ASSESSING VULNERABILITY TO POVERTY FOR HOUSEHOLDS IN WEST AFRICA: EVIDENCE FROM RURAL GAMBIA

Abstrak

This study assesses household vulnerability to poverty in rural Gambia using the Vulnerability as Expected Poverty (VEP) approach, based on cross-sectional data. Income and consumption were used interchangeably to estimate the probability that a household currently above the poverty line may fall below it in the future. A three-step Feasible Generalized Least Squares (FGLS) method was applied: first, estimating the ex-ante mean using OLS; second, modeling variance from squared residuals; and third, correcting for heteroskedasticity to predict vulnerability. Findings show that household size has a positive and significant effect on expected log income, while its squared term is negative and significant, indicating diminishing returns. Age of the household head has a slightly positive but statistically insignificant effect on income, with the squared term weakening this impact. Employment status is positively associated with income, where a one-unit increase leads to a 2.9% rise in expected log income and reduced vulnerability. Educational attainment also shows a positive, though insignificant, effect. Non-land production assets significantly reduce vulnerability, with a one-unit increase lowering it by 18.8%. The share of irrigated land contributes to a 4.4% reduction in vulnerability. To reduce vulnerability in rural Gambia, government efforts should focus on expanding access to irrigation and productive assets, especially for smallholder farmers. Employment programs targeting rural youth and women can enhance income stability. Strengthening vocational education and improving access to agricultural extension services will also help households build resilience and reduce future poverty risks.