On a typical pre-K morning, a teacher kneels beside a three-year-old at a low wooden table, guides her hand across a page of letters, and asks what sound the first one makes. It is easy to assume pre-K’s value lives in moments like this: patient attention, repetition, a child learning to hold a pencil. What’s harder to see is which parts of that morning actually build a child’s mind. Between 2013 and 2015, researchers ran a randomized trial across 120 public preschools in Colombia to test part of that question[Povertyactionlab]. The answer cuts against the idea that more funding, on its own, makes a better preschool.
The Assumption Most People Start With
A common way to judge a pre-K program is by what you can see: bright rooms, new toys, lower child-to-teacher ratios, extra staff.
Photo by It’s a reasonable guess. More adults in a room should mean more attention per child, and more attention should mean more learning. The goal is the right one. What the assumption skips is how that attention gets used once it is in the room.
Colombia tested this directly through its national preschool-improvement program, which added staff and materials to existing centers and raised spending per child by about 30 percent[Povertyactionlab]. Researchers wanted to know whether simply strengthening the basic program would move the needle on child development. The honest answer was no. The base preschool-improvement program had no positive impact on child development, and teachers actually reduced their own involvement in classroom activities once extra staff were added[Povertyactionlab]. Adding people to a room does not automatically add learning to a room. Teachers in the resourced centers ran 0.34 standard deviations fewer learning and development activities than teachers in the comparison schools[Povertyactionlab]. The researchers tie the drop to the hiring of teaching assistants, which led teachers to spend less time in the classroom, including on learning activities[Journal of Political Economy].
What Actually Moved the Numbers
The second arm of the trial is where the results changed.
Photo by Researchers kept the extra resources and added structured training for the teachers already in the room. The training focused on improving day-to-day interactions with children and covered topics including early literacy and early math, delivered through 16 monthly videoconference sessions, weekly online tutoring, and one observed lesson with feedback[Povertyactionlab]. Children in these classrooms gained 0.15 standard deviations in cognition, language, and school readiness compared to peers in standard programs[Povertyactionlab]. Researchers measured that gain four months after the training program was put in place. The children in the study averaged 29 to 30 months old, so it is a reading on early skill, not on later school results[Povertyactionlab]. The researchers suggest the teacher training drove most of the gain, but the trial never tested training without the extra resources, so it cannot show how training would do on its own[Povertyactionlab].
The gains were not evenly spread. For children from below-median-wealth families, the combined program produced roughly double the effect, around 0.30 standard deviations[Povertyactionlab]. These preschools already served low-income families, so the largest gains went to the poorer half of an already disadvantaged group[Povertyactionlab].
Reconciling the Two Pictures
These two findings don’t contradict each other.
Photo by The base program (extra staff, more materials) failed to help on its own. Adding the teacher-training layer, at an estimated ongoing cost of about US$13 per child per year, turned it into a program with real gains that favored lower-income children[Povertyactionlab]. J-PAL’s summary concludes that extra resources alone may not be enough to improve children’s outcomes, and that adding teacher training can offset the program’s negative effects and improve preschool quality[Povertyactionlab].
Why This Matters Beyond One Country
It would be easy to file this away as a Colombia-specific result.
Photo by The trial ran only in Colombia and does not show that the same gains would appear elsewhere. Its broader value is the policy question it tested. The Colombian findings separate two things that often get bundled together in policy debates: spending more money, and spending it well. The version that worked added about one percent to spending per child, on top of a roughly 30 percent increase that had not improved children’s development by itself[Povertyactionlab].
That distinction matters for anyone designing or funding early childhood programs.
What the Trial Shows
None of this means facilities are irrelevant. The extra staff and materials were part of the program that worked, and the trial did not test training without them. On their own, those resources left children’s development unchanged and reduced the time teachers spent on learning activities. Combined with teacher training, the same resources raised cognition, language, and school readiness by 0.15 standard deviations, and by about 0.30 for children from the poorer half of families[Povertyactionlab].
The training also changed what happened in the classroom. Observers rated teaching practices and classroom processes about 0.4 standard deviations higher in preschools that received both the resources and the training than in the comparison preschools or in those that received resources alone[Povertyactionlab].
