Estimating household vulnerability to tropical cyclones: An investigation of tropical cyclone shocks in coastal villages of eastern India

Abstract

In east-coast India, tropical cyclones are perennial phenomena. The conditions of rural coastal households remain precarious as they become vulnerable to poverty. The present study examines the vulnerability of the affected households to recurrent cyclones. The study is based on 504 sample households in selected coastal villages on the Indian east coast. We employ the three-step feasible generalized least-square (3FGLS) estimation in a multilevel model to get consistent and asymptotic efficient estimates of consumption and find their determinants. Similarly, we use the multilevel model to find the factors affecting the vulnerability of each household. We further decompose the estimated vulnerability into idiosyncratic and covariate, and poverty into chronic and transient. We also estimate the vulnerability severity and gap of the households. We find that the current poor are less than those at risk of becoming poor due to tropical cyclones. Fishing as the main occupation affects the most vulnerable population as they are exposed to high consumption volatility and poverty. Risk attitude forms an integral part of the coastal population, and risk-taker households are more vulnerable to cyclonic shocks. However, households with multiple sources of livelihood stay less vulnerable. Unlike other studies, we find covariate vulnerability almost equal to idiosyncratic vulnerability. From the policy perspective, we need to mitigate the covariate shocks by ex-ante mitigation interventions, which may help consumption smoothening for households having high consumption volatility. Households under chronic poverty need to be provided with social safety nets to protect themselves from falling into poverty traps.