Association for Public Policy Analysis & Management

Background: This study presents findings about the implementation and impact of Child Development Accounts (CDAs) in the SEED for Oklahoma Kids (SEED OK) experiment. SEED OK, which began in 2007, is a large-scale policy test of universal, automatic and progressive CDAs. The CDA in SEED OK consists of an Oklahoma 529 College Savings Plan (OK 529) account automatically opened for all newborns in the treatment group with an initial deposit of $1,000. It also includes an optional individual OK 529 account, a $100 account-opening incentive, a savings match, and educational materials.

Methods: The sampling frame for SEED OK consisted of birth records for all children born in Oklahoma during certain periods in 2007. Study participants are the primary caregivers—mostly mothers—of the infants identified in the sampling frame. Mothers and their infants were assigned randomly to the treatment group or control group after mothers completed a baseline survey. Children in the treatment group received the CDA (n = 1,358), and children in the control group did not (n = 1,346). The study collected the follow-up survey data in 2011 and received quarterly account data from the OK 529 plan. Random assignment and probability sampling from a full state population make SEED OK an ideal test for a universal policy.

Results: The CDA in SEED OK has very large impacts on financial outcomes 7 years after the intervention began. Treatment children are 30 times more likely than control children to have college savings (99.9% vs. 3.3%), and the average total amount of college savings is about six times larger ($1,851 vs. $323). Investment earnings also contribute to the treatment-control difference in asset amount. The $1,000 initial CDA deposit increased by more than 40% over 7 years. The CDA’s financial impacts are especially strong for disadvantaged children. For example, in many disadvantaged groups, the CDA increases asset holding rates from 0% to 100%, and even treatment children who received no family deposits earned $426 through market growth. The CDA also increases the likelihood that parents save for their children’s future college expenses, and this is true for disadvantaged as well as advantaged families. The CDA has important nonfinancial outcomes: it helps mothers maintain or increase their expectations for their children’s postsecondary education, boosts mothers’ mental health, and improves disadvantaged children’s early social-emotional development. In most cases, these impacts occur regardless of parental saving behavior.

Conclusion: Because account opening and initial deposits were automatic for all newborns, the CDA in SEED OK is the first CDA in the US to achieve full inclusion. Demonstrating full inclusion paves the way for widespread participation in asset building and more equitable distribution of public resources. The automatic features of the CDA are also largely responsible for the CDA’s financial and nonfinancial impacts over time, impacts which are often greater for disadvantaged families. Future research will examine whether the CDA in SEED OK increases college graduation rates, but early impacts on college savings, educational expectations, and child development suggest that this long-term impact is possible.