Potential for Application of a Probabilistic Catastrophe Risk Modelling Framework to Poverty Outcomes : General Form Vulnerability Functions Relating Household Poverty Outcomes to Hazard Intensity in Ethiopia

Porter, Catherine ; White, Emily

World Bank, Washington, DC

关键词: POVERTY THRESHOLD;    LIVING STANDARDS;    RISKS;    HOUSEHOLD SIZE;    HOUSEHOLD SURVEY;   

DOI  :  10.1596/1813-9450-7717
RP-ID  :  WPS7717

学科分类：社会科学、人文和艺术（综合）

来源: World Bank Open Knowledge Repository

【 摘 要 】

This paper analyzes the potential tocombine catastrophe risk modelling (CAT risk modeling) witheconomic analysis of vulnerability to poverty using theexample of drought hazard impacts on the welfare of ruralhouseholds in Ethiopia. The aim is to determine thepotential for applying a derived set of damage(vulnerability) functions based on realized shocks andhousehold expenditure/consumption outcomes, onto aforward-looking view of drought risk. The paper outlines theCAT risk modeling framework and the role of thevulnerability module, which describes the response of anaffected exposure to a given hazard intensity. The need toexplicitly account for different household characteristicsthat determine vulnerability within our model is considered,analogous to how a CAT risk model would differentiate damagefunctions for buildings by different classes ofconstruction. Results for a regression model are presented,estimating ex-post drought impacts on consumption forheterogeneous household types (e.g. with cattle, safety-netaccess, illness). Next, the validity/generalizability of thederived functions are assessed, to infer applicability ofthe derived relationships within a CAT risk modellingframework. In particular, the analysis focuses on externalvalidity: whether the relationships established in thedataset can be used for forecasting outside of the sampleused for analysis. The model is stress-tested usingstatistical methods of resampling. This involves randomlysplitting the data into “training” and "testing"datasets. The tests show consistency of results across thedatasets. Finally, future plans are outlined with regard todeveloping a fuller catastrophe risk model to combine withthe consumption results.