Abstract

We match survey data and administrative data on true program participation for a large welfare program in Germany to determine misreporting on welfare receipt in survey data and analyze the effects on measured welfare dynamics in survey data.