Four studies use new survey data from Wave 3 of the SEED for Oklahoma Kids experiment (SEED OK). In 2007, the experiment randomly selected a sample of 2,704 individuals from the population of newborns in Oklahoma (oversampling populations of color) and randomly assigned their parents to the treatment (n = 1,356) or control group (n = 1,348). The treatment included a CDA consisting of an automatically opened state-owned Oklahoma 529 College Savings Plan account, an initial deposit of $1000, and savings matches for deposits from lower income families. These features follow a policy concept in which CDAs are universal and automatic (including everyone), progressive (providing greater subsidies for vulnerable populations), and potentially lifelong (starting at birth).
The Wave 3 survey was conducted between January and August, 2020 when the children were about age 14. About 67% (n=1,799) of mothers completed the survey (Treatment=921 and Control=878). About 30% of families could not be reached due to their moving to other states, loss of contact, impacts of the COVID-19 pandemic, or the death of the child. The COVID pandemic greatly affected data collection (e.g., call center closed, field visits cancelled) and familiesââ¬â¢ daily living and activities. To better assess the CDA impacts, we conduct analyses on the pre-COVID sample (n= 707) unaffected by the pandemic and also the whole sample. Listwise deletion is applied to remove missing values.
Study #1 reports familiesââ¬â¢ historical harm experience (measured by stolen land, enslavement, housing restrictions, and family exclusion) by race/ethnicity, and provides a background to understand asset building and racial inequality.
Study #2 updates previous results on financial outcomes, and finds that the CDA experiment has a very large impact on overall OK 529 account holding and also the mean balance across all OK 529 accounts.
In Study #3, mothers in the treatment group have higher educational expectations and are more likely to prepare for childrenââ¬â¢s postsecondary education.
Study #4 suggests that, especially in the pre-COVID sample, treatment children are more likely to have better behavioral competencies. Analyses stratified by race/ethnicity suggest larger effects for minorities on some outcomes.
SEED OK has the most rigorous research design possible with random assignment in a full population sample; results can be generalized to the full population because of the research design. Consistent with previous results, new findings from Wave 3 show that asset building for all children (and its positive impacts) can be achieved through a universal CDA model built on the 529 plan. SEED OK evidence has already informed policies in seven U.S. states and in other countries. Future policy development is explored.