Encouraging Tax-Time Savings with a Low-Touch, Large-Scale Intervention: Evidence from the Refund to Savings Experiment

Abstract

Low- and moderate-income households often struggle to save, but the annual tax refund represents a prime opportunity for these households to save toward their financial goals or build their emergency savings. This paper presents the results of a randomized, controlled experiment embedded in a free tax-preparation product offered in 2013 to low- and moderate-income households. The experiment involved approximately 470,000 filers and assessed the impact of behavioral interventions on their savings behaviors. The results show that filers exposed to the treatments, which involved the established behavioral-economics techniques of anchoring, choice architecture, and persuasive messaging, were more likely than a control group to save their tax refund and, on average, saved more of the refund. A follow-up survey of these tax filers found that the treatments were associated with saving more of the tax refund six months after filing. The findings also show that anchors encouraging filers to deposit certain amounts are more effective than persuasive messaging emphasizing savings.