The Impact of a Digital Message on Tax Compliance: Evidence from a Swedish Field Experiment Using Model-based Bayesian Inference

How to cite

Angelov, Nikolay; Johansson, Per. 2022. "The Impact of a Digital Message on Tax Compliance: Evidence from a Swedish Field Experiment Using Model-based Bayesian Inference." AEA RCT Registry. AEARCTR-0009408. https://doi.org/10.1257/rct.9408-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Nikolay Angelov* , Per Johansson

* Primary investigator

Registry ID
AEARCTR-0009408
DOI
https://doi.org/10.1257/rct.9408-2.0
First registered
Last updated
Status
In development
Trial dates
2022-03-17 – 2022-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/9408

Abstract

The purpose of the present study is to estimate the effect of a low-cost intervention (either one of two versions of a digitally disseminated message) on the subsequent tax compliance of Swedish taxpayers. The population consists of taxpayers who, according to automatically exchanged data from tax agencies abroad (CRS-data), had foreign dividends or interest from any of the three neighbouring countries Norway, Finland, or Denmark, during the tax year 2020. Within the CRS, tax authorities obtain information from financial institutions in their own jurisdiction and automatically exchange that information with other jurisdictions on an annual basis.

Experimental design

The sample frame consists of 10,000 individual taxpayers, 5,000 of whom are randomly assigned to one of two possible treatments. Treatment is defined as receiving one of two digital messages. Half of the treated (2,500) receive message A and the other half of the treated (2,500) receive message B.

Design details

Randomization method
Simple random sampling dome by a computer, using R.
Randomization unit
Individual taxpayers
Clusters
10,000 individual taxpayers (not clustered)
Observations
10,000 individual taxpayers
Arms
5,000 controls; 2,500 treated by receiving "Message A"; 2,500 "Message B"
Primary outcomes
Declared capital income in the income declaration for 2021

Cited by

No citation data available for this registration yet.

Keywords