Investments in education: A political economy approach

Author: 
Hasnain, Zahid
Year: 
2000

     This dissertation attempts to explain why certain developing countries invest more than others in basic education. Given that numerous studies have established the importance of education in improving growth and labor productivity, and reducing fertility and infant mortality, this question is crucial to explaining why certain countries are poorer than others. The analysis examines the role of politics and policies in explaining differences in educational outcomes. Two sets of arguments are advanced: one for democracies and one for dictatorships. In democracies, the analysis focuses on the interaction between voters and legislators. It is assumed that the majority of the electorate demands education, that a legislator's primary concern is to be elected to political office, and analyzes how the pursuit of this goal can, paradoxically, lead to under-investment in education. The theoretical argument shows how party structure, specifically the extent to which political parties offer their members stable careers, shapes the incentives of legislators and determines these investment decisions. The validity of this argument vis-a-vis some standard alternative explanations is then tested on two states in India--Uttar Pradesh and Tamil Nadu. For dictatorships, it is assumed that the leader maintains power with the support of a small group of elites, usually the upper echelons of the military or civilian bureaucracies, and the analysis investigates how investments in education affect these actors' self-interest. It is assumed that the dictator cares about staying in power, and that bureaucrats care about advancing their careers. It is argued that the scale of ethnically motivated distributional conflict among the ruling elite determines public investments in education. The plausibility of this claim is then empirically examined in the case of the military dictatorships in Indonesia and Nigeria.

Advisor(s): 
Laitin, David
Department: