What are the long-term impacts of early-childhood education?
We study the long-term effects of Project STAR Experiment, which randomly assigned children
into different kindergarten classrooms.
Even at an early age, test scores have important predictive power for eventual outcomes,
like earnings at ages 25-27, college attendance, neighborhood quality, and even chances of getting married.
Kids who were randomized into a higher-quality classroom, as measured by the average test scores of their peers, earned more as adults.
Higher quality kindergarten classrooms raised test scores initially, but the gains evaporated after a few years. The effects of higher quality education then re-emerged in adulthood. The data suggest that this pattern may be driven by the effect of education on non-cognitive skills, such as the ability to cooperate well in groups.
These results suggest that government policies focused on increasing the quality of early-childhood education
may have large returns over the long run for student outcomes.
To learn more, please read our paper, our non-technical summary, or our slides (PDF, PPT).