The paper investigates two issues regarding household expenditure on primary education of own children using the Second Malawi Integrated Household Survey (IHS2) data. Firstly, we look at factors which influence a household's decision to spend or not (the participation decision), and by how much (the expenditure decision). This is done for urban and rural households. We fi?nd that there are differences in the factors which influence both decision levels for the two groups of households. Secondly, to get a deeper understanding of these rural-urban spending differences, the study develops the Blinder-Oaxaca decomposition technique for the independent Double Hurdle model. The proposed decomposition is done at the aggregate and disaggregated levels. The aggregated decomposition allows us to isolate the expenditure differences into a part attributable to differences in characteristics and a part which is due to differences in coefficients. The detailed (disaggregated) decomposition enables us to pinpoint the major factors behind the spending gap. At the aggregate decomposition level, our results show that at least 66% of the expenditure differential is explained by differences in characteristics between rural and urban households, implying that an equalization of household characteristics would lead to about 66% of the spending gap disappearing. At the disaggregated decomposition level, the rural-urban difference in household income is found to be the largest contributor to the spending gap, followed by quality of access of primary schools. Besides, rural-urban differences in mothers education and employment are found to contribute more to the spending differential relative to the same for fathers.