New Research: Did Colombia’s Nationwide Preschool Expansion Benefit Students?

Which strategies work to improve the quality of public preschools? In new research published in the Journal of Political Economy, Yale Economist Orazio Attanasio and coauthors partnered with the Colombian government to evaluate a nationwide expansion of its preschool program. In collaboration with a private foundation, they also evaluated the impact of slightly modifying the government intervention. Prior to this research, few studies have addressed which aspects of preschool programs are most important for child development or whether specific improvements to existing programs are effective.

Their paper directly addresses key education policy issues: Does providing schools with additional resources necessarily improve education quality? Do teacher professional development programs work? They find, strikingly, that a costly national government program that provided resources to hire teaching assistants (TAs) had no impact on child development. However, also including—at little extra cost—a professional development training program for existing preschool teachers resulted in significant positive overall impacts on children’s cognitive development. They show that, even within the same institutional setting, different approaches to improving the quality of early-years education can have very different effects on child development.

The article and interview with Professor Attanasio below dive into the research process, the key results, and the implications for local and international education policy.

• The TA program alone had no positive impacts on child development, despite high compliance and the fact that it represented a large increase in government investment in preschools.
• The authors found that teachers responded to the program by reducing their overall involvement in classroom activities, delegating much to (often untrained) TAs, thus inhibiting child development.
• However, moderate extra training of the existing teachers did have significant positive impacts on child development.
• This training program prompted teachers to delegate only specific activities to TAs, and to focus more on improving teaching quality and spending more time on learning activities.
• Overall, additional school resources can be effective when accompanied by guidance on how to utilize them. Without guidance, such provision might generate unintended consequences.

Early childhood education (ECE) plays a critical role in child development: evaluations of well-designed ECE programs have often shown substantial and long-lasting positive effects on children. Yet, there is limited evidence on the effect of these programs in lower- and middle-income countries, where services are of widely variable quality, with many children receiving poor-quality center-based care. As global momentum around investing in early education builds, and preschool enrollment rates continue to rise, many questions remain around how governments should allocate resources to ECE, and how they can improve overall education quality.

In a new study recently published in Journal of Political Economy—“Preschool Quality and Child Development”—researchers evaluated two strategies to improve the quality of public preschools in Colombia. The first, designed by the government and planned to be rolled out nationwide, provided extra funding earmarked mostly for hiring teaching assistants (TAs). The second complemented the first by additionally providing low-cost professional development training for existing teachers. The authors used a randomized evaluation to estimate the impacts of these two interventions in improving children’s cognitive development, especially for more disadvantaged children. Throughout the study, they collected rich measures of child development, the classroom environment, and teaching practices.

The study took place in the eight largest cities in Colombia: Bogotá, Cali, Medellín, Barranquilla, Bello, Palmira, Itagüí, and Soledad, where the national government implemented large scale ECE development projects. This preschool program is one the oldest public center-based childcare providers in Colombia, and the centers have enrolled an average of 125,000 children per year over the past decade. The study population included partially subsidized government preschools for children between the ages of 18 months and 5 years from low–socioeconomic status families. To evaluate the government’s expansion of the program, the authors worked with the government to embed a randomized control trial (RCT) within the program rollout and study the outcomes. This design enabled them to evaluate rigorously the impact of the Colombian government’s approach to quality improvement as it was, in practice, implemented nationwide.

To analyze the effects of the programs on child development, the authors first gathered baseline data using eight different assessments given to the preschoolers. After the study period, they compared these data with a new set of child development tests. The authors also collected detailed measures of classroom activities in order to assess whether and how the interventions changed the routines and quality of instruction among the teachers and TAs, and measured the quality of teaching activities through direct observation of the teachers.

The authors found that the first intervention had no effect on child development. Surveys revealed that teachers relied on TAs to substitute their work, which led teachers to reduce their classroom time, including care and learning activities. However, the additional training improved teaching quality and efficient use of TAs in the classroom. This led to improved cognitive development, especially for more disadvantaged children. The authors noted that the additional professional development gave teachers the skills needed to delegate tasks to TAs appropriately, and the result was greater involvement in learning activities and improved quality of teaching.

These findings suggest that provision of additional human resources can trigger changes in teachers’ time use that may counteract the positive impact of the new resources. In contexts where teachers are poorly trained, additional school resources should be accompanied by guidance on how to utilize them. Without guidance, such provision might generate unintended consequences, like the reduction in effort that was seen among teachers in the first program. On the other hand, training can lead to improvements in the efficiency of how teachers utilize their and the TAs’ time, delivering improvements in child development.

These results offer encouraging evidence on the potential of teacher training programs to change teaching practices in ways that translate into improvements in children’s outcomes. Overall, the study provides evidence on a concrete, scalable way in which the government could improve education programs to deliver significantly better outcomes for children at little extra cost. This study has significant relevance beyond Colombia, as governments in developing countries face the challenge of how to reimagine existing ECE services with the goal of improving child development.

Collaborating and working with government may be difficult, in particular when performing rigorous evaluations. The overall agenda and policy timelines may be difficult. In this case, we built on many years of collaboration with ICBF, the government agency that runs most early years programs in Colombia and that is very well organized. The collaboration and intermediation of the Fundacion Exito was also very important—in addition to the fact that they funded a big part of the evaluation.

At a high level the main message is that quality in early years interventions matter. And quality is not always easy to achieve. Our work showed explicitly the changes induced by different interventions on teachers’ behavior that led to the impacts we saw. On the positive side, we showed that a small additional cost can trigger important impacts on the basis of the substantive resources invested. The big challenge is to do this at scale. It is possible, but it is not easy.

We have shared our results with policymakers, and while the changes are not immediate, we think that our results are slowly changing the way things are done in this context.

Many of the interventions for the early years seem to be successful when changing the behavior of key agents, such as parents and teachers. An important and I think exciting area of research is to understand what are the drivers of these agents’ behavior. Is it misperceptions and beliefs? Or social norms? Or simply resources. Studying how subjective beliefs change is a very exciting area of research.