May 29, 2020, 11:00–12:30
Room Zoom Video
Public Economics Seminar
Abstract
Sweden introduced a phase-out of the earned income tax credit in 2016. As a consequence, taxpayers belonging to the top 5 percent of the earnings distribution, already facing high taxes, experienced a 7 % reduction intheirnet-of-taxshares. Whileexploitingrichfull-populationadministrative data up to 2017, we evaluate earnings responses to the reform. When graphically and econometrically comparing earnings growth at different segments of the distribution, we estimate a significant relative earnings reduction in the treatment group immediately appearing in 2016, and growingin2017. Theimpliedearningselasticityisfairlylowandaround0.1. We interprettheessentialfeaturesoftheresponseusingasimulationmodel,in which people have noisy perceptions of the piece-wise linear tax code. To simulate the empirically observed response, we need to add more noise to perceptions than what is motivated by earnings uncertainty alone.