The paper examines the relationship between household income and schooling costs in the presence of intrahousehold schooling bias against non-biological children. To this end, we construct a two-period model of intrahousehold schooling bias. The model predicts that there is an asymmetry in the impact of changes in costs and income on schooling in the sense that the impact is larger for the non-biological child. It predicts that the asymmetry increases as the relationship distance between the non-biological child and the parents gets wider. It also shows that an increase in cost of schooling leads to a bigger reduction in schooling for poor households, and that the difference in the impact of cost changes between the biological and the non-biological child declines as household income increases i.e. there is convergence. And the convergence is faster the more distantly related to the parents the non-biological child is. An empirical investigation of these predictions using the Second Malawi Integrated Household Survey (IHS2) data, shows that when current enrolment and grade attainment are used to measure schooling, the price and income elasticities of schooling are larger for non-biological children. The results also indicate that households in the lowest income quintile (the poorest) have the largest price elasticities, and households in the highest income quintile (the wealthiest) have the smallest price elasticities. We also find that the price elasticities for biological and non-biological children converge as we move from the lowest income quintile to the highest income quintile, and that the convergence is faster for non-biological children who are non-relatives.