Fortunately, there are two approaches that can assist parents in preparing financially for their children's college education. The first are government college savings plans that have tax advantages. The second are child development accounts (CDAs), which are an asset-building instrument from which the proceeds can be used to finance postsecondary education or other endeavors. Previous research has shown that financial capability, which is a combination of financial knowledge and financial access, has a strong influence on parents' decision to use the government-sponsored plans and CDAs to save for their children's secondary education.1 Jin Huang, Yunju Nam, Michael Sherraden, and Margaret Clancy build on this research and examine the association between financial capability and asset accumulation in CDAs for children's postsecondary education.2 The following is a summary of their study.


The news about the rising levels of student loan debt has been alarming. One estimate by the Consumer Financial Protection Bureau has the total amount of debt at nearly $1.2 trillion with many borrowers in default.3 A couple of strategies exist for those planning for college costs. One strategy is government-initiated, tax-advantaged college saving plans known as 529 plans. These plans are operated by state governments. The second strategy is the CDAs, which are a broader asset-building instrument, the proceeds of which can be used to finance postsecondary education, purchase a home, or start a small business. A compelling facilitator for individuals to use one of the aforementioned methods to save for college is financial knowledge and financial access, which, when taken together, is financial capability. Both components are critical: Financial knowledge allows the "ability to act," while financial access permits the "opportunity to act." Empirical evidence has underscored the virtue of financial capability and saving for college expenses. Using data from the SEED for Oklahoma Kids (SEED OK) experiment, it was found that financial capability is positively associated with the probability that participants will hold a college savings account for their child.4

Huang et al. expand on this finding by focusing on asset accumulation and financial capability when using CDAs to save for postsecondary education.

Data and Methodology

The authors used data from the SEED OK experiment in their investigation. The SEED OK intervention is a statewide randomized policy experiment that offered CDAs to treatment participants. It is based on the Oklahoma 529 College Savings Plan (OK 529 Plan). The benefits of the CDAs intervention are similar to those in the OK 529 Plan, namely annual contributions up to $10,000 ($20,000 for a married couple) are deductible on the state income-tax return. The intervention includes four additional components: 1) the state treasurer's office of Oklahoma opened a state-owned OK 529 Plan account for all treatment children and deposited $1,000 of SEED OK funds in the account, but treatment participants are not allowed to make deposits into the state-owned accounts; 2) treatment participants are encouraged to open their own participant-owned OK 529 Plan accounts, which requires a $100 initial contribution to open, but participants are allowed to make contributions to these accounts; 3) the SEED OK experiment provided a savings match based on participant's income for deposits made in participant-owned accounts; and 4) the Oklahoma treasurer's office communicated regularly (e.g., letters, postcards, and brochures) with treatment participants.

Participants in the control group received no information from SEED OK about the OK 529 Plan. They were not eligible for the state-owned account and were not offered any SEED OK financial incentive. However, they were allowed to open a participant-owned OK Plan.

Sample. The SEED OK experiment randomly selected 7,328 children from the population of infants born in Oklahoma in two three-month periods: April 1 through June 30, 2007 and August 1 through October 31, 2007. Racial and ethnic minorities were oversampled. Once the 2,704 primary caregivers of these children agreed to participate in the experiment and completed the baseline survey, they were randomly assigned to treatment (1,358) and control (1,346) groups.

Research Questions. The authors used data from the SEED OK experiment to answer three questions: 1) Does the intervention with CDAs, an effort to expand financial access, promote asset accumulation for a child's postsecondary education? 2) Does financial knowledge increase the participants' asset accumulation for their children? 3) Does financial knowledge's effect on asset accumulation change due to the expansion of financial access via SEED OK intervention?

