Sending reminders to reduce money wasted on overdraft fees

Background

Americans paid an estimated $34 billion in overdraft fees last year. The Consumer Financial Protection Bureau (CFPB) finds that the majority of these fees were incurred on transactions of $24 or less and were repaid within three days. Most banks charge steep fees for overdrafting, making the overdraft operate like a very small loan with a very high interest rate. According to the CFPB, “If a consumer borrowed $24 for three days and paid the median overdraft fee of $34, such a loan would carry a 17,000 percent annual percentage rate (APR).” What’s more, the overdraft fee is often charged by overdraft incident, rather than proportional to overdraft amount, causing consumers to rack up fees faster than they may realize.

This research suggests that most overdraft fees are incurred unnecessarily and come at an exorbitant cost, especially for low-income consumers. To understand what can be done to reduce unnecessary overdrafting and non-sufficient funds (NSF) fees, we partnered with Freedom First Credit Union, a Community Development Financial Institution (CDFI) credit union based in Virginia.

As a CDFI dedicated to helping their credit-challenged members build up their finances through programs like financial coaching and a Payday Relief Loan, Freedom First was eager to explore ways of helping members to avoid unnecessary overdraft fees. Freedom First already offers a robust money management platform including online banking, overdraft protection services, and the ability to limit debit card transactions that would cause accounts to go into overdraft. Together, we were interested in building on this success to further enable their members’ financial success.

Key Insights

To gain a better understanding of factors that lead to members overdrafting and incurring NSF fees, we conducted site visits to Freedom First branches, interviewed staff members, and analyzed fee data for all 4,207 Freedom First members who had overdrafted or incurred NSF fees in the previous six months.

The behavioral diagnosis revealed that different groups of members overdrafted and incurred NSF fees for different reasons:

• Some members called into the Freedom First branches to complain when they were charged these fees. These members were often unaware they were being charged fees for overdrafting or were unaware that their account balance was low.

• However, over 250 members had more than 50 instances of overdraft or NSF fees in the previous six months. These customers were clearly suffering from deeper financial issues that could not be solved by a heightened awareness of fees or low account balances.

Experiment

We hypothesized that providing members with a reminder when their balance was low would help more members navigate their finances in ways that avoided unnecessary overdrafts. We randomly selected 5,200 credit union members and assigned them to one of three conditions.

• A control where the member did not receive a reminder.

• A balance reminder that highlighted fees to take advantage of loss aversion.

• An “opportunity cost” reminder that prompted them to think of other things the money could be spent on.

Those in the treatment groups received one of two different text message reminders if their balances dropped below $100. Reminders were sent no more than once every two weeks, to avoid overloading those who have continually low account balances.

To account for the different customer groups identified in the behavioral diagnosis, we implemented three different treatment conditions, along with a control.

Results

After a few weeks of running the experiment, Freedom First shut down the pilot in response to reports of problems occurring on January 23rd. There was a high volume at the call center related to the text messages, and given the available data, it seems that customers may not have received prior overdraft warnings as scheduled and perhaps multiple text messages were delivered on January 23rd. Not all of the calls were negative (some people called in to transfer money to the account that was low, which was the purpose of the text), but the call volume for the two hours after the messages were released was surprising.

While obviously, the way our experiment about using reminders to prevent overdrafting ended was disappointing, it also creates an important learning opportunity. Here are several important lessons learned:

• Include testing staff on the recipient list so that they can see the actual version that people in the test see. If we had been included on the send lists, we would have realized early on that the messages were not going out as planned in the early weeks. We also didn’t have clear evidence of what actually happened on January 23rd that caused the phone calls, and being on the recipient list could have helped clarify that.

• Pre-test with a small pilot if possible. Running the test with a hundred participants first might also have helped us to identify the sending errors earlier. A small pilot might have also brought to light some important data collection errors from the vendor early on, and then we could have fixed the issues in time for the larger test.

• Vet multiple vendors. The vendor that Freedom First used to execute the test had not previously run any broad text messaging services before. They were very eager to help and to build out new texting capabilities for their client, but ultimately their platform was not prepared for this volume or this test design. Freedom First already had an existing relationship with this vendor so chose to continue to work with them. In hindsight, we should have either found a vendor with ample experience and flexibility in text messaging or learned the limitations of the existing vendor to design a test within those constraints.

• Anticipate reactions to any change. The call center staff had been notified about the text messaging test but was not intimately familiar with the details. Surprisingly, the high call volume came several weeks into the test (we think because prior scheduled messages had not actually been sent) but having more details available about the test, particularly during the scheduled launch date, may also have helped prepare staff who are more peripherally involved in the test.