International Topics

• Agriculture
• Democracy and Governance
• Education
• Health
• Infrastructure
• Social Protection
• Capabilities
• Clients

Burkina Faso: Improving Education for Girls

The Burkinabé Response to Improve Girl’s Chances to Succeed (BRIGHT) program was designed to improve the educational outcomes of children in Burkina Faso. It focused on girls in particular and was implemented in 132 rural villages throughout the 10 provinces in which girls’ enrollment rates were lowest. It consisted of constructing primary schools with three classrooms and implementing a set of complementary interventions. These included inputs such as separate latrines for boys and girls; canteens; take-home rations and textbooks; and “soft” components, such as a mobilization campaign, literacy training, and capacity building among local partners. The program was implemented between 2005 and 2008.

Mathematica conducted an impact evaluation of BRIGHT, which answered three key questions: (1) What was the impact of the program on school enrollment? (2) What was the impact of the program on test scores? (3) Were the impacts different for girls than for boys?

The program has been extended and BRIGHT II is constructing three additional classrooms for grades four through six and providing complimentary activities in the same villages. We are designing and implementing a rigorous impact evaluation of BRIGHT I and II effects on girls’ school enrollment, attendance, and learning, using a regression discontinuity design.

Evaluation Design and Data Collection

The evaluation design involved comparing children in the 132 BRIGHT villages (participant group) with children in 161 similar villages that had applied to participate in BRIGHT but were not chosen (comparison group). The statistical technique used to estimate program impacts is called regression discontinuity, which takes advantage of the fact that all 293 villages that applied to the program were given an eligibility score by the Burkina Faso Ministry of Education based on their potential to improve girls’ educational outcomes.
Evaluation data were collected from January through April 2008 by a team from the University of Ouagadougou with oversight from Mathematica. The data collection included two surveys:

• A face-to-face household survey, which included questions on household demographic information, children’s educational outcomes (such as enrollment and attendance), and parents’ perceptions of education. The target sample for the survey was a random sample of 30 households with school-aged girls in each of the 293 villages that applied to the BRIGHT program—8,790 households in total. That survey included tests in math and French administered to all children ages 5 to 12 who lived in households interviewed, regardless of whether they were enrolled in school. A total of 21,730 children took the tests. The response rate was 97 percent.
  
• A school survey administered in two waves. In the first wave, information on the schools’ characteristics was collected from school officials. In the second wave, attendance and enrollment data were collected for children who were enrolled in school, based on parents’ reports from the household survey. The target sample for the survey was the three primary schools closest to the villages that applied to the BRIGHT program (within 10 kilometers) and regularly attended by children from those villages. This yielded 360 schools. The response rate for the school survey was about 99 percent.

Findings

By and large, we found that the BRIGHT program was implemented as intended. The schools were constructed and the set of complementary interventions were implemented for the most part according to the original plans. We found that the program had the following impacts:

BRIGHT had a positive impact on school enrollment. The impact of BRIGHT on enrollment was an improvement of about 20 percentage points, based on household survey data. The impact on whether a child was present on the day we visited the school was about 16 percentage points. The effects imply that BRIGHT was responsible for increasing enrollment rates from about 35 percent to 55 percent (household-reported outcome) or from about 31 percent to 47 percent (school-based outcome).

BRIGHT had positive impacts on math and French test scores. The impacts on both outcomes were approximately 0.4 standard deviations. This increase in test scores is larger than many other successful education interventions in the developing world, which have effect sizes typically between 0.1 and 0.3 standard deviations (although BRIGHT is a more costly intervention, since it involves building schools). In this context, an impact of this size implies that, for a student who started at the 50th percentile of our sample, attending a BRIGHT school is predicted to increase his or her test score to approximately the 80th percentile.

The impacts of BRIGHT were positive for both boys and girls. In terms of enrollment, the impacts for girls were about 5 percentage points higher than the impacts for boys. In terms of test scores, the impacts for girls and boys were statistically indistinguishable.

Publications

"Impact Evaluation of Burkina Faso's BRIGHT Program" (June 2009)