Dataset for the paper "Does voting on tax fund destination imply a direct democracy effect?" (2021)

Data creators : Nicolas Jacquemet [1] [2], Stéphane Luchini [3], Antoine Malézieux [4]
[1] : Centre d'Économie de la Sorbonne, University Paris 1 Panthéon-Sorbonne (UMR8174)
[2] : Paris School of Economics
[3] : Aix-Marseille School of Economics, Aix-Marseille Université (UMR7316)
[4] : Centre de recherche sur les entreprises (Burgundy School of Business)
Description :
Previous experiments have shown that giving participants the opportunity to select the organization that receives their tax funds tends to increase tax compliance. In this study, we assessed whether this increase in compliance is induced by the sole fact of giving subjects a choice — a “direct democracy effect”. The present dataset is from a new tax evasion game designed to control for self-selection into the chosen policy.
Discipline :
economics (social sciences)

General metadata

Data acquisition date : from Jan 2016 to Jan 2018
Data acquisition methods :
  • Experimental data :
    The design of the experiment aims to provide evidence on the Direct Democracy Effect while controlling for self-selection. The treatment variable is the choice offered to subjects about how collected taxes are used. The structure of the tax evasion game (TEG) is common to all treatments, and proceeds in two steps. First, participants earn their income in a task borrowed from Alm et al. (2008), in which they have to sort digits in ascending order as quickly as possible. The quicker a subject is the more money he earns, according to the following compensation rule (labeled in an Experimental Currency Unit): 150 - (time x13). Second, participants are asked to report the amount of income they earned in the first stage, knowing that it will be taxed at a 35% rate. The amount of money collected through taxes in the experiment is donated to either of two organizations (between which subjects are asked to choose before the declaration stage in all treatments but the baseline): the WWF (World Wide Fund for Nature, www.wwf.fr) or ASPAS (Association pour la Protection des Animaux Sauvages), a French organization for the protection of wildlife.
    Our between-subject treatments implement changes in the way the organization is chosen. This choice adds an additional step in the experiment, which happens just before the income reporting part of the TEG. There is no such step in the Baseline treatment, in which a random draw decides (with equal probability) whether WWF or ASPAS is selected. As an additional control variable, we ask participants of the Baseline which organization they would have chosen if they had the opportunity at the end of the experiment.
    In our main treatment of interest, Vote, participants are randomly and anonymously matched into groups of three participants. After having earned their income, they learn about the tax simulation that is about to come but are asked, before the declaration stage, to vote in favor of the organization (either WWF or ASPAS) that will receive the taxes collected from all members of the group. At the end of the experiment, the organization is selected based on majority voting.
    In the Vote+ treatment, our aim is to reinforce the social effect of the voting stage. To that end, we add cheap talk between participants, in the sense that participants enter a computerized and anonymous group chat with the members of their trio for up to three minutes, before participating to the Vote treatment. Subjects are explicitly told that the aim of this step is for them to collectively decide on the organization they would like to vote for. In both the Vote and the Vote+ treatments, the uncertainty that subjects actually face regarding the selected organization depends on their beliefs about the vote of other subjects.
    The data is made of ten experimental sessions, two for Baseline and Choice and three for Vote and Vote+, each including between 24 and 27 subjects (variations are due to the show-up rate among invited participants). We over-sampled the Vote and Vote+ conditions because the voting groups induce within subjects correlations that lowers the power of the statistical analysis. All sessions took place in the Strasbourg University lab (LEES) from January 2016 to January 2018
Update periodicity : no update
Language : English (eng)
Formats : application/pdf, application/vnd.openxmlformats-officedocument.wordprocessingml.document, application/x-stata-dta
Audience : Research
Publications :
Projects and funders :
Additional information :
Dataset, glossary of all variables for database and procedure of the experiment

Data

3 files

Dataset_DBVoteIRLE.dta

Published : 22/01/2025 09:46 Size : 180.67 kB

Description : Dataset for the paper "Does voting on tax fund destination imply a direct democracy effect?".Stata-backed data file developed by StataCorp LLC (.dta)

Dataset_key_VariablesVoteIRLE.docx

Published : 22/01/2025 09:50 Size : 21.27 kB

Description : Glossary of all variables for Database corresponding to the paper "Does Voting on Tax Fund Destination Imply a Direct Democracy Effect?"

InstructionsVoteChoixVotePlus.pdf

Published : 22/01/2025 09:52 Size : 98.37 kB

Description : Procedure of the experiment

DOI and links

10.25666/DATAUBFC-2025-01-21-02
https://dx.doi.org/doi:10.25666/DATAUBFC-2025-01-21-02
https://search-data.ubfc.fr/FR-13002091000019-2025-01-21-02

Quotation

Nicolas Jacquemet, Stéphane Luchini, Antoine Malézieux (2021): Dataset for the paper "Does voting on tax fund destination imply a direct democracy effect?". dataUBFC. doi:10.25666/DATAUBFC-2025-01-21-02

Record created 21 Jan 2025 by Cécile Schweitzer.
Last modification : 22 Jan 2025.
Local identifier: FR-13002091000019-2025-01-21-02.

dat@UBFC

dat@UBFC is a metadata catalogue for research data produced at UBFC.

République Française
dat@UBFC
Université de Bourgogne, Université de Franche-Comté, UTBM, AgroSup Dijon, ENSMM, BSB, Arts des Metiers