Method. The authors used regression analysis to examine the relationships among financial access, financial knowledge, and assets accumulated in CDAs.5 They focused on two dependent variables: total savings, which is defined as the total amount of contributions that study participants made into participant-owned accounts, and total assets, which combines the account balances in the state-owned and participant-owned accounts. The regressions include two independent variables: one is the SEED OK treatment status (1 = treatment group, 0 = control group) and the other is a measure of the participants' financial knowledge. The latter is based on responses to three baseline-survey questions drawn from the 2004 Health and Retirement Survey.6 The questions evaluate an individual's ability to understand basic financial concepts, such as compound interest, inflation, and risk diversification. If participants answer all three questions correctly, they are deemed to have a high level of financial knowledge and are assigned the value of 1. All others are thought to have lower financial knowledge and are assigned the value of 0. The analysis also controls for demographic and other characteristics of the children, participants, and households.7


In light of the aforementioned research questions, Huang et al. offered the following two hypotheses: 1) both financial access and financial knowledge increase the amount of assets that participants accumulate in CDAs and 2) the association between financial knowledge and asset accumulation is stronger among treatment participants than among those in the control group. Their statistical analysis revealed the following findings:

  • The relationship between financial knowledge and saving behavior differs across the treatment and control groups, which implies that financial knowledge and financial access may interact in their relationship with asset accumulation.
  • The SEED OK significantly increased savings in the participant-owned accounts of about 7 percent of participants.
  • Financial knowledge and the CDA intervention interact to a statistically significant degree in their effect on the assets that parents accumulate for children's postsecondary education.
  • It appears that financial knowledge may have a greater influence on saving behavior than does the CDA intervention.
  • Very few participants with a low level of financial knowledge make contributions in participant-owned accounts, and few control-group participants make such contributions.
  • Conversely, participants with a high level of financial knowledge are 12 times more likely to make contributions than those with a low level of such knowledge.
  • The authors' analysis underscores the notion that the combination of financial access and financial knowledge plays an important role in changing financial behavior.
  • As witnessed by the participants in the control group, few families save without a CDA intervention. Thus, they are not future-minded enough to start saving in an OK 529 Plan account when their child is young. An implication is that the effectiveness of CDAs and other education savings programs diminishes if a family does not begin saving for a child's postsecondary education early.
  • SEED OK's automatic enrollment feature and $1,000 seed deposit increase the total assets of almost all treatment participants, regardless of their level of financial knowledge.
  • A key finding is that the effect of financial capability (the combination of financial literacy and financial access) is more important than the effect of either component alone.

Policy Implications

The authors suggest two policy implications based on their findings. First, the results "call for the creation of policy tools that enhance parents' financial preparation for their children's college by improving parents' financial knowledge but also by expanding their financial access." Second, "future CDAs should include program features that benefit socioeconomically disadvantaged participants and those with low levels of financial knowledge."

The views expressed here do not necessarily represent the views of the Federal Reserve Bank of Philadelphia or the Federal Reserve System.

[1]Jin Huang, Yunju Nam, and Margaret S. Sherraden, "Financial Knowledge and Child Development Account Policy: A Test of Financial Capability," Journal of Consumer Affairs, Spring 2013, 47(1), pp. 1–26.

[2]Jin Huang, Yunju Nam, Michael Sherraden, and Margaret Clancy, "Financial Capability and Asset Accumulation for Children's Education: Evidence from an Experiment of Child Development Accounts," Journal of Consumer Affairs, Spring 2015, 49(1), pp. 127–155.

[3]Rohit Chopra, "A Closer Look at the Trillion," Consumer Financial Protection Bureau Blog, August 5, 2013, available at

[4]See Huang, Nam, and Sherraden, 2013.

[5]The authors used quantile regressions as they are more robust than linear regression in dealing with outliers. This is important because only a small proportion of each group opened and made deposits in a participant-owned account (16 percent of the treatment group and 1 percent of the control group). Thus, the distribution of assets is highly skewed.

[6]Annamaria Lusardi and Olivia Mitchell, "Financial Literacy and Planning: Implications for Retirement Wellbeing, 2006," Working Paper 2006-01, Pension Research Council, University of Pennsylvania, Philadelphia.

[7]These include, among others, age, race, education, marital status, employment status, household size, homeownership status, and income.