There is a new Best Evidence in Brief with amongst onthers this study, written up by Sooyeon Byun:

In 2003 and 2004, a team of researchers implemented a year-long social and emotional learning (SEL) intervention with 192 children within 22 classrooms within 12 Head Start programs. This intervention, called Head Start REDI, was an integration of PATHS, which is a well-known SEL intervention, and a daily interactive reading program using books aligned with PATHS’ social and emotional themes. Teachers in the intervention group received a four-day training and weekly mentoring. An additional 164 children within 22 classrooms within 13 Head Start programs served as a control group.

Karen Bierman and her team followed these children after 8 to 10 years to estimate the long-term effects of the intervention. Among the original 356 children, 281 children (81%) were reassessed in this study. Children who were 4 years old at the time of intervention were in grades 7 and 9 when they were assessed again. The researchers found that children who participated in Head Start REDI displayed significantly lower levels of conduct problems (ES = -0.20) and emotional symptoms (ES= -0.25) than the control group. However, the two groups did not demonstrate significant differences in hyperactivity/inattention or peer problems.

It is important to note that the findings need to be interpreted with caution because of the substantial loss of students over time. Only 78% of children in the intervention group (144 children), and 88% of children in the control group (145 children) were retained, showing considerable differences in the attrition rates between the two groups. However, the intervention and control groups did not have significant differences in the outcome measures or demographic characteristics when they were first assessed at age 4, and the children who dropped out and who remained in the study showed no meaningful differences, except for race/ethnicity. Altogether, with reservations, this study provides evidence that a high-quality and high-dosage SEL intervention in preschool years can potentially have long-lasting effects through children’s early adolescence.

Daniel Willingham shared another study by Sara Amadon and colleagues that looked at a different aspect of early childhood education:

Early childhood education contributes to improved school readiness but impacts on high school remain unclear. This study estimates the effects of Tulsa, Oklahoma’s universal pre-K and Head Start programs through the junior year of high school (in 2018/2019; N = 2902; M_age = 16.52, SD = .39; 48% female; 28% white, 34% Black, 27% Hispanic, 8% Native American). Propensity score weighted regressions suggest students who attended pre-K, but not Head Start, missed less school, were less likely to fail courses and be retained in grade, were more likely to take an Advanced Placement/International Baccalaureate course, but did not have higher test scores or grades. Subgroup analyses by race/ethnicity demonstrated some differences in the pattern of associations favoring students of color.

While these results seem to be a bit disappointing, but IMHO Dan is correct in his reply:

No effect of Head Start was observed. But why would you expect such long term effects? Would you expect good nutrition at age 5 to lead to good health at age 18? You need quality throughout.

It’s the difference between the Heckman Curve (invest as young as possible) and the Heckman Equation:

• + INVEST: Invest in educational and developmental resources for disadvantaged families to provide equal access to successful early human development.
• + DEVELOP: Nurture early development of cognitive and social skills in children from birth to age five.
• + SUSTAIN: Sustain early development with effective education through to adulthood.
• = GAIN: Gain a more capable, productive and valuable workforce that pays dividends to America for generations to come.