By Caitlin Anzelone & Jonathan Timm
The field of applied behavioral science continues to grow, with examples showing up from the private sector to the White House. Seminal research studies that nudged individuals into retirement savings plans and organ-donation registration have prompted people from diverse backgrounds to wonder how they can use these insights in their programs and services. The goal of the Behavioral Interventions to Advance Self-Sufficiency (BIAS) project was to learn how behavioral insights could be applied to human services agencies that serve low-income and vulnerable populations in the United States. What did BIAS learn about applying behavioral insights in this field?
BIAS has now completed 15 randomized controlled trials in child care, child support, and work support programs. While each intervention was designed to respond to unique challenges faced by particular programs, seven behavioral concepts were used in almost every site. These seven concepts led to the “SIMPLER” framework: Social influence, Implementation prompts, Mandated deadlines, Personalization, Loss aversion, Ease, and Reminders. This article summarizes each part of the framework and shares examples of how these concepts were used in BIAS interventions. Practitioners and researchers may find this guide useful when designing behavioral interventions to improve the outcomes of social programs. All of the findings discussed are statistically significant.
Persuasion by society, peers, or a person of influence can affect people’s decisions and actions.
In Texas, the BIAS intervention encouraged incarcerated parents who owed child support to submit applications for child support order modifications. Among other strategies, the new outreach made use of social influence by explaining that “Other parents like you have had their orders lowered!” The intervention led to a large increase in the percentage of parents who submitted applications.
Implementation prompts — which encourage people to plan the precise steps they will take to complete a task — can help move people from intention to action.
In Indiana the BIAS intervention prompted parents to develop specific plans to make it to their child care subsidy renewal appointments on time, with the proper documents. This strategy was one of a set of behavioral interventions that increased both the percentage of parents who attended their first scheduled renewal appointment and the percentage of parents who completed the process in one appointment.
Without a fixed deadline for accomplishing a task, it can be easy for people to procrastinate or assume that they will get around to doing it eventually.
In the BIAS Paycheck Plus test in New York, the program advertised an artificially early deadline for attending a meeting, knowing that people who missed the first due date would have time to catch up before the real deadline. People who were given the behavioral intervention that included the earlier deadline attended the meeting sooner and in higher numbers than those in the control group.
Efforts to personalize information or give customers personal assistance through a difficult task can improve outcomes.
BIAS interventions often sent out communications that were personalized with people’s names or prepopulated with information relevant to their cases. In one of the Indiana child care studies, parents were sent personalized referrals to quality-rated child care providers close to their homes, accompanied by a handwritten note to call the agency with any questions. These notes contributed to an increase in the percentage of parents who selected a highly rated provider.
In some sites BIAS interventions also included active personal assistance. In child care studies conducted in both Oklahoma and Indiana, having a staff member make contact with parents to assist them led to greater impacts than written communications alone.
Loss aversion is the tendency to prefer avoiding losses to achieving equal-sized gains, relative to a reference point. Sometimes just rewording a message can lead people to a different outcome.
In the Los Angeles Temporary Assistance for Needy Families study, the team designed two notices that conveyed the same information except that one emphasized the benefits participants would gain by attending an appointment, and the other emphasized the losses they might incur by not attending. When compared with the status quo outreach, the notice that emphasized losses increased positive engagement while the notice that emphasized gains did not produce an impact.
Studies in psychology have shown that people can process, absorb, and recall only a limited amount of information at one time. Thus, a central tenet in behavioral design is that making things as easy as possible can increase the likelihood that people will act.
The BIAS team’s work in Washington sought to increase the number of incarcerated parents who submitted applications for child support order modifications. The intervention made the process easier for parents by eliminating the need for them to request an application; it mailed them the materials before they even asked. It included a tip sheet specifying which questions in the application had to be answered and sent several other communications, including e-mail messages. This intervention led to a large increase in the number of submitted applications and completed modifications.
Reminders reduce the mental effort required to complete an action by providing a cue that the task still needs to be completed. Studies in behavioral science have found that reminders can spur people to action in many fields including health, voting, and personal finance.
Almost all of the BIAS sites employed some form of reminder, most notably the child support interventions in Franklin and Cuyahoga Counties, Ohio. Simply sending a monthly reminder through mail, robocall, or text message to parents who owed child support but were not getting any other monthly invoice increased the number of parents making payments.
Behavioral science has the potential to help improve a range of social programs, but there is no one-size-fits-all solution. Evidence suggests that the SIMPLER framework is a helpful place to start. Let us know what techniques and frameworks you are using and what you’ve found most helpful by tweeting us @CABS_MDRC.