What kind of messages get ignored? What kind prompt you to do something?
Those are questions that a small group of behavioral scientists at the White House have been working on since early last year.
The Social and Behavioral Sciences Team is seeking ways to improve government efficiency and access to government programs through easy, low-cost interventions.
The team issued its first annual report Tuesday, and Maya Shankar — chair of the team, spoke with Robert Siegel about the findings.
Interview Highlights
On how to improve rates of re-enrollment in the Thrift Savings Plan, a 401(k)-like plan for federal workers, for 140,000 members of the military who had to re-enroll by January 2015 or risk losing the ability to contribute
We redesigned the Department of Defense's business-as-usual letter to include clear action steps for enrolling. We also put a postscript into the email message. ... Research shows the postscript is the second thing the visual system looks at when we open a letter, open an email. So you can highlight key messages and key action steps in that particular location on the letter.
We redesigned the letter and then we tested the impact of the standard business-as-usual letter with the redesigned letter that used behavioral insights. And we found within a week that there was a 22 percent increase in re-enrollment rates for those who received the redesigned letter. The DOD immediately took those insights and brought them to scale for the remainder of the email outreach campaign so that everybody was able to benefit from those insights moving forward.
... [The rate of re-enrollment] was from 23.5 percent to 28.7 percent — 3,770 service members who were re-enrolling as a result of the redesigned letter.
On the scale of the federal government, where a change of a few percentage points can affect thousands of people
That's what makes the nature of our work so thrilling — which is the scale of impact that we can have from these very small tweaks that are very low cost and very easy to implement within federal programs.
On a failed attempt to use peer pressure to discourage improper drug prescriptions
The Center for Program Integrity at the Centers for Medicare and Medicaid Studies sent letters to doctors simply notifying them of how their prescribing rates compared to that of their peers ... sort of similar to when you might receive an energy bill in the mail and it shows you how your energy consumption compares to that of your neighbors. We ran this and found that actually it didn't work. ...
So it's compelling a re-evaluation of how can we better communicate with physicians. So we're thinking of the envelope. Are people even opening these letters?
[There are features — for example, use of automatic paid postage stamp — ] we're exploring in a follow-up pilot in which we try to figure out way to get these letters in the hands of the physicians, so they can learn from the information and maybe change their prescribing patterns accordingly.
On how pervasive a role behavioral sciences can have in government actions and policies
We think that government programs work best, and Americans are best served, when they are easy to access and when program information is presented clearly. I think the critical thing that behavioral science teaches us is that when programs are not designed in this way, the consequences for Americans [are] much larger than we might think.
Every year 20 to 30 percent of college-accepted high school graduates in urban districts fail to matriculate in college in the fall, because they haven't filled out required college enrollment materials. This phenomenon is known as summer melt. What research shows us, and what the Social and Behavioral Sciences Team worked on, is sending these students eight low-cost text messages that reminded them of these deadlines and provided them more information and access to personalized links so that they could complete this information. Now, some might say it's so straightforward and common sense to simply send students text messages. But this can make the difference between a student going to college or not going to college.
These are students who have applied to college, have gotten in, and might not have had access to the kind of professional assistance a lot of people might have had access to within their own schools or within their communities. And all we're doing is providing that sort of assistance, so that students who want to go to college can actually go.
Copyright 2015 NPR. To see more, visit http://www.npr.org/.
