What are the best ways to deploy finite resources for the betterment of young children? What inputs provide the most beneficial outcomes later in life? These are big, important questions whose answers matter to individuals, families, and society. A new report from Urban Institute’s Income and Benefits Policy Center aims to provide the answers.
Authors Kevin Werner, Gregory Acs, and Kristin Blagg build on a large body of previous research showing the well-established link between childhood and adult outcomes. They add to the literature by constructing a model that simulates changes in several variables at different points in children’s lives and projecting their outcomes as adults. Data come from the Social Genome Model (SGM), a microsimulation model of the life cycle that tracks the academic, social, and economic experiences of individuals from birth through middle age in order to identify the most important paths to upward mobility. They translate SGM data to match real world data from the Early Childhood Longitudinal Survey-Kindergarten Cohort (ECLS-K) for childhood outcomes at ages five, eight, and eleven, and the National Longitudinal Survey of Youth 1997 Cohort (NLSY97) for adult outcome data.
Their model focuses on a single adult outcome: earnings at age thirty. While this is mainly because age thirty is currently the extent of the SGM data, they also consider earnings a useful proxy measure for other aspects of adult life, “given the strong relationship between higher income and other indicators of well-being, such as physical and mental health.” Additionally, research has shown that earnings at age thirty are predictive of lifetime earnings.
They look at sixteen different factors across four different stages of childhood: birth, preschool, early elementary school, and middle childhood. The model contains variables from five different domains for each life stage, comprising cognitive and academic development, emotional and psychological development, physical health and safety, mental health, and social behaviors. The changes their model introduce to childhood circumstances are all improvements, and all important. Examples include increasing birth weight to safer levels, boosting math and reading test scores by 0.5 standard deviations, and improving health indices by 0.5 standard deviations. Changes are introduced at various points in childhood to determine their adulthood impact.
Findings indicate that improving children’s math scores outshines all other factors and increases adult earnings the most. Improving preschool (age five) math scores by 0.5 of a standard deviation yields an impressive 2.5 percent increase in earnings at age thirty; doing so in middle childhood (age eleven) raises earnings by 3.5 percent, which corresponds to about $1,200 per year (in 2018 dollars) in additional earnings for the average adult. Roughly similar impacts are seen in children of all races and ethnicities across life stages, with Hispanic children consistently seeing the largest gains. Girls tend to see a higher earnings boost than boys. By comparison, improvements in childhood health indices result in small but steady earnings increases of 0.6 to 0.7 percent in adulthood, with larger gains for Black and Hispanic children as compared to their White peers. Improvements in reading test scores (again, 0.5 of a standard deviation) give a nearly 1 percent bump to adult earnings when they happen at age five, but decline by half when they occur at age eleven.
The authors do not discuss potential mechanisms at work, given the exploratory nature of their analysis, and the findings as discussed are simplified for the same reason. “Increasing math achievement by 0.5 standard deviations” is far more easily done on a computer than accomplished in practice, but real-world research has shown that there are many ways both inside and outside the classroom to help students improve their math performance. Thus two important points emerge from this exercise. First is that policymakers can and should direct resources toward interventions and supports that are proven to provide the largest boost for the families they serve. This analysis suggests that math success in childhood is critical to adult—and lifetime—success. Second, we have multiple ways to accomplish in the real world what computer scientists can do with the click of a mouse. Urgent dedication from all stakeholders to the goal of improving math scores for students is required, with the broadest support and resources possible, as an investment in children’s (and families’ and society’s) futures.
SOURCE: Kevin Werner, Gregory Acs, and Kristin Blagg, “Comparing the Long-Term Impacts of Different Child Well-Being Improvements,” Urban Institute (March 2024).
