A spatial analysis of risk and vulnerability in the COMESA region

The socio-economic condition in most COMESA countries is characterized by persistent high poverty levels and low food security. This is further compounded by the susceptibility of agriculture to climatic variability and the vulnerability of impoverished and malnourished households to HIV/AIDS, market shocks and prolonged violent conflicts. In response to these challenges, the Comprehensive Africa Agricultural Development Programme (CAADP) has been endorsed by the African Heads of State as a framework for the restoration of agricultural growth, food security, and rural development in Africa. At the same time the CAADP agricultural growth agenda must ensure that the marginalized are the ultimate beneficiaries of growth and are not further marginalized by rapid developments. It is therefore crucial to include a special focus on those with immediate needs to address hunger and malnutrition.

The purpose of this study is twofold. On the one hand it is providing policy makers with a collation of baseline information on risks and vulnerability in the COMESA region (type of shocks and stresses, and population groups and geographical areas most affected by them). The presented maps, tables and graphs can inform policy discussions and eventually improve resource allocation. On the other hand, the study aims to improve the understanding of the sources and consequences of vulnerability and how they differ across space and endowments. The mapped and characterized risks and vulnerabilities will be used to geographically target future research to enhance the understanding of the co-evolution of risk profiles, coping strategies and socio-economic development.

The vulnerability definition adopted for this study is the exposure to risks, mitigated by the ability to cope. Vulnerability is thus comprised of the risks that people confront face in pursuit of their livelihood, the risk response or the options that people have for managing risks and finally the outcomes that describe the loss in well-being. The risk response or available risk management options are determined by livelihood assets, strategies and policy and institutional environments. In this framework, vulnerability has to be seen as a dynamic process that represents the conditions set by the environments they inhabit and the choices of the vulnerable populations themselves.

The study builds on the work of Thornton et al. “Mapping climate vulnerability and poverty in Africa” (ILRI, 2006), but focuses more rigorously on the risk component of the vulnerability framework. Risk cannot be avoided and at the same time risks continue to evolve and change. Assuming and managing risk is therefore at the core of any decision-making process. The generation of socio-economic profiles of geographical areas and communities, including their risk profile, is beneficial when response measures are considered. Given the recurrence of certain types of disasters (e.g. floods and droughts) it makes even more sense to generate risk maps of the most prevalent shocks and stresses and their potential impact on communities.

From the literature the main sources of risks were identified and grouped into five categories: natural disasters, pest and diseases, human health, socio-economic risks and governance and conflict risks. For each of the identified risk criteria in these categories spatially disaggregated data were collected and probability maps developed. Thereafter, multi-criteria evaluation (MCE) techniques were applied to combine these surfaces into risk hotspots. Two different methods of assigning weights to the risk criteria were applied and compared. Firstly, principal component analysis was used to scale down the original list of criteria to an operational, non-redundant set and weights assigned according to the variance explained before combining the different variables into one map. Secondly, expert opinion and pair-wise comparison was used and compared with the results from the PCA-assigned weights. The resulting hotspots of risk were characterized and stratified in terms of farming systems, market access, population density, poverty and malnutrition.

Although progress is made in the identification and mapping vulnerable people and locations, the state of knowledge and methodologies are still limited. Especially the sub-country level mapping techniques are rather novel and in its initial stages. This study shows that the innovative combination of a variety of spatial data layers, multi-criteria analysis and statistical techniques can yield useful results that enable policy makers and other stakeholders to compare, understand and analyze risk and vulnerability components within and across national boundaries and integrate this in adaptive planning systems.