| As developing country education systems shift their focus from enrollment towards learning outcomes, developing fair and accurate measures of school performance is increasingly critical. This study uses publicly available test score data from Tanzania to construct school value-added scores, providing an alternative to the current measure of school performance, average test scores. Average test scores tend to be biased in favor of schools in relatively well-off areas and biased against schools serving less well-off student populations. However, performance rankings by value-added scores rather than average test scores are shown to create substantially different outcomes for many schools. This study uses value-added scores, along with a number of school-level controls, to compare the value provided by private schools versus government schools in Tanzania. Multivariate regression results suggest that private schools tend to produce value-added about one standard deviation better than similar government schools, but this difference is smaller than when schools are measured by average test score rather than value added. Data and results from this study are being used in current discussions with Tanzanian education policymakers to demonstrate the importance and feasibility of using value-added scores as one measure of school performance. |
