The Interaction of Communication, Social Preferences, and Inequality: An experimental study

How to cite

Zhazhin, Kirill; Vargas, Kristian Lopez; Putterman, Louis. 2025. "The Interaction of Communication, Social Preferences, and Inequality: An experimental study." AEA RCT Registry. AEARCTR-0014317. https://doi.org/10.1257/rct.14317-1.0

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

Kirill Zhazhin* , Kristian Lopez Vargas , Louis Putterman

* Primary investigator

Registry ID
AEARCTR-0014317
DOI
https://doi.org/10.1257/rct.14317-1.0
First registered
Last updated
Status
In development
Trial dates
2025-07-08 – 2025-08-10
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/14317

Abstract

This experiments studies the impact of communication on voter behavior in a lab setting. First, participants complete a series of tasks that can determine their starting position later. Then, they interact in a fixed group by sending messages to each other in an open-forum. Last, they vote for a tax that serves a redistributive and public-good creation purposes. Communication and voting stages are repeated for several rounds. We study the individual choice of messages, voting behavior, and distributional outcomes relative to the theoretical predictions. We wary access to communication and the initial income determination method across treatments.

Intervention

We use a 3×2 design that varies access to communication and the method of initial income determination. We consider a society of size N=9 individuals, who complete real-effort tasks and then interact with each other by communicating and then voting. Communication access is a between-group treatment, with three conditions: (i) free communication with no message cost, (ii) affordable communication with costly options to enhance the message, (iii) no communication. The initial income determination method is a within-group treatment. It varies across four rounds of repeated communication and voting stages, and can either be random (R) or effort-based (E), i.e., dependent on the relative performance of participants in the tasks that precede communication. We use a crossover approach with RREE and EERR options across four round in each group. Therefore, each treatment is a combination of one of the (i)-(iii) and RREE/EERR conditions. The payoff of participants depends on their initial income, their message and its cost, and the voting outcome.

Experimental design

In the experiment, each participant individually completes four real-effort tasks, then gets assigned to a group of size N=9. Then, participants interact inside each group in four rounds, while groups remain unchanged. In each round, participants get informed about their income and then vote on a tax that could either result in smaller or larger final payoff. Depending on the treatment, a communication stage precedes voting. The participants are allowed to send a message to others in an open-forum message, which can also be enhanced for small but noticeable price.

Participants earn experimental points based on their performance in the real-effort tasks, communication decision, and voted tax. At the end of each round, they receive feedback about their own choices, group outcome, and payoffs. Their final earnings are determined based on the post-tax outcome from a randomly selected round.

During the voting stage in each round, participants are also asked to predict the voting outcome. At the end of the experiment, participants fill out a survey about demographics, risk-, redistribution- and social- preferences.

The experiment is conducted online with participants recruited on Prolific, using o-Tree as software.

Design details

Randomization method
By the random order of arrival of participants to participate in the experiment on Prolific and "random" Python module.
Randomization unit
Group level
Clusters
Number of groups: 32
Observations
Number of observations: 32x9=288 participants
Arms
Number of groups per each treatment arm: 6 in Free Communication & RREE, 6 in Free Communication & EERR, 6 in Affordable Communication & RREE, 6 in Affordable Communication & EERR 4 in No Communication & RREE, 4 in No Communication & EERR
Primary outcomes
Messages of individuals Votes of individuals Payoffs of individuals
Secondary outcomes
Message content Group outcome efficiency Distributional outcomes

Cited by

No citation data available for this registration yet.

Keywords