Tekalign Gutu Sakketa is a researcher at the German Institute of Development and Sustainability (IDOS), Germany. He was a visiting fellow at Cornell University as part of the Structural Transformation of African Agriculture and Rural Spaces (STAARS) program. Co-authors include Dan Maggio (Cornell University), and John McPeak (Syracuse University).
Pastoralists rely on animals for their livelihoods, making them susceptible to climate risk. Extreme weather events, particularly prolonged and recurrent droughts, can force pastoralists into poverty traps as livestock die or degrade in value. These droughts are likely to induce food insecurity and may intersect with additional hazards, such as those presented by instances of violent conflict. One can easily envision how these interrelated occurrences may impede the economic progress of the affected populations.
To address this challenge, policymakers and stakeholders have long sought cost-effective and scalable risk-reduction strategies. Recently, index-based insurance products have gained popularity among the interventions to improve resilience. In a new working paper titled The Protective Role of Index Insurance in the Experience of Violent Conflict: Evidence from Ethiopia, we study the effects of one such intervention on the risk of violent conflict. This intervention, Index-based Livestock Insurance (IBLI), was developed and expanded to the Horn of Africa to allow pastoralist populations to hedge the risk of livestock mortality created by adverse weather events. We ask if it may also help pastoralists hedge conflict risk. Indeed, we find that IBLI has the potential to be an effective protective factor in fragile and conflict-affected settings.
Against this background, we ask whether households with insurance are less likely to experience violence. We also study whether bad weather increases conflict risk among herders. We use a household panel survey associated with the IBLI pilot collected from 520 pastoralists and agro-pastoralist communities in southern Ethiopia over four years (2012-2015). We also utilize administrative data on IBLI purchases, remotely sensed datasets on rangeland conditions in the IBLI index areas, and geocoded conflict events data obtained from the Armed Conflict Location and Event Data Project (ACLED).
The International Livestock Research Institute (ILRI) implemented the IBLI pilot program in Southern Ethiopia in eight woredas of the Borana zone. The scheme provided insurance to families who own herds of animals (cows, camels, goats, sheep), indexed by the amount of vegetation available to in their area. When a composite index showed insufficient vegetation for the animals to graze, the insurance company paid money to the insured households. The insurance was sold biannually, before the start of the short rainy season and long rainy season, and coverage lasted for 12 months (see Figure 1).
To encourage uptake and create exogenous variation in the program uptake, households in the survey area were randomly given non-transferable, discount coupons before each insurance sales period that could be used when buying a contract.
Our findings demonstrate evidence of a robust association between rangeland conditions, conflict experience, and program adoption. Households that have faced drought are at a higher risk of conflict and exhibit a greater propensity to acquire insurance coverage in the following season. Specifically, a one-unit decrease in rangeland conditions per index area, proxied by the standardized normalized difference vegetation index (zNDVI) in the previous season, is associated with a 0.3-1.7 percentage point increase in the probability of household conflict exposure.
Our second goal is to demonstrate the protective role of IBLI against conflict. However, since there are other factors that are likely to affect both IBLI participation and conflict exposure; it makes it difficult to determine whether the observed decrease in conflict is truly due to the IBLI adoption or if it is the result of these other factors. Failure to take into account these factors can lead to biased estimates of the impact of IBLI participation, making it difficult to draw accurate conclusions about its effectives. To overcome this challenge, we instrument for IBLI uptake in our estimations using randomly distributed discount coupons that were distributed as part of the pilot program’s encouragement design and have been shown to be predictive of program uptake. Using this design and implementing a novel method for causal mediation analysis, we find that IBLI is indeed protective against conflict and reduces the conflict risk created by drought by 17 to 50 percent.
Overall, our findings suggest that social protection programs such as index insurance, can serve as a protective measure in areas where households face both climatic and conflict risks, which themselves may interact. We will continue to study whether we find comparable effects of this intervention are observable in the context of Kenya pastoralists, where IBLI is also in operation. These forthcoming analyses will aid in providing further insight into the role of index insurance programs in reducing conflict and fostering social cohesion.
