Poverty Imputation in Contexts Without Consumption Data: A Revisit With Further Refinements
利用埃塞俄比亚等五国的14项家庭调查数据，改进了调查间插补模型，发现加入家庭公用事业支出能准确估计贫困率，优于常用插补和机器学习方法，并揭示了空间异质性。
Survey‐to‐survey imputation has been increasingly employed to address data gaps for poverty measurement in poorer countries. We refine existing imputation models, using 14 multi‐topic household surveys conducted over the past decade in Ethiopia, Malawi, Nigeria, Tanzania, and Vietnam. We find that adding household utility expenditures to a basic imputation model with household‐level demographic and employment variables provides accurate estimates, which even fall within one standard error of the true poverty rates in many cases. The proposed imputation method performs better than several commonly used multiple imputation and machine learning techniques. Further adding geospatial variables improves accuracy, as does including additional community‐level predictors (available from data in Vietnam) related to educational achievement, poverty, and asset wealth. Yet, within‐country spatial heterogeneity exists, with certain models performing well for either urban areas or rural areas only. These results offer cost‐saving inputs into future survey design.