An economic experiment: Enhancing the Capital gains tax on property compliance.

Last registered on April 08, 2021

Pre-Trial

Trial Information

General Information

Title
An economic experiment: Enhancing the Capital gains tax on property compliance.
RCT ID
AEARCTR-0007464
Initial registration date
April 08, 2021

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
April 08, 2021, 6:18 AM EDT

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

Locations

Region

Primary Investigator

Affiliation
City, University of London

Other Primary Investigator(s)

PI Affiliation
Tax Authority Slovak Republic
PI Affiliation
Prague University of Economics and Business
PI Affiliation
Institute for Fiscal Policy
PI Affiliation
Prague University of Economics and Business

Additional Trial Information

Status
On going
Start date
2021-02-23
End date
2021-09-30
Secondary IDs
Abstract
For many countries and public institutions, efficient tax collection is essential for their existence. Even though their tax codes seem always complete and precise in specifying terms and procedures of tax collections, in reality there often exists space for improvements. In Slovakia, if a property is sold within five years from its acquisition, the Capital gains tax on property (further just “Tax”) should be paid. The Tax should be self-reported by taxpayers in an annual tax report for the year in which the property was sold, which opens space for evasion.
In this project, we cooperate with the Financial Administration Slovak Republic and aim at increasing the Tax collections by sending informative letters to the potential evaders that remind them of their obligations. We test several text versions of the letter that may improve the rate of response, each appealing to certain personal motivation. To shed light on what type of information motivates people the most, five variations of the letter were created on top of a baseline letter and were combined orthogonally in a 2x4 manner.
External Link(s)

Registration Citation

Citation
Celik Katreniak, Dagmara et al. 2021. "An economic experiment: Enhancing the Capital gains tax on property compliance.." AEA RCT Registry. April 08. https://doi.org/10.1257/rct.7464-1.0
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
In this project, we cooperate with Financial Administration Slovak Republic (TASR) and aim at increasing the Tax collections by sending informative letters to the potential evaders that remind them of their obligations. We test several text versions of the letter that may improve the rate of response, each appealing to certain personal motivation. To shed light on what type of information motivates people the most, five variations of the letter were created on top of a baseline letter and were combined orthogonally in a 2x4 manner.
Intervention Start Date
2021-03-23
Intervention End Date
2021-09-30

Primary Outcomes

Primary Outcomes (end points)
On the extensive margin, the primary outcome variable is whether individuals or joint owners decide to pay the Capital gains tax. On the intensive margin, the primary variable of interest is the amount of the tax paid from the capital gains. We will test for the existence of gender differences in the measures.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
The study includes four closely related experiments that differ with respect to (i) repayment period and (ii) the nature of the subject pool. First, we aim at the taxpayers who sold their property more than one year ago, and thus the period for reporting has already expired (the “ex-post” experiment), and at the taxpayers who sold their property in the last year and are supposed to report it in the current standard reporting period (the “ex-ante” experiment). The second dimension consists of whether the recipient is an individual or spouses selling a jointly owned property.
The treatments T1-T4 consist of presenting different information in one extra paragraph in the middle of the front page of the cover letter. The first treatment (T1) provides information about the probability of being audited, as it may deter tax evasion (Bérgolo et al., 2017); the second (T2) describes the structure of fees for late payment, which can also improve tax compliance (Castro & Scartascini, 2015; De Neve et al., 2019); the third (T3) contains moral suasion inspired by Bott et al. (2019); and the last one (T4) contains a sentence reframing the act of non-payment as a commission rather than omission (Hallsworth et al., 2015). Finally, in an orthogonal treatment (B) we provide the recipients with an extra leaflet that contains a flow-chart aiming to simplify the decision-making situation and help them find out whether they are exempt or not. The selection of the treatments reflects the interests of the TASR. Two control groups are introduced - first, a control group whose group members receive the baseline letter, and second, no-letter control group, whose members do not receive any letter at all.
Experimental Design Details
The study includes four closely related experiments that differ with respect to (i) repayment period and (ii) the nature of the subject pool. First, we aim at the taxpayers who sold their property more than one year ago, and thus the period for reporting has already expired (the “ex-post” experiment), and at the taxpayers who sold their property in the last year and are supposed to report it in the current standard reporting period (the “ex-ante” experiment). The main difference important for us is that we expect the ex-post control group basically not to declare the tax unless they receive a letter in our project, but in the ex-ante control group we expect a rate of 30% response independent of our letter, based on data from recent years as provided by TASR. The second dimension consists of whether the recipient is an individual or spouses selling a jointly owned property. In the case of spouses, we send a letter either to the female or the male owner. We decided to study them separately from the individual owners because it allows us to observe the effect of sending a letter to either husband or wife on the reaction of the whole couple, be it the recipient, their spouse, or both.

