Cleverness of loopholes

Last registered on October 18, 2024

View Trial History

Pre-Trial

Trial Information

General Information

Title

Cleverness of loopholes

RCT ID

AEARCTR-0013772

Initial registration date

October 08, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

October 18, 2024, 4:38 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

United States of America

Region

Primary Investigator

Name

Renke Schmacker

Affiliation

WZB & DIW Berlin

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Tobias König

PI Affiliation

Linnaeus University

Contact Investigator

PI Name

Lydia Mechtenberg

PI Affiliation

University of Hamburg

Contact Investigator

Additional Trial Information

Status

In development

Start date

2024-10-08

End date

2024-11-30

Keywords

Behavior

Additional Keywords

JEL code(s)

Secondary IDs

Prior work

This trial is based on or builds upon one or more prior RCTs.

Abstract

In this experiment, we examine whether the acceptability of exploiting a tax loophole depends on how “clever” it is. The experiment involves two types of participants: taxpayers and spectators. Taxpayers are matched into groups of three and are supposed to pay a tax designed to equalize their incomes. However, they have the option to engage in a calculation task to reduce their tax rate at the expense of other participants. After the calculation task they have to report the number of correct calculations. Depending on the treatment, the calculation task contains different loopholes, which allows taxpayers to strongly lower their tax rate. Spectators judge whether exploiting the respective loophole is appropriate or not. They can decide whether to sanction workers who exploited the loophole.

This experiment relates to experiment AEARCTR-0011828 (https://www.socialscienceregistry.org/trials/11828). In contrast, we manipulate the cleverness of the loophole in this experiment.

External Link(s)

Registration Citation

Citation

König, Tobias, Lydia Mechtenberg and Renke Schmacker. 2024. "Cleverness of loopholes." AEA RCT Registry. October 18. https://doi.org/10.1257/rct.13772-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

2 treatments: Baseline, Non-Clever

In Baseline, the loophole is clever and difficult to detect. In Non-Clever, the loophole is not clever and easy to detect.

Intervention (Hidden)

Intervention Start Date

2024-10-08

Intervention End Date

2024-11-30

Primary Outcomes

Primary Outcomes (end points)

Share of spectators that sanction loophole exploiters: Spectators are matched with a group of taxpayers from a previous session and can decide whether loophole exploiters should be sanctioned.

Appropriateness of using the loophole: Spectators’ first-order and second-order beliefs of the appropriateness of exploiting the loophole.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

As manipulation checks, we estimate treatment effects on agreement with the statements relating to the cleverness of using the loophole, whether they successfully detected the loophole and their guess of how many others detected it.

We will investigate mechanisms explaining the decision to close or sanction the loophole using the factors above, as well as the belief of how many subjects exploit the loophole and by how much.

Exploratively, we will investigate whether behavior in the experiment is correlated with political views and voting behavior in previous presidential elections. We will also investigate the relationship with their beliefs regarding the appropriateness of loopholes, preferences for redistribution, and perceptions of the efficiency of public spending.

Exploratively, we will also investigate the behavior of the taxpayers. Here, we focus on their choice to use the loophole and on whether they consider using the loophole appropriate.

Secondary Outcomes (explanation)

Experimental Design

Spectator-worker design with randomized interventions

Experimental Design Details

Participants are matched into groups of three (one green workers and two orange workers). Green workers receive a higher gross income than orange workers. Green workers are supposed to pay a tax of 60% and orange workers of 20%, such that post-tax incomes are equal. However, green workers can engage in a calculation task to adjust their tax rate. For each correct calculation, they reduce their tax rate by 3 percentage points. Depending on the treatment, there are different loopholes in the calculation task, which allow workers to decrease their tax rate strongly without making any calculations.

Spectators are shown the instructions of the workers and they can try out the task as well. Afterward, they are asked whether they would like to sanction workers who exploited the loophole. Moreover, they are asked how appropriate they consider the exploitation of the loophole.

Worker sessions for the ex-post decision are conducted prior to the spectator experiments. The choices made by randomly selected spectators are then implemented in the first round of these worker sessions.
In a post-experimental questionnaire, we elicit the motives for spectator decisions, general political attitudes, and background characteristics.

For the experimental interventions, we test the following hypotheses (relative to the baseline treatment):
H1a: In the Non-clever treatment, individuals are more likely to sanction loophole exploiters.
H1b: The Non-clever treatment reduces the appropriateness of exploiting the loophole.

We will estimate heterogeneous treatment effects by political affiliation, voting behavior in previous presidential elections, income, education, gender, and age.

We are interested in heterogeneous treatment effects by political affiliation, however, Prolific has a substantially higher share of Democratic participants. Therefore, we will use quotas to target the vote share of Donald Trump in the presidential election 2020.

We will do robustness checks by excluding participants who answered the comprehension questions incorrectly on the first try.

Randomization Method

Computerized randomization using Qualtrics. We use the option “evenly present elements.”

Randomization Unit

Individual

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

Target sample: 230 to 280 spectators per treatment conditional on the ability to fill our quotas (see experimental design).

Sample size: planned number of observations

See above

Sample size (or number of clusters) by treatment arms

See above

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Assuming a baseline share to sanction the loophole exploiter of 0.72, the MDE with N=280 per treatment is 0.10 at power=0.8 and alpha=0.05. If we end up with N=230, the MDE is 0.11. Assuming a baseline appropriateness rating of -0.37 (sd 0.7), the MDE is 0.166 at n=280 per treatment, power=0.8 and alpha=0.05. If we end up with N=230, the MDE is 0.183.

Supporting Documents and Materials

IRB