A recent evaluation in Mozambique shows that Early Childhood Development (ECD) interventions can be successfully implemented, rigorously evaluated, and have a direct impact on national policy dialogue, even in challenging environments with limited resources.
Starting in 2008, Save the Children implemented a community based preschool program in rural communities of the Gaza Province in south-eastern Mozambique. My co-authors Sophie Naudeau, Vitor Periera and I ran what is to our knowledge the first randomized controlled trial of preschools in Africa. The program provided seed funding for the construction, equipping and staffing of preschools in an area hit hard by the HIV-AIDS epidemic and where 1 in 10 children are orphaned. The preschool model is low cost. Each classroom costs $1000 dollars and communities donated labor and local materials. All in all, we estimate that the program costs less than $30 dollars a year per student.
For the evaluation we surveyed 2000 households with children ages 3 to 5 and around 1000 first grade students in 76 evaluation communities. Two years later, when comparing children in communities randomly selected for the program to children in communities not selected, we find that preschool participation increases the probability of enrolling in primary school, and children are better prepared for school based on measures of their cognitive and problem solving skills, socio-emotional development and fine motor skills. Furthermore, the program had positive externalities for older siblings of preschoolers, who were more likely to go to school, and parents and caregivers, who increased labor supply.
So what lessons can we draw from this experience for ECD in Latin America and the Caribbean? Mozambique is really far away from LAC! However I think there are important lessons from the way that this program was implemented and evaluated that are worth highlighting:
First, high quality programs can be implemented at low cost in challenging environments by leveraging local resources. In Mozambique, with no formally trained teachers available to staff the preschools, the program recruited and trained instructors from within communities and provided continuous monitoring and supervision of classroom activities to help ensure quality instruction. Most of the preschool’s instructional materials such as games and puzzles were manufactured by local artisans with locally available materials. And community ownership of the preschools helped ensure that parents were invested in contributing time to help in the upkeep of the facilities.
Second, the introduction of new and innovative ECD models should go hand in hand with the generation of rigorous evidence. Small pilot programs such as this one are particularly good candidates for experimental evaluations, since limited resources mean that the program cannot reach all the potential communities and children. In this particular case, funding was available to support only 30 of 167 eligible communities in the intervention areas. The critical step to generate a solid impact evaluation was to coordinate the selection of communities before the program rolled out, and agreeing that the selection procedure would be random. The small operational costs involved in setting up an experimental evaluation had a high pay off, namely knowing whether this innovative program actually works or not.
Third, in order for the research results to have an impact on policy, it is important to involve policy audiences in the program and evaluation design from the get go. In this particular case, although this was not a government program, we formed a steering committee with members of government ministries and other donors to help advise the evaluation starting at the baseline. During each visit to Mozambique we updated the committee on progress and shared early results. By the time impact results were completed, key policy makers were familiar with the project, understood the research, and were able to readily use this experience to inform public policy. Mozambique is now in the process of scaling up a much larger, government-led ECD initiative, and this topic is now firmly on the policy agenda.
I believe we can learn a lot from these results for the Latin America and the Caribbean context. It would be interesting to learn from your experiences and how do you think the success in Mozambique could be replicated in our region. Any ideas?
Sebastian Martinez is a Lead Economist in the Office of Strategic Planning and Development Effectiveness at the Inter-American Development Bank, where he supports the unit’s work on impact evaluation in the social sectors, including areas related to human development and infrastructure.