AdmitHub: How to design for behavior change

How can we increase the number of on-time FAFSA applications?

Using behavioral design, our intervention increased odds of applying on time

The Challenge

Every year college students have the opportunity to apply for financial aid to use towards college costs. Submitting an application for aid is often a big part of student success. It not only reduces the financial burden of college, but it is also one of the strongest predictors of whether or not a student enrolls in college, and ultimately persists to graduate.   Sadly, at least 2 million eligible students fail to submit The Free Application for Student Federal Aid (FAFSA). By failing to submit the application, students are leaving over $2.5 billion on the table in the form of loans, grants and scholarships.

Our team at  Irrational Labs was tasked with helping to improve FAFSA application rates among students at West Texas A&M University, a school where the vast majority of students would be eligible for financial aid, but the year-over-year repeat FAFSA application rate is only 54%.

Applying for the FAFSA is free, easier than ever before, and makes college more affordable—so what was holding West Texas A&M students back from applying?  Solution designers might attempt to solve this problem by asking students why they don’t apply and then building solutions based on their answers.  What happens when you take this approach? Students commonly say they don’t apply because they don’t think they are eligible to receive money. The logical solution, based on this data, is to give these students more information about financial aid eligibility so that they can correct their beliefs. However, when Harvard researchers tried this approach by giving students information about their eligibility, they found that information alone had no positive effect on application rates.

At Irrational Labs, we believe that principles from behavioral economics should be leveraged to help design solutions to overcome one of the biggest hurdles to attending and graduating from college: obtaining the loans or grants that make it affordable.  Behavioral economics has shown that the environment of decision-making can have a stronger influence on behavior than someone’s personal attitudes and beliefs.  Instead of studying what people say, behavioral economics pushes solution designers to study the decision environment and what people actually do.

As a first step, we conducted a Behavioral Diagnosis, a step-by-step analysis of all the actions someone goes through in order to perform a desired behavior.  In this case, we mapped out every action and decision a student at West Texas A&M makes in order to file a FAFSA by the deadline. There are over 20! The first decision, the decision to apply, happens prior to knowing if one actually qualifies for aid. If a student wants to find out if they qualify, they have to fill out a form that requires a pile of paperwork, including bank statements, W2s and tax records for the student and their parents.

Next, we identified the psychological barriers and biases at play for each step in the application process. We identified several obstacles that account for why students fail to file (and yes, it goes well beyond the complexity of the form). One key barrier we identified is that the decision to submit requires a great deal of effort.  Students need to put in time and effort to gather information to determine if they’re eligible and seek out information from others. They have to be confident enough in the probability that they will receive aid, to set aside the time required to collect all of their documents.   Students may have wanted to apply, but due to the time, effort and uncertainty involved, they procrastinate and end up not filing.

Procrastination happens when there is an asymmetry between the costs that someone has to pay now and the rewards expected in the future. Since the rewards to applying for a FAFSA are uncertain (a student has no way of knowing whether she will receive aid, and if so, how much) the cost of applying for students who have never applied before outweighs the rewards. How could we mitigate this?

One strategy would be to decrease the actual costs of applying, by streamlining the process through hands-on, one-on-one assistance. A study within the same Harvard research paper found that this strategy can have a positive impact. We believed that we could take this strategy one step further by decreasing the perception of the costs of applying.

Hypothesis

In the case of applying for the FAFSA, the act of needing to make a decision is itself a cost, since students need to spend time going back and forth trying to figure out if aid is an option or if it’s worth the upfront time investment to apply. Not only that, but when the decision is framed as being optional, it suggests that many students don’t need to apply. What if we tried to remove this decision point for students by eliminating the sense that applying is optional? Research has shown that an active decision point can be removed by implying that a behavior is simply the next course of action that someone should take. For example, doctors can increase vaccination rates when they imply that getting a vaccine is just the next step for the child at their doctors’ visit. We thought that a similar strategy could work to improve FAFSA application rates.

We hypothesized that converting students from an optional mindset (this is something I could do) to a required mindset (this is something I’m supposed to do) would lead to improved application rates. By removing the sense of optionality, filing a FAFSA is normalized, and students are freed from feeling like they should weigh their options or seek out further information.

Experiment

To test our intervention strategy, we ran an experiment in partnership with AdmitHub, a higher education company that uses AI and mobile messaging to provide students with more accessible resources. We sent over 2,000 students mobile messages reminding them about the FAFSA deadline on April 15th.

The students were randomly assigned to one of two experimental conditions. In the Treatment Condition, two mobile reminders emphasized that applying for a FAFSA was part of the enrollment process (removing the sense of optionality). In the Control Condition, students were sent similar reminders, but the language about this being part of the enrollment process wasn’t included. It was just a reminder. We predicted that students in the Treatment Condition (those who saw messages that made applying feel required) would submit a FAFSA by the deadline at a higher rate compared with students in the Control Condition.

Results

In line with this prediction, we found that for students who did not apply the previous year, the odds of applying for a FAFSA on time were triple for students who felt that applying for aid was just a part of the enrollment process.  Among this group of students, only 2% applied on time in the Control Condition, whereas 5% applied on time in the Treatment Condition.

Based on these early results, if this low-cost intervention were to be scaled across the United States, an estimated 230,000 more students would apply for financial aid. 

Assuming that 85% of these students would be awarded aid, this translates to almost $200 million dollars that could help students afford college, persisting to their next semester and graduating with less debt.

While our behavioral intervention could have a major impact on FAFSA application rates, it does have a potential limitation if it were to be rolled out at scale: It could create an unnecessary burden on students who don’t qualify for the FAFSA, who could end up wasting an hour of their time applying.  Institutions would need to consider whether it’s worth adapting the practice of implying that filing is part of the enrollment process, given this caveat.

Given that financial challenges are a top predictor of student drop-out, we speculate that this intervention would have the long-term impact of increasing future college enrollment.  While the data from this study don’t allow us to address this education outcomes directly, we did find some evidence that applying for the FAFSA has immediate positive spillover effects:  Students at West Texas A&M were more likely to register earlier for classes (on average 3 days earlier) if they had applied to the FAFSA than if they did not.

By not filing the FAFSA, millions of students miss the opportunity to receive grant, loan or work-study financial aid that could help them afford and attend college. Our findings underscore the impact of how the design of our decision-making environment can massively influence our behavior:  In this case, one small change could increase the likelihood of college attendance.  We would not have figured this out, and unlocked millions in aid, if we had relied solely on interviewing students.  This experiment highlights the incredible impact that can be captured when solution designers put their energy on studying what people actually do vs what they say they do and focus the solution design on removing the psychological barriers in their way.

Interested in applying behavioral science?

Read More Case Studies

Health

Changing health outcomes requires understanding human psychology.

Financial Services

Design for the psychology of money.

Loneliness & Social Isolation

We use behavioral economics to design the environment for happiness.