Early childhood is a critical period for long-term cognitive and psychosocial development. Developing in a responsive, stimulating, and caring environment is vital for children’s positive development and growth. Conversely, inadequate stimulation, lower opportunities for learning, stunting, anemia, exposure to violence, maternal depression, among others, can stifle children’s development. In Latin America alone, 11 million children under 5 are at risk of poor development. Studies show that differences in socio-economic status result in gaps in cognitive skills during early childhood in several Latin American countries.
Today, COVID-19 has exacerbated threats to young children’s development under lockdowns, economic hardship, caregiver stress, and disruptions to ECD programs and health care and social services. COVID-19 could also, however, provide an opportunity to rethink ECD programs and policies, experiment with new delivery platforms, and make new models more effective. Collaborations between researchers, governments, NGOs, international organizations, and Evidence to Policy Organizations that focus on evidence-generation and innovation to design scalable Early Childhood Development (ECD) programs can hold the key to tackle these challenges.
We suggest that randomized evaluations can – and should – be used to design innovative and scalable early childhood development programs.
– Why are randomized evaluations effective?
2019 Nobel Prize in Economics winners, Esther Duflo, Abhijit Banerjee, and Michael Kremer highlighted the promise of the experimental approach to solving daunting policy challenges. Randomized evaluations are a powerful tool to understand which interventions work and to test if programs are as effective at scale, ensuring that they deliver the promised benefits. Not only can they provide answers about whether a specific program is effective in a particular context, but they can also shed light on why the intervention works, thus deepening our knowledge about the causal mechanisms at work. Understanding these mechanisms allows us to draw lessons about human behavior and assess if these interventions could be replicated in new contexts. Following an appropriate framework of analysis, evaluations of interventions around the globe can inform the design of a program in a specific country.
– How can randomized evaluations be applied to ECD?
In a complex field such as ECD, we can see how randomized evaluations are helping us to better understand whether the interventions designed and implemented around the world are effective and why. For instance, several randomized evaluations in Latin America, Africa, and Asia have consistently shown that programs that motivate parents to engage in early childhood stimulation activities succeeded in improving parents’ interactions with children, and in turn, enhanced children’s developmental outcomes. Although these interventions are very promising, implementation at scale is still a significant challenge. Taken together, these evaluations shed light on how to move forward: building on previous efforts and innovations, new research should explore variations in the delivery model focusing on improving the quality of implementation at scale and sustain its effects.
– How is randomized evaluation being used in preschool? A few concrete examples
When it comes to early learning opportunities in preschool, though there is a limited number of studies from developing countries, an example from Latin America illustrates how randomized evaluations can generate valuable lessons on how to improve pedagogical practices in the classroom. In Peru, a team of researchers, including J-PAL affiliate Francisco Gallego, tested a cost-effective intervention that relies on a scaffolding approach that tailors activities to children´s ability. Researchers evaluated the impact of an innovative program that changed the pedagogical approach used to teach the existing national mathematics curriculum on preschoolers’ numeracy and ability to recognize shapes. The program used an inquiry- and problem-based learning approach to tailor instruction to preschoolers in Peru. The intervention was found to improve mathematical results of preschoolers. Given that the intervention was implemented during regular school hours, using existing curricula, and relying on current teachers, the results of this evaluation highlight the potential for scalability and a promising path of interventions that aim to change pedagogical practices without changing existing inputs.
The examples above show us how randomized evaluations are helping to answer specific policy-relevant questions about the interventions evaluated while opening a path to keep innovating to achieve impact at scale.
– The way forward
COVID-19 has disrupted the status quo and demanded innovation and resilience to protect children during this critical period of development. It is inspiring to see how actors in the field of ECD have responded to the crisis and continue to work towards achieving impact at scale. This kind of learning process requires long-term collaborations among governments, researchers, NGOs, international organizations, and Evidence to Policy organizations and thus entails significant coordination challenges. However, the continuous application of lessons learned drawn from this approach can build effective evidence-based programs that have the potential to substantially improve the lives of millions of children in Latin America and the world.