Is there a better way to offer incentives for college savings?

Background

Financing a child’s college education is increasingly difficult, especially for low- and moderate-income families. The rising cost of attending college has outpaced inflation for the past three decades, meaning college has never been more expensive. Child Savings Accounts (CSAs) offer families a vehicle for families to start saving for their future education and, perhaps more importantly, set expectations and foster a college-bound identity.

While effective, encouraging parents to save for their child’s future college is notoriously hard. Enrollment rates and account usage tend to be quite low, despite how many people for whom they would be beneficial. This year, we continued our partnership with St. Louis Office of Financial Empowerment (OFE) to explore how we might encourage greater numbers of families to open and save in a CSA.

Key Insights

Encouraging families to save with a child savings account poses several particularly challenging issues to grapple with:

• People often struggle to think about and save for their future selves – there is high psychological distance to our future selves. In child savings, this disconnect between now and our futures is even greater because the future savings are for another person (the child).

• People are particularly unmotivated to engage in activities where the future benefits are ambiguous. This is especially relevant for college savings, where individual deposits have such a limited impact on the future. A child attending college is the collection of multiple actions in concert and not solely dependent on savings.

• Making multiple, one-off payments is a difficult process. A person either needs to go into branch or they have to navigate often complicated online interfaces; both of these methods require non-trivial amounts of logistical and cognitive effort.

Experiment

While our previous work focused on smaller tweaks to communication, we wanted to focus on larger, more structural changes. Most CSAs offer some amount of a match as an incentive to encourage deposits. We thought that the incentives of the match could be designed to be more behaviorally effective and worked with the St. Louis OFE to re-design the incentive structure for the CSA accounts.

The new system emphasized recurring contributions, since automated savings is less onerous than making large lump sum deposits and will continue beyond the initial set-up behavior. If someone sets up a recurring transfer under the new incentives, they received $50 cashback immediately to spend on personal expenses. After they have completed six direct deposits, or an equal number of individual deposits, another $50 would be contributed to their child’s college savings account.

On top of the cash reward, the system utilized a “regret-lottery” rewards program, where the winner is drawn at random but only individuals who have already signed up for an automatic deposit or have contributed a one-time deposit can claim the prize. The prize for individuals who have automatic deposits set up is larger than the prize for one-time depositors. When a person won, but had not set up a direct deposit or made any deposit, they were given two months to do so in order to claim half of the prize.

We tested the new incentive structure through a randomized controlled trial where participants were evenly and randomly assigned into one of the three groups shown to the left.

Results

Starting in December of 2018, we implemented the new incentives structure across all 53 schools in the College Kids program in St. Louis and across approximately 2,500 children.

We found the new incentive structure did not appear to lead to more families to contribute to the child’s college savings account. However, the analysis indicated that the new incentives encouraged families to save more often and more when they saved. When looking only at families who contributed at least once to the savings account, we saw the families offered the new incentives tended to save more in total. The ‘savers’ made significantly more deposits into their accounts than the control and the total size of their deposits were marginally greater.

Taken together, the new incentive structures did not significantly influence families to start saving into the college savings account, but the intervention did lead to increased overall dollar amounts contributed into the account and the number of deposits made by the families. The number of families who made any deposit remains relatively small, but we believe that over time the families who contribute are more likely to create lasting savings behaviors that can lead to increased total savings in preparation for college.