Book Description

This dissertation focuses on the evaluation of three distinct developing country policies, which, despite having been implemented in different parts of the world, are uniformly relevant to the field of development economics as well as effective policy design. Consequently, each essay contributes to the literature in its own way, either by evaluating a new and innovative intervention, by enhancing the theoretical understanding of policy-relevant interactions between public and private investment behavior, or by assessing the effectiveness of a broadly implemented, but yet unevaluated development policy. The results of the individual essays therefore convey one common conclusion; that a truly effective development policy must review carefully how its components interact with the behavioral responses of beneficiaries to identify the pathways through which impacts can be achieved.Chapter 1 examines whether computer-assisted instruction has a positive impact on the cognitive development as well as literacy and numeracy skills of early grade students. The analysis focuses on an educational intervention implemented in the rural region of Eastern Zambia that integrated technology into classroom activity in order to mitigate weaknesses in teaching skills and address specific unmet student needs. Using two control groups to compare the program's success to both standard government schools and lower-quality community schools, a difference-in-difference approach combined with inverse propensity score weighting is used to identify impact. While the program is unable to significantly advantage students in treated schools with respect to literacy, numeracy, and cognitive skills, estimates indicate that the program does succeed in leveling out initial differences, especially in comparison to government school students. This leveling by the program is accomplished at a third of the cost of government schools. An analysis of the heterogeneous impact further shows that effects are stronger for grade two students than for first graders. This is potentially because benefits take time to accumulate or because computer-assisted instruction becomes more important in supporting teachers as teaching becomes more complex and requires more materials. These results drive home the importance of integration of technology into curriculum and teaching methodology and how this can be a cost effective means to improve student learning.Chapter 2 examines how public education expenditures affect household spending on schooling and provides new theoretical underpinnings that highlight the importance of incorporating models of household decision-making processes into policy design. The study takes advantage of two country cases, Indonesia and Peru, which offer sufficient variation in public expenditures at the local level, and therefore lend themselves to assessing the important effect of government education expenditures at the district level on household spending on schooling. Employing a fixed-effects regression and an instrumental variable approach in Indonesia and a pseudo-panel approach in Peru, results indicate that a 1% increase in public-level education expenditures per school is estimated to decrease household-level spending on schooling per school-aged child by approximately 0.5% in Indonesia and 0.04% in Peru. This suggests that government spending may crowd out private investment in schooling, which represents an important indirect effect of any educational policy and can potentially diminish policy effectiveness. A closer look at household expenditures revealed that the specific (country) context will determine how parents reallocate their resources in response to changes in public spending levels.Lastly, Chapter 3 presents the first rigorous impact evaluation of a shoreline stabilization program in Barbados and attempts to assess whether shoreline stabilization investments indeed have beneficial effects on medium-term economic growth in Small Island Developing States through stimulating tourism demand and real estate development. The analysis relies on a carefully designed geographic information systems (GIS) dataset, which comprises extensive panel data from Barbados' touristic West and South Coasts on key infrastructure, beach characteristics, and real estate activity, as well as remotely-sensed luminosity data as a proxy of economic growth. The synthetic control method is employed to construct a counterfactual from a combination of all control beach sites and subsequently estimate program impact on per capita luminosity as a proxy for GDP per capita. Results indicate that even in the first three years after treatment, economic effects are positive and indicate a strong positive trend. This suggests that shoreline stabilization works may not only help preserve fragile ecological conditions, but further lead to sustainable growth in the local economy.