The incidence of VAT evasion: experimental evidence from Italy

Last registered on March 20, 2024

View Trial History

Pre-Trial

Trial Information

General Information

Title

The incidence of VAT evasion: experimental evidence from Italy

RCT ID

AEARCTR-0012415

Initial registration date

January 25, 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

January 31, 2024, 11:43 AM EST

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

Last updated

March 20, 2024, 11:09 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

Italy

Region

Rome

Primary Investigator

Name

Leonardo Maria Giuffrida

Affiliation

ZEW Mannheim

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Giacomo Brusco

PI Affiliation

University of Tuebingen

PI Name

Albrecht Bohne

PI Affiliation

ZEW Mannheim

Additional Trial Information

Status

In development

Start date

2023-02-17

End date

2024-04-30

Keywords

Behavior, Crime, Violence, & Conflict, Firms & Productivity, Other

Additional Keywords

Tax Evasion, Value Added Tax, Tax incidence.

JEL code(s)

H22, H26.

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

Who benefits from VAT tax evasion, and how are the rents distributed between sellers and buyers? How much of the “saved” taxes does the seller pass on to the consumer, and what does this depend on? When the tax of a commodity or service is evaded, it is unclear who reaps the benefits. While the general distributional effect of taxes (incidence) has been frequently studied in the literature, little is known about how the rents of evasion are split between consumers and producers. While some theoretical general-equilibrium models would predict factor prices adjust so that firm owners do not benefit from evasion in equilibrium, we have very limited empirical evidence, which so far has only been based on online surveys. This project aims to quantify the incidence of evasion by conducting an innovative survey-based field experiment in Italy. Thanks to a randomized treatment design, we elicit average price differentials by the method of payment in order to back out what portion of evaded taxes are passed on to consumers.

External Link(s)

Registration Citation

Citation

Bohne, Albrecht, Giacomo Brusco and Leonardo Maria Giuffrida. 2024. "The incidence of VAT evasion: experimental evidence from Italy." AEA RCT Registry. March 20. https://doi.org/10.1257/rct.12415-1.1

Sponsors & Partners

Interventions

Intervention(s)

Specifics of the intervention will be pre-registered as hidden and released upon completion of the study.

Intervention Start Date

2024-01-29

Intervention End Date

2024-04-01

Primary Outcomes

Primary Outcomes (end points)

Price quotes.

Primary Outcomes (explanation)

Quoted prices will be directly elicited from in our survey.

Secondary Outcomes

Secondary Outcomes (end points)

Issuance of receipts.

Secondary Outcomes (explanation)

For a randomly chosen subsample, enumerators will conduct additional visits and record whether a receipt was issued on a small transaction.

Experimental Design

Specifics of the experimental design will be pre-registered as hidden and released upon completion of the study.

Experimental Design Details

The protocol followed by enumerators is the following:

_________________________________________________________________________________________
-Enumerator enters the restaurant

-Enumerator: "Good morning! I am organizing a dinner for a friend, can I talk to the manager?"

-Waiter reports to manager or says you can ask him/her directly

-Manager: [Greets customer].

-E: "Hello, I would like to arrange a dinner in a couple of weeks - we were thinking [date]. There should be [8 or 12] of us. I'm asking a couple of restaurants to see how much it might cost us..."
---Control visits: "...I was thinking an appetizer and an entree -- we don't have any vegetarians or allergies. Could you tell me how much is per person? I would like to pay cash if possible. "
---Treatment visits : "...I was thinking an appetizer and a first course -- we have no vegetarians or allergies. Could you tell me how much is per person? And by the way, can I pay by debit card?"

-M: [discusses menu details, makes a price].

-E: Thank you very much! Could I also see the table you will give us?

-Manager shows the venue to the enumerator.

-E: Thank you again. I think I will ask for a couple more quotes, but I should contact you soon to confirm our plans. How can I contact you?

-M: [provides phone/email details].

-E: Very well, have a nice day!

Enumerator leaves the restaurant.
_________________________________________________________________________________________

At the end of the survey, enumerators will record the outcome of the survey as well as details regarding the restaurant. The full list of questions they will answer can be seen in the attached document, "enumerator questionnaire".
_________________________________________________________________________________________

Our empirical strategy will be relatively straightforward, thanks to the experimental nature of our setting. We will, however, conduct two visits per restaurant, one in the control group and one in the treatment group, with the main purpose of improving our statistical power, as we expect large variation in prices across restaurants even for comparable meal plans. In order to make the re-sampling of firms less suspicious, we plan to ask for slightly different meal plans, by introducing small changes the courses enumerators ask for and the size of the dinner party. This, in turn, allows us to include meal-plan fixed effects. As is standard in similar settings, we also plan to include enumerator fixed-effects.

Randomization Method

Each establishment will receive two interviews by two distinct enumerators, one in the treatment group and one in the control group. The order of the two visits will be randomized ex-ante through a pseudo-random number generator.

Randomization Unit

Randomization will occur at the establishment level. Each establishment will be visited by our enumerators twice, once for a control visit and once for a treatment visit. The order of the visits will be randomized for each establishment. In a third visit, a distinct third enumerator will conduct a small transaction with each establishment and record whether a legally valid receipt is issued.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

Each establishment will receive two visits. Randomization will occur at the establishment level (no clustering) and will determine in which order the two visits will occur.

Sample size: planned number of observations

500 establishments will be surveyed two times each, totalling 1000 observations.

Sample size (or number of clusters) by treatment arms

1000 establishment visits. Each of 500 establishments will be visited twice.

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

According to our calculations, MDE for price is 2.39 Euros. This MDE is based on variance in the price quotes we gathered during our pilot survey. With a baseline average price quote of 31.89 euros per person and an applicable VAT rate of 10%, this means we would be able to identify full pass-through of evasion to consumer prices. In our analysis plan we also give details on how to perform this calculation again conditional on restaurant fixed effects. Restaurant fixed effects are bound to substantially decrease the conditional variance of our outcome and thereby increase our statistical power. Unfortunately, we did not re-sample restaurants in our pilot survey which means we cannot perform this calculation currently. We provide more details in our attached analysis plan.

Supporting Documents and Materials

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

Request Information

IRB