Empirical Analysis of Education Demand: Costs, Returns, and Policy Interventions

The lecture delves into the empirical analysis of the demand for education, positing that educational choices are a function of the costs of education (marginal cost of human capital acquisition) and the returns to education (financial wages and human capital for the next generation). The speaker emphasizes the importance of understanding both actual and perceived benefits. The discussion begins by examining direct costs, such as school uniforms, and indirect costs, like the opportunity cost of schooling, using examples from their own research in Kenya and Indonesia. A key argument is that people respond to costs, but sometimes in ways that challenge the simple economic model, suggesting the presence of credit constraints or other behavioral factors. A significant nuance is the distinction between direct and indirect costs, and how different interventions target them. Direct costs are addressed by subsidies like school uniforms, while indirect costs are tackled by conditional cash transfers (CCTs) or reducing distance to school. The lecture highlights the surprising magnitude of responses to small direct costs (e.g., $5 uniform), suggesting that simple cost-benefit models might be insufficient without accounting for severe credit constraints or social norms. Furthermore, the effectiveness of conditional versus unconditional cash transfers is debated, with some studies showing little difference, implying that the pure income effect or credit constraint alleviation might be more potent than the conditionality itself. Practical insights emerge from the case studies. The uniform paper in Kenya suggests that simply keeping girls in school (by removing the uniform barrier) is more effective at reducing HIV risk than complex educational programs like the ABCD strategy. This implies that basic access barriers can have profound, multi-faceted impacts. The Indonesian school construction program demonstrates how large-scale infrastructure investments, driven by political economy considerations (national unity, equitable resource distribution), can significantly increase educational attainment, particularly for younger cohorts. The speaker also offers advice on research: write papers you wish you had read, and don't be discouraged by perceived simplicity, as foundational work can be highly impactful. The lecture touches upon broader implications for development economics and policy. It underscores the challenges of identifying causal effects in education, particularly the "ability bias," and introduces instrumental variables and Difference-in-Difference (DiD) as key econometric tools. The political economy behind large-scale social programs, such as Suharto's use of oil money in Indonesia to foster national identity through education, provides crucial context for understanding policy design and implementation. The discussion on the limitations of simple economic models when faced with large behavioral responses to small costs suggests a need for integrating insights from behavioral economics into development policy. The evolution of econometric methods, from simple DiD to event study graphs, also highlights the continuous refinement of empirical research.

"the level of Education chosen by a family for their kids is going to be a function of the costs of Education the marginal cost for them of acquiring one more unit of human capital and of the returns to education both the immediate financial return in the form of wages as well as the human how is it it's going to acquire the human capital of the Next Generation"

"the impacts are really large really really large so you get for example in our uniform paper you get reduction in in Dropout of a few percentage points and the cost of a school uniform is is five dollars or so"

"it is better to just pay for the uniform and do nothing else that's worse that it is worse than trying to teach them the ABCD program while they're in school"

"people are very responsive to conditional cash transfer you do they are now been evaluating in in dozens of countries starting with the evaluation in Mexico which was one of the first big prominent City in development"

"if people are closer to school I probably are in more urban setting to what you really need is a change in the distance to school well where would that come from well if a country started building a lot of schools that would reduce the distance to school"

"I want to write the people that I wish to read and then you know that at least one person will like it and a pretty important person because that's yourself and you're going to spend a lot of time with that paper"

"the trick is going to both justify this assumption from a priori principle and kind of gather as much evidence as possible at the assumption is in fact real"

"his view for fixing that was to try and create a national identity which was uniform for the whole country and to do that it sort of invented a whole ideology called the pankasilla really promoted the Bahasa Indonesia language"

"in the places that that got one more school for each extra School per thousand kids the increase in education between The Young and the old court is 0.12 years faster"