Poster #54 - Can Health-Incentivized Children’s Savings Accounts Move Pediatric Outcomes? Evidence from Santa Cruz County

• In Event: Saturday Poster Luncheon

Abstract

A growing body of research links health and wealth, yet evidence for children’s savings programs built around this link remains limited. The one published randomized trial of a health-linked children’s savings account (CSA) – the Early Bird program in Texas – found no effect on milestone completion despite comparable per-milestone rewards (Olmstead et al. 2025). A related concern is that conditional incentive structures may disproportionately benefit families already equipped to navigate the health system, potentially reinforcing rather than reducing disparities.
This paper evaluates the health milestones offered by Semillitas, a CSA program in Santa Cruz County, California, that automatically enrolls every child born in the county and sets aside money for their future education. For Medi-Cal (California’s Medicaid program) enrolled families, Semillitas awards $25 deposits for completing CIS-10 (ten childhood immunizations by age two) and W15 (six wellness visits in the first fifteen months) if they give consent to share their health records. 
We conduct two analyses. First, we link administrative enrollment records with immunization and wellness visit data  from the Central California Alliance for Health (CCAH) and annual family surveys. We estimate linear probability models with robust standard errors, controlling for language, ethnicity, geography, birth cohort – and, in a survey-matched subsample, parental education and household income – to identify patterns in completion. Second, we compare Santa Cruz’s completion rates to those of four neighboring counties to assess whether the county with Semillitas outperforms demographically similar counties without it. A planned difference-in-differences analysis will exploit the June 2023 shift to hospital-based consent, which raised consent rates from roughly 11% to over 80%. Once post-consent cohorts age into CCAH data, updated data expected in May 2026 will enable causal estimation.
We find that 75% of eligible children complete W15 and 57% complete CIS-10. Completion gradients consistently favor less-resourced families. Residents of Watsonville – the rural, majority-Latino, lower-income portion of the county – complete W15 at 80% compared to 65% elsewhere, a 16 percentage point gap that is the strongest predictor in the regression (p < 0.001). Spanish-speaking households complete CIS-10 at 65% versus 48% for English-speaking households (p < 0.001). Families without a high school diploma complete at higher rates than college graduates, and middle-income households ($30,000-$75,000) complete CIS-10 at significantly higher rates than those above $75,000. These patterns contradict the concern that conditional CSA incentives necessarily favor the already-resourced. In the cross-county comparison, Santa Cruz’s W15 rate averages 80%, compared to 76% in Monterey and 58% in Merced; for CIS-10, Santa Cruz averages 49% versus 44% in Monterey and 19% in Merced – higher than every comparison county despite serving a predominantly low-income immigrant population.
These results provide the first evidence that a health-incentivized CSA can be associated with high pediatric milestone completion, and that the families completing at the highest rates are those the conditionality critique predicts would be left behind. The findings point to program design – specifically automatic enrollment, hospital-based consent, and trusted bilingual community outreach – as the factor distinguishing effective from ineffective health-linked savings programs.

Authors

• Eunjung Jee, Brandeis University
  Presenting Author

• Madeline Smith-Gibbs, Brandeis University
  Non-Presenting Co-Author