naulu2020/poverty-analysis-eicv5-rwanda

# Rwanda Household Poverty Analysis ## Overview This project analyzes the determinants of household poverty in Rwanda using data from the Integrated Household Living Conditions Survey (EICV5). A logistic regression model is applied to examine how household and demographic characteristics influence the likelihood of being poor. Results are interpreted using average marginal effects. ## Dataset - Source: Integrated Household Living Conditions Survey (EICV5), NISR - Type: Cross-sectional household survey - Observations: 14,580 households Sample characteristics: - 64.16% of household heads are married - 76.7% of household heads attended school - 74.46% of household heads are male and 25.54% female - 91.56% of household heads are non-disabled ## Variables The dependent variable is household poverty status (1 = poor, 0 = non-poor). Independent variables include region (urban/rural), education of the household head, sex of the household head, household size, marital status, and disability status. ## Methodology The study uses a logistic regression model estimated with survey weights and robust standard errors. Average marginal effects are computed to interpret results in terms of changes in the probability of being poor. ## Key Findings Households in rural areas, female-headed households, and larger households are more likely to be poor. Education of the household head significantly reduces the likelihood of poverty. Marital status and disability status are not statistically significant but are included as control variables. ## Conclusion The results highlight the importance of rural development, education, gender-sensitive policies, and household size management in reducing poverty in Rwanda. ## Notes This project was completed as part of econometric coursework and is shared to demonstrate applied data analysis using real-world survey data.