The replication package contains data that we are making publicly available. We use two main datasets: census data (Swiss Federal Statistical Office) and taxpayer data (Swiss Federal Tax Administration). Census data are aggregated at the municipality level. Taxpayer data are proprietary. We aggregated these data into four income bins for each municipality. There are fewer than 5 observations in one of these income bins for a bit less than a third of tax data units (municipality-year) and the Swiss Federal Tax Administration asked us to code these bins as “missing” in the dataset we are making publicly available. Hence, the full dataset we are using and the dataset we are making publicly available differ slightly. The ReadMe file includes an Appendix with all Tables and Figures of the paper produced with the restricted, publicly available data.

We acknowledge support from the Swiss National Science Foundation (Sinergia Grants No. 130648 and 147668; Division I Grant No. 182380).