Post-Audit Tax Compliance: Endogenous Audits and Audit Effectiveness

How to cite

Mohr, Aline. 2023. "Post-Audit Tax Compliance: Endogenous Audits and Audit Effectiveness." AEA RCT Registry. AEARCTR-0012026. https://doi.org/10.1257/rct.12026-1.1

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

Aline Mohr*

* Primary investigator

Registry ID
AEARCTR-0012026
DOI
https://doi.org/10.1257/rct.12026-1.1
First registered
Last updated
Status
In development
Trial dates
2023-10-10 – 2023-10-15
Countries
Private
JEL codes
H2, H3
Registry record
http://www.socialscienceregistry.org/trials/12026

Abstract

As tax evasion remains a large problem in many countries with immediate consequences for the economy and living standards, tax authorities make use of audits to deter non-compliance.
Thus, it is crucial to examine the determinants of the post-audit effect on the tax compliance behavior of audited taxpayers. In this light, I investigate experimentally the causal effect of audit effectiveness on post-audit tax compliance in presence of an endogenous audit selection rule. Furthermore, I study whether this effect differs from the one observed in a setting with an exogenous audit selection rule. To do so, I will implement a randomized control trial within a tax evasion game in the form of an online experiment using a Census-Matched sample of at least 399 US citizens. Based on their order of entry in the study, participants are randomly assigned to one of the three treatments regarding audit effectiveness, low (T1), medium (T2) or high (T3). Additionally, participants experience both, a setting with an exogenous and an endogenous audit selection rule. The outcomes of interest are subjects’ (post-audit) tax compliance rates. I hypothesize that the direction of the post-audit effect depends on audit effectiveness, i.e., effective audits increase, and ineffective audits decrease post-audit tax compliance rates, in both settings. I expect that, comparing the endogenous to the exogenous audit setting, overall compliance is higher, while the effect of effective audits is stronger and the one of ineffective audits is weaker.

Experimental design

I investigate the causal effect of audit effectiveness on post-audit tax compliance in presence of an endogenous audit selection rule. Additionally, I study whether this causal effect differs from the one observed in a setting with an exogenous audit selection rule. Therefore, I implement an online experiment simulating voluntary income tax reporting. Hence, a Census-Matched sample of at least 399 US citizens will be recruited using CloudResearch’s Connect.

This online experiment consists of two parts with each multiple decision rounds. In each decision round participants are endowed with a certain income and need to decide on the amount of income reported to the tax agency. Thereafter, participants face the risk of being audited. In the first part, participants are confronted with an exogenous audit selection rule, while the audit selection rule becomes endogenous in the second part.

In the beginning of the experiment, participants are randomly assigned to one of the three treatment groups. Participants in the first treatment group experience audits of low effectiveness, whereas participants in the second and third treatment group face audits of medium and high effectiveness, respectively. The main outcomes of interest in this study are participants’ (post-audit) tax compliance rates. Compliance behavior is identified from the decisions taken by the participants in the multiple decision rounds and allows to determine the researched causal effects by adding respective dummy variables to the regression.

Design details

Randomization method
The random assignment of the participants to the three treatments is based upon their order of entry in the study. The order of entry in the study is random, because subjects are free to choose when to participate in the online experiment after its implementation. The data for each treatment group is collected in a different experimental session. In each decision round and for each participant the outcome of a decision round, i.e. whether or not an audit takes place, is based on randomization done by a computer. This randomization takes the respective audit probabilities into account.
Randomization unit
Subjects participating in the online experiment are randomly assigned to the three treatments upon their order of entry in the study. Furthermore, individual randomization takes place in each decision round in order to determine whether or not an audit takes place.
Clusters
I plan to recruit a Census-Matched sample of at least 399 US citizens.
Observations
The planned minimum number of observed income reporting decisions amounts to 3990.
Arms
The sample of at least 399 US citizens will be assigned to three treatments. Thus, at least 133 participants will be assigned to each treatment. This leads to at least 665 observations in each part and treatment of the experiment and amounts to a total of at least 1330 observations per treatment arm.
Primary outcomes
I consider (post-audit) tax compliance rates (declared income over total income), obtained from the online experiment, as primary outcomes of the study. Several dummy variables and their interaction terms are added to the regression in order to identify the post-audit effect, the effect of audit effectiveness on post-audit tax compliance and the difference between a setting with an exogenous and an endogenous audit selection rule.
Secondary outcomes
Additionally, I investigate in both, the setting with an exogenous and an endogenous audit selection rule, whether a taxpayer's prior reporting behavior affects the effect of audit effectiveness on post-audit tax compliance by limiting the regression sample to relatively compliant and noncompliant taxpayers respectively.

Cited by

No citation data available for this registration yet.

Keywords