Treatments

The baseline letter (T0) is formulated as clearly as possible, stating the reason for sending it to a recipient which is a change of ownership of a property in the land registry and the absence of a corresponding tax declaration. It contains all important information on how, when, and where to declare and pay the tax in the simplest way possible, at the same time meeting the conditions of full and proper information from tax authorities. Additionally, it includes a list and simple definitions of possible exceptions to the capital gains tax on property on the second page.
We designed four text treatments (T1-T4) where we add an extra paragraph in the middle of the baseline letter (T0), and one orthogonal treatment (A vs B) where we add an additional navigation leaflet to the letter which contains a flow-chart of possible situations of taxpayers, thereby aiming at further simplification of the decision whether or not to pay the Tax.
● Treatment T1 includes one paragraph with a reminder of tax authorities having the legal right to conduct an audit and inspect all types of income and the fact that by not paying the capital gains tax on property the chance of an audit is being increased. In the case of self-reported income, threat-of-audit letters are expected to have a positive deterrence effect on tax evasion, in line with the extended standard model of (rational) tax evasion (Kleven et al., 2011).
● Treatment T2 adds a paragraph with information on fees for late payment and a late tax declaration and encouraging information about these fees being significantly lowered if the income is declared before the stated deadline. This message emphasizes the possibility to lower costs if an immediate action takes place, which is known to be an effective mechanism in tax correspondence (De Neve et al., 2020).
● Treatment T3 adds a paragraph about taxes being an important part of the duties of citizens and necessary to offer free education, healthcare, and social benefits. It also states that not paying the taxes is unfair towards all paying citizens. By appealing to injunctive social norms and highlighting that the recipients' behavior might be considered as a deviation from moral standards, we aim to raise explicit moral costs of not paying the tax. Moreover, by listing the public goods we encourage taxpayers to derive utility from a public service provision (Hallsworth et al., 2017).
● Treatment T4 adds an extra sentence to the baseline letter framing non-payment of the tax being viewed by tax authorities clearly as commission instead of omission. This reframing addresses the omission strategy of taxpayers by changing the perceived nature of the action, resulting in the expectation of greater punishment for evasion of paying taxes (Hallsworth et al., 2015).



Randomization Method
Randomization is done by a computer.
Randomization Unit
In the two experiments with individual owners, we randomized the subsample at the individual level. In the other two experiments with joint owners we used households as our clusters and we randomized the subsample at household level.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
We have 1,365 spouses in the ex-post experiment who sold a property more than one year ago and 686 spouses in the ex-ante experiment who sold a property in the last year.
Sample size: planned number of observations
The target sample of the ex-post experiment is 8,254 individuals who sold the property more than one year ago. The sample of individuals who sold a property less than one year ago in the ex-ante experiment includes 3,969 taxpayers. The sample of spouses in the ex-post experiment which sold a property more than one year ago includes 1,365 couples and the sample of spouses in the ex-ante experiment who sold a property in the last year includes 686 couples. In each of these four experiments, we include a separate control group that receives no letter.
Sample size (or number of clusters) by treatment arms
The natural goal of the TASR is to maximize the chance and value of the (re)payment of the Tax and therefore their preference was to minimize the size of the control group which does not get any letter. The size of the control (no-letter) group is different across the four scenarios: in the ex-post experiment it is 5% of individuals and 15% of households and in the ex-ante experiment 10% of individuals and 20% of households. The proportion of people selected for the control group depended on a) the expected proportion of control-group people who would pay the capital gains tax once the letters are sent (we expect no or a minimal number of individuals/households (spouses) in the ex-post experiment to pay without any reminder, and approximately 30% of the individuals/spouses in the ex-ante experiment), and on b) the minimum treatment effect size we would be able to detect given the resulting treatment-control proportions.
The overall sample size was determined as the total number of taxpayers identified by the TASR as potential evaders. The final number of addresses per experiment and treatment cells are shown in Table 1. The total sample size per experiment, the control/treatment proportions, and the minimum detectable treatment effects based on power calculations are presented in Table 2. Within the treatment group, people had equal chances to be selected into any of the treatment cells. We stratified all treatment cells using multiple observable characteristics (gender, age below/above 45, disposable income below/above 25k€, tax base higher/lower 10k€, employment type, and subjects’ residency in case of experiments with individuals; and gender, age group, tax base, and disposable income in case of experiments with spouses) to strengthen the randomization balance.

Table 1: Randomization and the list of hypotheses (see the attached Analysis plan)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
The number of hypotheses we are interested in differs across the four experiments. In the power calculations, we adjust the significance levels using the Bonferroni correction to minimize the problem of false-positive discovery due to multiple comparisons. The minimum detectable treatment effect describes the minimum difference between the expected value of our dependent variable in the compared treatment groups that would be detected using 80% power and adjusted significance levels when the initial α=0.05. First, we evaluate the extensive margin – the percentage of recipients of letters to fill in the tax declaration in the upcoming period – and second, the intensive margin – the size of the paid tax. The minimum detectable effects on the extensive margin calculated for each experiment are shown in Table 2, when we assume a two-sided chi-2 test of proportions in two independent samples. The range of minimum detectable treatment effects reflects the expected response rate either of the control-group people (in the evaluation of the overall effect of receiving a letter), the treated people without the navigation leaflet (in the evaluation of the orthogonal treatment effect of the navigation leaflet), the treated people that received the baseline letter (in the evaluation of the text treatment effects of 4 types of letters) or the response of male recipients of a letter (in the evaluation of the gender differences in the experiments with joint owners). Table 2: Sample sizes, the control-treatment proportions, and the minimum detectable treatment effects (see the attached Analysis plan)
IRB

Institutional Review Boards (IRBs)

IRB Name
IRB Approval Date
IRB Approval Number
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials