Using Behavioral Economics to Inform the Integration of Human Services and Health Programs under the Affordable Care Act . Eliminating paperwork by completing it on consumers’ behalf


A randomized, controlled experiment tested the impact of tax preparers actually filing SNAP applications for eligible consumers, compared to completing applications and providing instructions about how and where the consumers could file them. Low-income clients at H&R Block offices were provided information about the monthly SNAP benefits for which they qualified. In addition to filing tax returns for these clients, H&R Block gathered additional information needed for the SNAP application.

The clients were randomly assigned to one of three levels of assistance in completing the SNAP application: Basic (where only a blank application was given to the client), which served as a control group; Plus (where the clients were interviewed and given an application completely filled out by H&R Block staff, plus information about where and how the clients could file the application); or Full Assistance (where the H&R Block employee not only completed the application but also, with the client’s permission, filed it for the client by sending it to the SNAP office, which completed the enrollment process by contacting the client for a telephone interview).

Researchers found that participation in the Full Assistance model substantially increased the probability of completing a SNAP application. Roughly 40 percent of Full Assistance recipients applied for SNAP, compared to 22 percent of the control group, representing an 80 percent relative increase. By contrast, Plus-level help, which gave consumers a fully completed SNAP application and instructions about where and how to file it, yielded no statistically significant difference in application filing, compared to the control group.43

A similar randomized, controlled experiment involved college student aid application forms. When H&R Block staff completed and filed college student aid forms on behalf of tax clients, 55.9 percent successfully completed the application process. This represented a 40 percent relative increase, compared to the 40.2 percent who applied in the control group that received only general information about college aid. By contrast, no statistically significant effects were observed when H&R Block gave families written, personalized estimates of their likely eligibility for student aid, data about potential tuition costs at nearby colleges, and information about how to complete the application process on their own.44

A randomized, controlled experiment in a low-income, predominantly Latino community in Boston came to similar conclusions. This study compared Massachusetts’s normal Medicaid outreach methods with having case managers from a community-based organization file applications for children, then following up over time to address emerging problems.45 The state’s normal outreach methods involved mailings, door-to-door canvassing, radio advertisements in Spanish, grants to community organizations, and a toll-free call center. Among the children who received assistance from community-based case managers, 96 percent enrolled in Medicaid, and 78 percent retained coverage continuously throughout the study’s one-year follow-up period. By contrast, only 57 percent of the children receiving the state’s standard outreach enrolled, and just 30 percent retained coverage continuously throughout the following year.46

43 Diane Schanzenbach, “Experiment Estimates of the Barriers to Food Stamp Enrollment,” (Madison, WI: Institute for Research on Poverty, University of Wisconsin- Madison 2009).
44 Eric P. Bettinger, Bridget Terry Long, Philip Oreopoulos, and Lisa Sanbonmatsu, “The Role Of Simplification And Information In College Decisions: Results From The H&R Block FAFSA Experiment,” National Bureau of Economic Research Working Paper 15361 (September 2009). The results described in the text refer to dependents. Among independent adults, the form completion intervention nearly tripled the likelihood of filing for student aid, raising the form submission rate from 13.8 percent to 39.5 percent.
45 Glenn Flores, Milagros Abreu, Christine E. Chaisson, Alan Meyers, Ramesh C. Sachdeva, Harriet Fernandez, Patricia Francisco, Beatriz Diaz, Ana Milena Diaz and Iris Santos-Guerrero. “A randomized, controlled trial of the effectiveness of community-based case management in insuring uninsured Latino children.” Pediatrics. 2005;116:1433–41.
46 In a final example, researchers looked at the behavior of people interested in opening their first basic bank account. The researchers hosted informational workshops for those interested in opening an account. In some of the workshops, the attendees were given a referral letter and told to take it to a branch to complete the application. In other workshops, a bank representative was present and applicants were given the entire application and encouraged to fill it out there. Not only were those who attended a workshop with a bank representative present more likely to open an account, they were also more likely to use different features of the account, such as direct deposit and ATMs. Bertrand, Mullainathan, et al., op cit.

View full report


"rpt_BehavioralEconomics.pdf" (pdf, 805.46Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®