- What are the socioeconomic consequences of large educational expansions?
- Which policies can help improving education quality?
- What are the effects of providing different types of resources to schools?
Contribution to International Research
While the literatures on the determinants of education quality and on returns to education ar enormous, most of the evidence comes from high income countries. However, this work is unlikely to be of much guidance when considering the effect of education policies in middle and low income settings. In particular, little isknown on determinants of education quality in Southern Africa, and very little high quality evidence exists on returns to education in some regions, such as the Middle EAast and North Africa (MENA). Evaluating the returns to education or the effect of different policies on education quality is challenging in such settings, given the general lack of data availability. The challenge is even bigger when attempting to obtain causal effects of such policies, as there are likely to be confounding factors that obscure the most direct and straightforward analyses. This project seeks to address these issues by processing, merging and homogeneizing a variety of country-specific data sets and by applying quasi-experimental approaches to estimate the effect of education policies on schooling and socioeconomic outcomes.
Research Design and Methods
The project uses quasi-experimental approaches such as regression discontinuity designs and instrumental variables to answer specific evaluation questions. For instance, we evaluate the effect of the large educational expansions in the MENA on labour market outcomes and political attitutdes exploiting several specific education policy changes in Tunisia that affected some cohorts but not others. This allows us to address hypotheses regarding the role of education for grievances and political attitudes in the runup to the Arab Uprisings. Another example of analysis in this project concerns the effect of providing financial resources to schools in South Africa. In South Africa, these resources are allocated on the basis of school poverty score quintiles. This peculiar allocation rule can then be used to estimate the effect of these resources quasi-experimentally using a regression discontinuity design.
Preliminary results point at relatively limited labor market effects of the large education expansions in Tunisia, at least for men. While obtaining more education is associated with better labor market prospects, such as a higher likelihood of becoming a manager, we show that this is driven mostly by the public sector. It appears that the private sectors values relatively little the skills obtained in formal education in Tunisia, while the public sector values mainly credentials. Preliminary evidence from the effect of financial resources on education outcomes in South Africa also points at limited effect of these resources. We find that such resources appear not to affect results from standardized test taken at the end of secondary school. If at all, these resources appear to increase throughput during the last years fo secondary school. We conjecture this could be because funding is provided per pupil, so that higher resources give incentives to keep relatively weak students.