HERE’S WHAT THE WHITE HOUSE’S BEHAVIORAL SCIENCES TEAM DISCOVERED IN ITS FIRST YEAR

The report runs down a series of what are effectively experiments designed to gauge the effects of making various tweaks to the wording and user interfaces government agencies use. The results varied, but there's some promising stuff here: http://sciof.us/1JbVIIE
  1. One initiative helped encourage people to apply for government benefits to which they were entitled:
    "Compared to no message, the most effective message nearly doubled the rate at which Servicemembers signed up for Thrift Savings Plan (TSP). Emails informed by behavioral insights led to roughly 4,930 new enrollments and $1.3 million in savings in just the first month after the emails were sent. DOD is now scaling up this intervention by sending periodic emails informed by behavioral insights to Servicemembers about the benefits of TSP."
  2. Another one helped encourage low-income students to apply for federal student loans.
    "To help students enroll in college, SBST and the Department of Education’s office of Federal Student Aid provided technical expertise to researchers and nonprofit uAspire on messages notifying college-accepted, high school graduates of required pre-matriculation tasks for their respective colleges. A series of eight personalized text messages to low-income students reminding them to complete these tasks led to a 5.7 percentage point increase in college enrollment, from 66.4 to 72.1%."
  3. And another tweaked the government's approach to getting people to pay money they owed.
    "To increase debt recovery from individuals with outstanding non-tax debt, SBST worked with the Department of the Treasury’s Debt Management Service (DMS) to redesign a collection letter. No difference in payment rates was observed, but changes such as shortening the web address for making an online payment led 45 percent more individuals to pay online, representing an increase from 1.5 to 2.2 percent. DMS has now permanently shortened the web link in the collection letter."
  4. Some of this stuff sounds bureaucratic and unexciting, but the underlying logic is important: In theory, if you can run a big and robust enough experiment to test the effects of these sorts of wording and interface changes, the onetime cost of running that experiment and making the tweak can make a big difference in people's lives.