Voter Responses to Emotional Appeals

How to cite

Labonne, Julien; Trebbi, Francesco; Cruz, Cesi. 2019. "Voter Responses to Emotional Appeals." AEA RCT Registry. AEARCTR-0004137. https://doi.org/10.1257/rct.4137-1.0

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

Julien Labonne* , Francesco Trebbi , Cesi Cruz

* Primary investigator

Registry ID
AEARCTR-0004137
DOI
https://doi.org/10.1257/rct.4137-1.0
First registered
Last updated
Status
In development
Trial dates
2019-04-09 – 2019-06-30
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/4137

Abstract

This project tests the effect of microtargeting of political messaging in election campaigns in the Philippines. It explores how messaging affects voter preferences and beliefs.

The core of the project - conducted in collaboration with a political party - relies on conducting an informational treatment in the context of the 2019 Senatorial elections in the province of Laguna in the Philippines. Voters in randomly selected villages will receive scripted door-to-door visits by party volunteers. Voter in a randomly selected subset of those villages will further receive messages emphasising positive emotions.

The project seeks to answer questions such as:

1. What are the effects of door-to-door visits of party volunteers on voters beliefs and vote choices?
2. What the effects of positive campaign messages on voters beliefs and vote choices?

We will also estimate key parameters of a structural model of vote choice based on elicitation of voters beliefs and heterogenous preferences for policy.

Experimental design

The sample will be divided into three groups:
• Treatment 1: Receive door-to-door visits from party volunteers
• Treatment 2: Receive door-to-door visits from party volunteers. The message will be the same as in treatment 1 but the interaction between citizens and volunteers will be structured to emphasize positive emotions
• Control: do not receive any vists.

The door-to-door visits will take place a few days before the elections.

The randomization will be done within municipality.

Design details

Randomization method
The randomization will be done using a Stata code
Randomization unit
Randomisation at the village-level
Clusters
About 200 villages
Observations
10 respondents per village for a total of 2,000 observations
Arms
1/3 (pure control), 1/3 (door-to-door visits) and 1/3 (door-to-door visits and emotional messages)
Primary outcomes
Key outcome variables are: beliefs about candidate, candidate and party valence, vote choice. The survey instruments will also include a module necessary to estimate the structural model. It is designed to elicit voters' beliefs about candidates and parties. We will also collect official vote share at the village-level from the Electoral Comission (COMELEC)

Cited by

No citation data available for this registration yet.

Keywords