Factors Affecting Poverty in Nepal - A Binary Logistic Regression Model Study

One of the key factors in reducing monetary poverty is the identification of its determinants. Using a logistic regression model and considering household poverty status (poor/non-poor) as the response variable, this paper attempts to identify the most promising factors associated with monetary poverty based on nationally representative data of 5,988 households from the Nepal Living Standard Survey (2010/11). The goodness of fit, classification, discrimination, and diagnostics of the fitted model is performed. Six factors, namely illiteracy of household head (OR: 2.20; 95% CI: 1.86-2.61), households receiving no remittance (OR: 1.90; 95% CI: 1.64-2.20), households with no landholdings (OR: 1.53; 95% CI: 1.31-1.78), households with poor access to market centers (OR: 1.77; 95% CI: 1.52-2.07), households having more than two children under the age of 15 (OR: 4.69; 95% CI: 4.06-5.42) and households having no literate persons of working age (OR: 1.29; 95% CI: 1.07-1.56) are significantly associated with the likelihood of poverty. Male-headed households are not better positioned than female-headed households concerning poverty level. The developed regression model has satisfied the test of goodness of fit of the model and reasonably satisfied the regression diagnostics through visual assessment. As several risk factors associated with poverty increase, the likelihood of a household being poor increases substantially. This analysis is expected to be helpful for the concerned authority to reframe the policy.