| When will a government make policies to improve lives and spur development? This dissertation questions existing answers and proposes a new answer to this old question of political economy. I develop a new theory of how governments decide to make education policies, and test that theory with a half-century or more of education policymaking in three very-different developing countries: Taiwan, Ghana, and Brazil. These countries collectively provide settings in which common political-economic explanations of policymaking should do well: all have been through one or more regime transitions; they cover the full range of inequality in the world; and one, Taiwan, is an economic success story often touted for its wise economic governance. Yet regime type, collective action, and economic optimality fail to explain education policymaking in any of the three countries. Instead I find that two factors jointly determine education policy. The first is the success of political entrepreneurs at organizing the poor into a political force. This determines whether the government will depend for its power on poorer groups, more elite groups, or a cross-class alliance, and thereby determines most of the key features of the education system the government will try to create---whether the system will concentrate on primary education, higher education, or try to provide all levels, as well as sorts of schools, teachers, exams, fees, and access restrictions in the system and the role of the private sector. The second factor is the flexibility of the labor market faced by employers who need skilled workers, which determines the nature and extent of any worker training the government will try to provide. The evidence from Taiwan, Ghana, and Brazil supports this theory in very different contexts: relying on extensive data and field work from the three countries, I show that, over a half-century or more, governments facing similar conditions of political entrepreneurship and employer demands adjusted their education systems similarly, regardless of other factors. The results provide a new, more-accurate method of explaining and predicting policymaking, and point out the importance of factors relatively neglected by development research, such as the organization of the poor by political entrepreneurs. |
