Purpose
Enhancing household consumption and reducing inequality are among the fundamental goals of many developing countries. The purpose of this study therefore is to disaggregate household consumption expenditure into food and non-food and, thus, decompose inequality into within- and between-groups.
Design/methodology/approach
The study adopts generalised entropy (GE) measures. Second, the study uses regression-based inequality decomposition to ascertain the determinants of inequality in food and non-food expenditure using household demographic and socioeconomic characteristics as covariates.
Findings
The results show that non-food expenditure is the major source of inequality in household consumption expenditure in both urban and rural areas with inequality coefficients of above 0.6 compared to about 0.4 for food expenditure. The decompositions also show that within-group inequalities for non-food and food expenditure are, respectively, 0.97 and 0.365 using the Theil index, while between-group inequalities for non-food and food are, respectively, 0.016 and 0.035. Furthermore, the regression-based inequality decompositions show that variables such as living in rural areas, household size, household dwelling and household dwelling characteristics account for the significant proportion of inequality in food and non-food expenditure.
Originality/value
The policy implication of the findings, among others, is that policies should focus on addressing inequality within rural and urban areas, especially with respect to non-food expenditure than in inequality existing between urban and rural areas. These non-food expenditures include expenditure in education, health, energy, accommodation, water and sanitation.