In line with the rise of the "social investment" state, education has become a more important arena in welfare state reform. A large literature has studied the micro foundations of welfare states, but surprisingly, the study of individual preferences on education policy has largely been ignored so far. The paper is relevant for the topic of the panel in that it argues that understanding the micro-level dynamics of education policy preferences, and, in particular, their cross-level interaction with macro contexts has important implications for future reforms of welfare states in general. The paper provides a first step in this direction, engaging in a multilevel analysis of survey data for a large number of OCED countries. The paper argues that both the level of socio-economic inequality as well as the degree of educational inequality shape individual preferences over education spending. More specifically, I find that an increase in socio-economic inequality leads to an increase in overall support for education spending as well as a more pronounced negative effect of income on preference on the micro level, i.e. the rich/poor are more likely to oppose/support spending increases. In contrast, high levels of educational inequality mitigate the negative impact of income on support for spending. High levels of educational inequality imply more restricted access to higher levels of education for people with low-income backgrounds. Therefore, the rich/poor are more likely to support/oppose spending increases when levels of educational inequality are high.