If the poor are to benefit from economic growth, then they need the skills that are in growing demand, and the capacity to raise their productivity as smallholder farmers and micro-entrepreneurs. Yet, the poor seldom receive a satisfactory education. Too little is spent on primary education—the category of education of most direct benefit to the poor—while on average public subsidies to secondary education are roughly three times as high as subsidies to primary education, and subsidies to tertiary education are thirty times as high. In consequence, the higher income deciles benefit disproportionately from public spending on education—the share of the richest income quintile (28%) is roughly double that of poorest income quintile (13%) across countries. Why do such inequalities in public spending prevail? We argue that their wealth enables the affluent to buy favourable policies from politicians. In contrast, the poor lack the resources for lobbying and they face more severe collective action problems. We find strong empirical evidence for this interest group model of politics (as opposed to the median voter model which predicts a more redistributive pattern of public spending). We find that income inequality—which is a proxy for the political bargaining power of the rich versus the poor—is significant in explaining cross-country variance in the ratio of public spending on primary education to tertiary education. Holding everything else constant, a one standard deviation increase in the Gini coefficient would reduce the ratio of primary-school spending to tertiary spending by 0.20 percentage point. We also find that conflict is significant in skewing public spending away from primary education, and that increased ethnic diversity tends to reduce the relative share of public spending on primary schooling (although this effect may be mitigated if the political system is democratic rather than authoritarian). Our results raise some troubling issues for policy makers and aid donors. In particular, more attention must be given to reducing income inequality in order to reduce political constraints on pro-poor public expenditure reform (and on the effective implementation of the current wave of Poverty Reduction Strategy Papers). And reducing the prevalence of conflict would facilitate resource shifts from military spending to primary schooling, thereby lessening the need to introduce higher levels of cost recovery in the secondary and tertiary education sectors to facilitate higher investment in primary education. – poverty ; income distribution ; education ; development
