Do Think Twice, it´s All Right: Effects and Mechanisms of Tax Enforcement Policies

How to cite

Kotsadam, Andreas. 2019. "Do Think Twice, it´s All Right: Effects and Mechanisms of Tax Enforcement Policies." AEA RCT Registry. AEARCTR-0004817. https://doi.org/10.1257/rct.4817-1.1

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

Andreas Kotsadam*

* Primary investigator

Registry ID
AEARCTR-0004817
DOI
https://doi.org/10.1257/rct.4817-1.1
First registered
Last updated
Status
On going
Trial dates
2018-05-15 – 2022-08-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/4817

Abstract

We compare the tax compliance effects of two different tax enforcement policies, a relatively costly desk/correspondence audit and a cheap information treatment with a letter asking filers with “risky” filing behavior to take a second look at their tax returns. To assess the total effect of these enforcement policies, we need to take account of both their immediate effects and their long term effects on future filing behaviour. Audits will detect and correct non-compliance on the spot, but may also change subsequent filing behavior. A letter will not disclose non-compliance, but (some) taxpayers may adjust their filings. A letter may also have long term effects on tax compliance. Based on the short run effects of both treatments, which are easy to measure, we describe how we will test the behavioural effects on future filing. We describe the intervention and lay out some important decisions with respect to coding of variables, definitions of samples and the empirical strategy we will apply.

Intervention

We work together with the Norwegian National Tax Authorities to compare the effectiveness of these two enforcement policies; a desk based correspondence audit and a “take a second look” letter.

Experimental design

Randomly assigned audits and letters

Design details

Randomization method
In office by a computer
Randomization unit
Individual
Clusters
12,495 individuals
Observations
12,495 individuals
Arms
12,495 individuals
Primary outcomes
Claimed deductions for tax year 2018 (in NOK)
Secondary outcomes
Partner's claimed deductions for the tax year 2018 (in NOK) and Subjective Detection Risk

Cited by

No citation data available for this registration yet.

Keywords