Does CSA Participation Influence Savings Behavior in the Long-Term? Views from Families

Background: Child savings account (CSA) intervention programs have been implemented in a growing number of locations with the aim of improving the long-term well-being of families by promoting saving and financial stability. Much of the US-specific research on CSAs comes from the Saving for Education, Entrepreneurship, and Downpayment (SEED) initiative that implemented CSAs in 12 sites across the country beginning in 2003. We followed up with participants from a CSA quasi-experiment that was conducted in Michigan (MI-SEED) ten years after the program had ended to ask: Were families with accounts able to save and improve their financial stability? How do families perceive long-term influences on their savings behavior?

Methods: The MI-SEED program was designed as a quasi-experimental comparison of seven “treatment” and seven “control” Head Start centers. Caregivers of 790 children across all 14 centers were invited to complete a baseline interview in 2004 and a Wave 2 survey in 2008. Caregivers with children enrolled in treatment Head Start centers were offered a CSA with an initial $800 deposit, matched by a $200 state deposit, and were also encouraged to take advantage of a 1-to-1 match. Our 2014/2015 mixed method study followed up with several treatment and control families who took part in the original study. We conducted semi-structured interviews with and collected survey data from 50 households. We also obtained updated aggregate data on all SEED 529 accounts from TIAA (n = 496). Interviews were transcribed verbatim and coded thematically by three independent coders.

Findings: First, we found few families had made withdrawals from their SEED accounts and many had grown their account balances. The accounts on average had a total balance of $2,017 and a median value of $1,337. Second, participants identified that their savings behaviors were influenced by a myriad of factors other than participation in an incentivized savings program. To tease apart these factors, we constructed a typology of savers from interview data: always savers (n = 8), sometimes savers (n = 35), and never savers (n = 7). Institutional constructs that supported savings behavior, such as savings incentives built into the SEED CSA intervention, were only identified as consistent facilitators for savings by our most advantaged group, the always savers. Additionally, we found that different psychosocial and structural obstacles or supports were associated with each typology group. For example, the always saver group was the only group to have benefited from intergenerational transfers of wealth or savings knowledge.

Conclusions/Implications: Lending nuance to the idea that incentives and institutional access necessarily improve savings outcomes, these findings correspond with prior work identifying the determinants of asset building (Beverly et al., 2008). For families that did not experience transfers of intergenerational wealth or savings knowledge, accumulating savings was elusive. The study’s findings showed that strengthening institutional determinants of asset building through program features or directly intervening to influence intergenerational transfers of wealth and information may be keys to improving the long-term savings outcomes for CSA participants.