Nudges and Learning: Evidence from Informational Interventions for Low-Income Taxpayers

Do informational interventions create one-time nudges or permanent changes in behavior? We study how taxpayers respond to informational interventions that alert them of their eligibility for the Earned Income Tax Credit using population-level administrative tax data. The empirical analysis is based on a natural experiment in 2005, a randomized experiment in 2009, and quasi-random audits between 2006 and 2009. The evidence from each of these settings indicates that the informational interventions cause economically significant increases in EITC take-up in the short-term, but there are little to no long-term increases in EITC take-up.