Experimental Design
On June 24, 2015, Sallie Mae, a national financial institution specializing in student loans, launched the FICO Score Open Access program and began providing free score access to customers through the their website. Clients who logged in to the website saw a visual display that included their FICO Score beside a barometer showing the range of possible FICO Scores. The display also listed two “reason codes” that explain the key factors contributing to the individual’s score, such as limited credit history or account delinquency.
While all customers had the ability to log in and view this information, many borrowers may not have been aware of the new program. To test the effect of providing information about a borrower’s FICO Score, we experimentally vary knowledge of FICO Score availability through additional communication about the program.
A. Sample Population
The sample for the experiment consists of the 406,994 student loan borrowers who held a loan with Sallie Mae at the start of the FICO Score Open Access program and continued to hold that loan for the following two years. Table 1, Panel A presents summary statistics of the demographic characteristics of our experimental population provided by Sallie Mae. The average age of borrowers in our sample is 25 years old with just over half currently attending school, while the remainder are out of school and, therefore, have started paying off their student loan debt.
B. Experimental Conditions
Prior to the roll-out of the FICO Score Open Access Initiative through Sallie Mae, borrowers were randomly assigned to one of four experimental groups – three treatment groups and one control group. Roughly 90 percent of our sample was assigned to one of the three treatment groups, while the control group contained the remaining 10 percent of the sample. Borrowers assigned to the treatment groups received email communications from Sallie Mae alerting them to the availability of their FICO Score and providing instructions on how to access the information while control group members did not receive any communication about the program beyond what was provided on the provider’s website.
All emails included a short description of the FICO Score and informed borrowers that their score was available to view. The email also included a link to log in via the loan provider’s website. Treatment group members received these communications once per quarter on the date that scores were updated informing them that their FICO Score had been updated and, again, providing a link to log in to view the score. Due to privacy considerations, no personalized information was included in the email itself.
Borrowers who received an email were randomly assigned to be in one of three conditions: (1) baseline, (2) economic consequences, and (3) social influence. In the baseline condition, borrowers received only the information described above (Figure 3). The two additional conditions included the same information as the baseline email as well as additional messaging. In the economic consequences condition, clients received an email that was intended to emphasize the impact of the FICO Score on economic outcomes (e.g., “When you apply for credit – whether it’s a credit card, car loan, student loan, apartment rental, or mortgage – lenders will assess your risk as a borrower...”). Building on research demonstrating the effectiveness of messaging informing individuals of prosocial actions of their peers (Allcott, 2011; Ayres et al., 2012; Cialdini and Goldstein, 2004; Kast et al., 2012), the social influence condition included messaging informing readers that their peers were taking actions to improve their credit (e.g., “Many of your peers are building strong financial futures. You can, too, by effectively managing your student loans.”).
C. Experiment Timeline
The three treatment groups in the main sample received eight quarterly emails starting in June of 2015. Each treatment group received their assigned message for three consecutive quarters (June, September, and December of 2015). However, beginning in 2016, all three treatment groups received only the content included in the baseline email message. In other words, clients in the economic consequences and social influence conditions began receiving the baseline message starting in March of 2016; clients in the baseline condition continued to receive the baseline message. The control group never received any direct communications about the program.
The experimental design included a separate population of 37,393 borrowers – the “discontinued sample” – that received quarterly emails for only three quarters. This sample was also split into three treatment message groups, and received quarterly email communications in June, September, and December of 2015. Our main analysis focuses on the 326,609 treatment group members who received quarterly communications through the end of the intervention in June of 2017.