the same time, there was a massive turnover of international staff and poaching of skilled Afghan staff by the UN, private contractors and embassies. This weakened the ability of NGOs to do in-depth vulnerability assessments at just the moment when the financial resources were available. When the new UN mission to Afghanistan, UNAMA, was set up in early 2002, the influx of international staff meant that many national staff were effectively demoted. These new arrivals often lacked local knowledge and the experience of negotiating in Afghanistan.

Institutional distinctions between natural disasters and CPEs

Often, agencies become lopsided in their focus on one type of emergency rather than on both. In Atrato Media in Colombia, for example, some humanitarian agencies decided to focus on conflict alone, even when they were working in an area affected by natural disaster in the form of recurrent flooding. The responsibility for responding was left to state institutions or the Colombian Red Cross. Yet the response to the floods could have been strengthened by the involvement of these humanitarian agencies. They are often in a position to respond more rapidly and effectively than the state, have teams already in the area, and have knowledge of the local context.

Colombia also offers a positive example of humanitarian agencies taking a more integrated approach. After the earthquake in the Coffee Belt, a number of agencies sought to ensure good communication and transparency, to minimise the likelihood of armed groups interfering in their work. Caritas, for instance, paid special attention to the standard of housing they were building, recognising that new houses could become a source of conflict. They also included some beneficiaries who had not been directly affected by the earthquake, but were nevertheless equally vulnerable, again to reduce the likelihood of conflict.

Closing the gap between the natural disaster and CPE discourses: a way forward

The strong message that emerged from the case studies and the seminar discussions is that there is a need for aid agencies to strengthen their contextual – and especially political – analysis in these compound emergencies, regardless of whether they are labelled ‘natural disaster’ or ‘complex political emergency’, and to strengthen preparedness for all types of emergency. The humanitarian discourse has become overly focused on CPEs, just as the natural disaster discourse has remained tied to technical risk reduction at the expense of understanding the political complexities that are inevitably involved. If policies are to become more coherent in reducing vulnerability, these barriers must be reduced.

A more comprehensive analysis of risk, that breaks down the distinction between natural hazards and conflict, is one way of achieving this. Vulnerability analyses associated with natural disasters could break out of technocratic ruts by drawing lessons from the growing focus on political economy analyses in CPEs. Those working with CPEs could learn about the practical challenges facing disaster-affected people from those working with natural hazards. The common denominator for both is the need to focus on the risks faced by vulnerable people.

Seminar participants were strongly persuaded of the unifying potential of the vulnerability assessment, particularly if it promotes understanding of the underlying causes of vulnerability, on top of which an analysis of immediate causes of the emergency can be overlaid. This is possibly the most productive area for research, to strengthen vulnerability assessments (and other analytical tools) to capture the interconnectedness and complexity of ‘natural-disaster-plus-conflict’ in order to elicit a more timely and appropriate response.

Key questions to be addressed would include:

  1. How does conflict/political instability affect vulnerability to natural disasters, at household, district and national levels? How can this best be measured?
  2. How can risk assessment be broadened from a technical approach to incorporate social and political factors?
  3. How can practitioners communicate their findings to decision-makers convincingly, conveying the complexity of a compound crisis in an accessible and convincing way?
  4. What kind of funding support is required for this type of analysis/assessment, which takes a long-term perspective?

This agenda would lend itself to real-time research in countries where conflict and political instability are having an impact on the population’s vulnerability to recurrent natural disasters, for example in Indonesia, Nepal or Zimbabwe. In all of these countries the impact and interconnectedness of conflict and natural disaster are not sufficiently understood, nor are adequate preparedness measures in place.

References and further reading

M. Buchanan-Smith and I. Christoplos, ‘Natural Disasters Amid Complex Political Emergencies’, Report on a Seminar Hosted by the British Red Cross Society (BRCS), 2004. Background papers for the seminar are available from the BRCS.

I. Christoplos, ‘Natural Disasters, Complex Emergencies and Public Services: Rejuxtaposing the Narratives after Hurricane Mitch’, in P. Collins (ed.), Applying Public Administration in Development: Guideposts to the Future (Chichester: John Wiley and Sons, 2000).

S. Harrigan and Chol Changath Chol, ‘The Southern Sudan Vulnerability Study’ (Nairobi: SC-UK, 1998).

Independent Evaluation of Expenditure of DEC India Earthquake Appeal Funds. January 2001–October 2001 (London and Ahmedabad: HI, DMI and Mango, 2001). John Twigg, Disaster Risk Reduction: Mitigation and Preparedness in Development and Emergency Programming, Good Practice Review 9 (London: ODI, March 2004).

J. Darcy and C.-A. Hofmann, According to Need: Needs Assessment and Decision-making in the Humanitarian Sector, HPG Report 15 (London: ODI, 2003).