Rural household resilience to food insecurity and its determinants: Damot Pulasa district, Southern Ethiopia

Food insecurity is a global issue that the world has been attempting to address for decades but has yet to achieve a conclusive victory. The objective of this research was to assess the current levels of rural household resilience to food insecurity and to analyze its determinants. Multi stage sampling technique was employed to generate data from 187 sampled household heads. The collected data were analyzed by using descriptive statistics, household resilience index, and econometric model. Household resilience to food insecurity were selected as the dependent variables and measured by household resilience index by using principal component analysis resulting from five variables. The study employed 18 Independent variables; ordered probit regression was used to identify the key determinants of household resilience to food insecurity. The findings of the survey results revealed that 57.2% were non-resilient, 18.2% moderately resilient, 15.0% resilient and 9.6% highly resilient. Ordered probit model estimation results showed family size, diversity of income sources, livestock ownership, use of improved seeds with fertilizer, access to credit services, and assistance received; were six significant explanatory variables that determinants of household resilience to food insecurity at 95% CI or P < 0.05 probability level. In terms of gender, the study reveals that households headed by men are more resilient than those headed by women. However, the resilience components including Income per AE (in birr per annum); crop stock of maize equivalent per AE (in quintal); cash savings per AE (in birr); food expenditure per AE (in birr per week); and precautionary assets per AE (in TLU) are the most correlated with the households' resilience to food insecurity. Based on the significant explanatory variables findings of this study, stakeholders should focus on the concerns of household resilience to food insecurity as large extents of residents were non-resilient in the study area.