Stay Informed!

Posted on 03/28/2011 @ 01:00 PM

Tags: Behavioral Economics, Innovation, Innovators

Take Home Refunds Seven Times as Large: The Campaign for Working Families VITA Site Helps Self-Employed Clients Follow Through on Intentions to Prepare for Tax Day

A project with the Campaign for Working Families of Philadelphia.

Program background: The Campaign for Working Families (CWF) is a partnership that promotes increased resources for low-wage working families by providing free filing of the federal Earned Income Tax Credit (EITC) and connecting Philadelphia residents to other tax credits, work supports and asset-building resources.

The goal of our project was to help self-employed clients receive the refunds they earned by increasing their preparation for tax day by having their paperwork and receipts in order.

Behavioral insight: Preparing for tax time is a hassle and the short-term stress may eclipse the longer-term reward of a potential refund. Rationally, if we believe something is important, we should simply follow through on our intended plans. In a strict cost-benefit analysis, the hours it may take to collect tax information is surely worth the possible cash reward. But in reality, it is easy to let the present costs of hassle and stress eclipse future rewards.

CWF hoped the information and persuasive messages conveyed during orientation would motivate people to overcome the perceived stress of tax preparation and actually gather and organize the necessary tax information. But this was not happening. We hypothesized that the hassle of preparation might pose a significant barrier, and that people needed reminders and a sense of accountability to follow through on their intentions to prepare.

To make the reminders as powerful as possible, we wanted to leverage the following behavioral insights:

• Commitment and consistency: Telling others that we intend to behave in a certain way helps us keep our word. We like to appear consistent to ourselves and others, so we find it important that our actions and beliefs align, or at least appear to align. In fact, some early behavioral theorists considered the desire to be consistent as a central motivating human behavior. (Festinger, 1957; Heider, 1946; & Newcomb, 1953).
  
  For example, in one study people were called and asked to predict what they would say if asked to spend a few hours volunteering for the American Cancer Society. Most people wanted to appear charitable, and many predicted they would agree to help. This small commitment device produced a 700 percent increase in volunteers when representatives from the American Cancer Society came to their door asking for volunteers a few days later. (Sherman, 1980).
  
  Our desire to be consistent can be especially effective when faced with written evidence of a commitment in our own handwriting. As the well-known behavioral theorist Robert Cialdini explains, “There is something magical about writing things down.”
• Planning or "implementation intentions": There is evidence that implementation intentions can also help people follow through and accomplish a desired goal. (Gollwitzer, 1993). An implementation intention spells out the when, where and how of what one will do to reach a goal. For example, asking people to create a “voting plan” (What time will you vote? How will you get to the voting station?) significantly increased voter turnout in comparison to simply asking if someone would vote and encouraging the person to do so. (Nickerson & Rogers, 2010). Detailing the steps needed to implement our goals helps us follow through on our intentions.
• Channel Factors: Making things easy makes a difference. Research tells us that adjusting small nuances in our situation can have a surprising impact on our ability to close the gap between our intentions and actions. In a well-known study by Leventhal, Singer and Jones (1965), college seniors were given persuasive messages about the value of an inoculation against tetanus. While the messages were effective at changing the students’ beliefs and attitudes, few actually took the step of getting a tetanus shot. Other students received the same messages but were also given a map of the campus with the infirmary circled and urged to think about a particular time and route they would take to the infirmary. This small adjustment led to a significant increase in the percentage of students who actually got their inoculation.

Intervention: Building on the behavioral findings described above, the CWF project sought to improve self-employed tax assistance clients’ level of tax preparedness by asking then to formulate a tax preparation plan (implementation intentions), write it down for the CWF staff to see (commitment) and then sending the preparation steps back to the clients (consistency) along with another copy of the preparation worksheet (ease or channel factor).

Every self-employed tax client at CWF must complete an orientation before the tax session. We varied the orientations between treatment and non-treatment sessions. Every other orientation was a treatment or a control session, which takes advantage of the fact that a client’s selection of one class over the next is largely random.

At the close of the treatment pre-tax orientation, the group was asked to complete a form detailing their three next steps in preparing for their tax appointment. One of the steps had to be completion of a tax preparation worksheet handed out by CWF.

This form would later be used as their appointment reminder. A few days after orientation, clients received this sheet in the mail with their appointment date, an extra tax prep worksheet, and a list of preparation next steps written in their own hand. The control group went through the same orientation but did not go through the next steps exercise. They received a reminder letter in the mail with only their appointment time.

Findings:

Experimental Design, Sample Size = 41 clients

The treatment seemed to significantly impact clients’ tax refund amount. The treatment group had refunds that were significantly larger (in fact, seven times larger) than those in the control group: $1,837 for the treatment group compared to $241 in the control group. Note that because a taxpayer’s EITC amount is correlated with final tax refund, our comparison is based on their tax refund without the EITC.

This is a large effect, and we wanted to be confident that it was not the result of different populations in the treatment and control groups. That is, was there something about the treatment group that made them more likely to receive higher refunds? In statistical analyses, we found the two groups to be randomly distributed and similar. Each group had similar EITC refunds, W-2 incomes and business incomes.

Discussion: These findings may be very powerful, with the potential to increase refunds—and therefore incomes—for thousands of self-employed people. More broadly, behaviorally-informed reminders may help people follow through on a variety of high-stress, high-hassle goals like creating a budget or completing complex forms like the FAFSA. More research is needed before we can be confident that these results are replicable with a self-employed or even more general population. This is especially true given the study’s relatively small sample size. Because the intervention had such an unusually large effect, this is an especially exciting area for further exploration.