The Bolivian Social Investment Fund: An Analysis of Baseline Data for Impact Evaluation

The Bolivian Social Investment Fund (SIF) is a financial institutionthat promotes sustainable investments in the social sectors,principally in the areas of health, education, and sanitation.This article shows how to use preintervention data collectedfor evaluating the SIF to improve the targeting of a program,to test the quality of the evaluation design, and to definecorrective measures if necessary. It finds that among SIF interventionsthe benefits in education are distributed relatively equallyover the population, while the investments in health and sanitationfavor better-off communities. The article contributes to the methods used to evaluate socialinvestment funds and similar programs. It compares two typesof evaluation designs to assess social investment fund interventionsin the education sector. The authors demonstrate that a simplematched-comparison design introduces a bias in the estimateof the program effect, whereas an experimental design basedon random assignment does not. With preintervention data, theanalyst can select a quasi or indirect experiment, where thechoice of the indirect experiment coincides with the selectionof valid instrumental variables. The availability of preinterventiondata makes it possible to compare the two types of evaluationdesigns as well as to test the validity of the instruments andto determine the loss of efficiency due to the use of quasi-experimentaltechniques instead of random treatment assignment.