The economics of early response and disaster resilience: lessons from Kenya
by Catherine Fitzgibbon January 2013

In recent decades the drylands of the Horn of Africa have become one of the most disaster-prone regions in the world. Drought in particular affects more people, more frequently than any other disaster. Drought periods were not always so disastrous but, combined with the region’s underlying economic, social and environmental vulnerability, the impacts upon dryland inhabitants are extreme. Despite calls for greater investment in preparedness, early response and long-term resiliencebuilding, the 2011 drought crisis in the region illustrates how this has not yet been translated into reality.

It is an intuitive belief that investment in early response and resilience-building in drought-prone communities is more cost-effective than funding ever-increasing humanitarian responses. Yet little solid data exists to support this claim. To this end the UK’s Department for International Development (DFID) commissioned a study to examine the economic case for investment in early response and resilience-building in disaster-prone regions.+The report and related documents are on the DFID website at http://www.dfid.gov.uk/What-we-do/Key-Issues/Humanitarian-disasters-and-emergencies/Resilience/Economics-of-Early-Response-and-Resilience The study had two main components:

  1. To compare the costs of three different approaches or ‘storylines’:
    • late humanitarian response
    • early humanitarian response
    • building resilience to disasters
  2. To identify the types of interventions that provide the best value for money in building resilience to disasters.

The study looked at Kenya and Ethiopia, with a specific focus on the pastoral lowlands typical of many droughtaffected areas in the wider region. This article briefly outlines the findings of the economic analysis of the different scenarios from Kenya.

Approach to economic analysis

fitzgibbon table 1For each country the economic analysis was undertaken from both a top-down and a bottom-up perspective, using the three scenarios listed above. Top-down analysis was based on macro-economic data on emergency response and national drought eradication plans. The bottom-up analysis focused on a target community – in this case the 367,000 people that make up the Wajir southern grasslands pastoral household economy zone. Each of the three storylines was modelled using Household Economy Analysis (HEA) assessment data.+HEA is a livelihoods-based framework for analysing how people obtain the things they need to survive and prosper. It was designed to help determine people’s food and non-food needs in response to an event such as drought, and is built up based on household-level evidence. The HEA data was used in conjunction with the herd dynamics model developed by the Food Economy Group (FEG), who formed part of the study team. The FEG study comprises one of the background reports. Annual estimated response costs were then modelled over 20 years (ten years in some cases) and discounted using a rate of 10%. It was also assumed that a high-magnitude drought occurs every five years. This was a conservative assumption, as droughts are estimated to have a 3–5-year return period in both countries.

Analysis of findings

The study concluded that early response is far more costeffective than late humanitarian response. This was evident from the economic analysis for both countries from both the bottom-up and top-down perspective. As the figures show, storyline B2, which combines timely commercial destocking with other livestock support measures, is particularly cost-effective. The study found the figures to be consistent even when using highly conservative assumptions of costs and drought incidence.

This finding is not a great surprise as it reinforces a very evident rationale. Early response ensures that assistance arrives before households have to resort to negative coping strategies such as selling productive assets like core breeding stock. In pastoral economies facilitating early destocking (via commercial sale) of quality animals emerges as a particularly effective way to reduce aid costs. If pastoral households can convert high-value animals into cash before their condition declines they can use the income to maintain the condition of their remaining animals and feed themselves without food aid.

Another key factor was the inflated cost of buying food aid during a crisis, as against buying it beforehand. The study estimated that food (and cash) transfers usually represent 60-80% of total humanitarian assistance, so the combined effect of purchasing cheaply earlier and reducing the number of people in need drastically reduces overall costs. This is an important finding in addressing the reluctance of many governments and donors to release humanitarian funds early in response to early warning reports for fear that they may be funding a ‘non-disaster’. In fact, the study points out that donors could mistakenly fund two early responses in Kenya, and seven in Ethiopia, before the cost matches that of even one late humanitarian response.

The more comprehensive storyline, C, which combines early response with resilience-building, was slightly more expensive than storyline B2. However, resilience-building also emerges as significantly cheaper than late response. Again this is not surprising as early response and resiliencebuilding interventions both work to protect and build the asset base of vulnerable communities. In time this reduces the caseload of ‘vulnerable’ households that form the basis of humanitarian responses.

This being the case, why is there such limited investment in resilience-building? There are two key reasons. Firstly, investing in the key basic services and infrastructure that build resilience, such as water supplies, education and roads, is eye-wateringly expensive – in the short term. The Kenyan government, in common with others in the Horn of Africa, simply does not have these budgets to spare. Even if the money could be found, it is doubtful whether the political will exists to allocate it to the arid regions of Kenya, which are the most sparsely populated part of the country. Although there is an inherent understanding that such investments bring positive development gains, very little economic evidence exists to quantify the financial benefits and returns. The study recommends further work in this area. The second reason why resilience-building is not funded at the levels required is the fact that this cost must be added to the regular and significant costs of humanitarian response. This is because the number of vulnerable households dependent on food aid (drought year or no) would take some years to decline. This double-whammy of costs usually acts to undermine or reduce the ambition of any resilience-building plans.

The value for money of different resilience-building interventions

fitzgibbon table 2 fullThe second element of the study looked at the comparative value for money of different resilience-building activities regularly funded by donors. From a DFID perspective this entails examination of economy, efficiency, effectiveness and cost-effectiveness. The study did not involve a full cost–benefit analysis, but identified key issues in each sector. These are summarised in Table 2.

Some general value for money issues emerged in all sectors. The same intervention may be more or less cost-effective depending on the specific context. Participatory approaches, unsurprisingly, were found to maximise benefits and hence provide better value for money. ‘Soft’ resilience measures that build human capital and skills that can be reapplied are often more cost-effective than ‘hard’ capital projects. For example, teaching soil and water conservation and water harvesting techniques provides longer-term benefits than the construction of a borehole without the appropriate management and technical support. Other factors affecting value for money included:

  • Use of the private sector, which can ensure that private rather than public resources are used to build resilience. This is especially pertinent where communities already pay for services, such as animal health care.
  • The value for money of an intervention may change depending on the timescale over which it is evaluated. In the longer term certain initially expensive investments can emerge as very good value for money. Short funding timeframes may not facilitate this. For example, a teacher training/up-skilling project may not last long enough for increases in attainment and hence the earning capacity of students to be monitored.
  • An intervention that builds the resilience of one group may undermine the resilience of another; irrigation for agriculturalists, for instance, may reduce the grazing areas of pastoralists.

Catherine Fitzgibbon is a freelance consultant based in Kenya. She acted as the Kenya expert for this research.

Share
FacebookTwitterLinkedIn