Do Elections Shape People Views and Attitudes? Evidence from the 2022 Brazilian Presidential Elections

Last registered on September 22, 2023

View Trial History

Pre-Trial

Trial Information

General Information

Title

Do Elections Shape People Views and Attitudes? Evidence from the 2022 Brazilian Presidential Elections

RCT ID

AEARCTR-0010160

Initial registration date

November 02, 2022

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

November 02, 2022, 5:06 PM EDT

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

Last updated

September 22, 2023, 5:22 PM EDT

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

Locations

Country

Brazil

Region

Primary Investigator

Name

Matteo Francesco Ferroni

Affiliation

Boston University

Contact Primary Investigator

Other Primary Investigator(s)

Additional Trial Information

Status

On going

Start date

2022-09-27

End date

2023-10-31

Keywords

Behavior, Electoral, Welfare

Additional Keywords

Polarization, Political campaign, Fake news, Preferences for redistribution, Memory

JEL code(s)

B55, D31, D63, D72, D73, D84, D91, H11

Secondary IDs

Prior work

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

Abstract

Focusing on the 2022 Brazilian presidential election, I study how an election result can affect people's views, attitudes, and preferences. I collect more than 5,000 respondents during a 6-weeks window that goes from before the elections' first-round to after the elections' second-round. The 2022 Brazilian elections are being very polarizing and uncertain. By exploiting respondents' expectations of the election results and recontacting them right after the final result, I will manage to identify how the election's results affect people views, attitudes, and preferences. More precisely, the main outcomes are going to be respondents' perceptions of what people think, their affective polarization, and their trust towards the government are affected, and then how these effects might translate in a shift in policy preferences, distinguishing between economic and moral policies. I also elicit respondents' exposure and belief of fake news to then explore how this might affect polarization levels. In an additional wave conducted one year after the election, I conduct an experiment to identify the effect of providing good or bad news about the economy on the expectations about the country and attitudes towards violence and democracy.

External Link(s)

Registration Citation

Citation

Ferroni, Matteo Francesco. 2023. "Do Elections Shape People Views and Attitudes? Evidence from the 2022 Brazilian Presidential Elections." AEA RCT Registry. September 22. https://doi.org/10.1257/rct.10160-2.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

In Wave 4 and in the recontacts of Wave 3, the order of some questions are going to be randomized.

In Wave 6, I randomly assign respondents to one of two treatments (plus control group) with information on the Brazilian economy.
Respondents are also randomized between two different orders of the questions.

Intervention Start Date

2022-10-02

Intervention End Date

2023-09-30

Primary Outcomes

Primary Outcomes (end points)

Perceptions of what people think, affective polarization, trust towards the government, preferences for economic policies, preferences for moral policies.

Attitudes towards violence and democracy, expectations about Brazil.

Primary Outcomes (explanation)

Perceptions of what people think: share of people in your city supporting a given candidate, share of Brazilians who are a strong supporter of a given candidate.
Affective polarization: difference in feeling thermometer for both candidates and both candidates' supporters, how comfortable being friend of a supporter of a given candidate, how comfortable having son/daughter marrying supporter of a given candidate.
Preferences for economic policies: index built aggregating support for government intervention, higher taxes on the rich, increase expenditure on various welfare programs.
Preferences for moral policies: index built aggregating support for same-sex marriage, same-sex parents adoption, abortion, gun ownership, drug liberalization, environmental policies.

Attitudes towards violence: questions on whether violence is acceptable to express disagreement with the government, likelihood of participating in violent protests, support for January 8 attacks.
Attitudes towards democracy: questions on preferred form of government, support for a military coup.
Expectations about Brazil: ask about the economy, inflation, employment, inequality, criminality, political polarization.

Secondary Outcomes

Secondary Outcomes (end points)

Support for democracy vs strong leader or military rule.
Expectations of economic and social indicators.
Emotions.

Secondary Outcomes (explanation)

Experimental Design

I will exploit the election result of the second-round Brazilian elections as a treatment on those respondents that I will contact before and after the election result. To ensure variability of the treatment, I will exploit various characteristics of the respondents. Previous waves are going to allow me to test for pre-trends.

Within the survey, I also design a memory treatment where the order of certain questions is randomized. I will be able to test whether remembering what answered in the previous week is going to affect your current beliefs.

In Wave 6, I randomly assign respondents to two different treatments to see the effect on news about the economy on their expectations and attitudes towards violence and democracy. I also randomize the order of this latter two blocks to test whether just thinking about the future of the country makes people change views about democracy and violence.

Experimental Design Details

To ensure variability of the election result treatment, I will interact the post-election dummy with various characteristics and beliefs of the respondents. I will mainly focus on the following: political affiliation, strength of support for a given candidate, expectations of the election result (which will measure their level of surprise), memory adjustment of their previous expectations, support for a given candidate in their geographical area.

In Wave 6, to better identify the treatment effect, I will interact it with some perceptions elicited before the treatment.
To test for heterogeneous treatment effect, I will interact both the treatment and the ordering dummies with various characteristics of the respondent, mainly who they voted for in the elections and the strength of the support for that candidate.

Randomization Method

The randomization of the question order is automatically done by Qualtrics.

Randomization Unit

Individual respondents

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

1,200 respondents for Wave 1
500 respondents for Wave 2
2,000 respondents for Wave 3
Around 700 recontacts from Wave 3
1,000 respondents for Wave 4

3,000 respondents for Wave 6

Sample size: planned number of observations

1,200 respondents for Wave 1 500 respondents for Wave 2 2,000 respondents for Wave 3 Around 700 recontacts from Wave 3 1,000 respondents for Wave 4 3,000 respondents for Wave 6

Sample size (or number of clusters) by treatment arms

Around 350 controls and around 350 treated among recontacts of Wave 3
500 controls and 500 treated among respondents of Wave 4

For Wave 6:
500 controls, questions on expectations about Brazil first.
500 controls, questions on democracy and violence first.
500 treated with positive news about the economy, questions on expectations about Brazil first.
500 treated with positive news about the economy, questions on democracy and violence first.
500 treated with negative news about the economy, questions on expectations about Brazil first.
500 treated with negative news about the economy, questions on democracy and violence first.

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

Supporting Documents and Materials

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

Request Information

IRB