Attitudes toward Campaign Finance Quotas: Evidence from Survey Experiment in Brazil

How to cite

Tsoutsoplidi, Olivia; Cagé, Julia; Lauritzen, Felipe. 2022. "Attitudes toward Campaign Finance Quotas: Evidence from Survey Experiment in Brazil." AEA RCT Registry. AEARCTR-0010366. https://doi.org/10.1257/rct.10366-1.0

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

Olivia Tsoutsoplidi* , Julia Cagé , Felipe Lauritzen

* Primary investigator

Registry ID
AEARCTR-0010366
DOI
https://doi.org/10.1257/rct.10366-1.0
First registered
Last updated
Status
In development
Trial dates
2022-11-07 – 2023-03-01
Countries
Private
JEL codes
C93, D72, J45, J71
Registry record
http://www.socialscienceregistry.org/trials/10366

Abstract

We measure public awareness and support levels toward affirmative action for women in elected office across the Brazilian population. We evaluate the effect of providing citizens with information about recent reforms in the campaign finance system tying the public funding of elections to gender allocation rules (henceforth, “campaign finance quotas”). We compare support for campaign finance quotas against underlying preferences for female representation and disentangle it from attitudes towards public funding. We correlate both with political preferences and voting behavior in the past 2022 general elections. We conduct a survey experiment of 2,000 individuals with physical interviews across 120 towns in Brazil and give them information about the new Electoral Fund, on its size and allocation rules across parties based on the gender of their candidates. We aim first to assess the level of public awareness on these reforms, and then to evaluate their informational effect on citizens’ preferences for this affirmative action policy. The treatment will consist of exposing a randomized set of the respondents each to information about one of three reforms (three different treatments), while the rest are not exposed and used as a control group. We contribute to the literature on gender quotas by documenting empirically whether there is a principle-policy paradox and a backlash effect that would diminish their effectiveness in Brazil.

Intervention

We will conduct an in-person survey experiment of Brazilian citizens and voters on 7-11 November 2022 with a sample size of 2,000 respondents to assess public attitudes towards campaign finance quotas that reserve a proportional amount of public funds to female candidates in Brazil. We will first assess their level of awareness of this policy. Then we will divide our sample into four information treatment groups to disentangle their attitudes toward different aspects of the policy. We will provide them with information about the public fund created in 2017 (“Electoral Fund”) and reforms on its funding allocation rules passed in October 2020 and December 2021. Accordingly, we will have three treated groups, receiving information about each reform, and a fourth control group that receives no information. Then we compare their answers to a uniform set of outcome questions about their support for or against these policies.

Experimental design

Preference for Female Representation in Congress
Q1. In this year’s elections, women were elected in 91 out of 513 (or 18% of) seats in the Chamber of Deputies. In your opinion, is the number of women elected to Congress in this year’s elections more than enough, less than enough or just enough?

1. More than enough
2. Less than enough
3. Enough
99. Don’t know

Baseline Awareness about Electoral Fund
Q2. Are you aware of the existence of an Electoral Fund to finance political parties with public money? (IF YES) And would you say that you are well informed, more or less informed, OR badly informed about the Electoral Fund?
1. Yes, and you are well informed
2. Yes, and you are more or less informed
3. Yes, and you are badly informed
4. Not informed
99. Don’t know

Prior Beliefs about Quotas
Q3. In your opinion, the reservation of a quota in the amount of women that every party selects to compete in Congress elections is a great, good, fair, bad or terrible policy?
1. Great
2. Good
3. Fair
4. Bad
5. Terrible
99. Don’t know

Information Treatment Provision

Treatment group 1 will receive an information sheet that reads: “In 2017, an Electoral Fund was created, which is a public fund to finance electoral campaigns with public money. The amount of money invested in this Fund was of 1,7 billion reais in 2018, and of 4,9 billion reais in 2022. Parties receive the public money depending on the number of votes their candidates obtain.”
Treatment group 2 will receive an information sheet that reads: “Since 2020, the law has made it mandatory for parties to distribute their campaign resources proportionally between male and female candidates.”
Treatment group 3 will receive an information sheet that reads: “In 2017, an Electoral Fund was created to finance electoral campaigns with public money. Starting in this 2022 election, every vote you cast for a woman candidate will give her party twice as much public money as a vote for a man candidate.”
The control group will receive no information at all.

After the information provision treatment, all of the respondents will be asked the exact same set of questions that will serve as our outcomes of public support for the policy.

Attitudes towards Campaign Finance Quotas

Q4. Are you in favor or against that women candidates receive as much money as men candidates from parties to finance their election campaigns?
Q5. Are you in favor or against that a part of the public money from the Electoral Fund be reserved only for the candidacies of women?
Q6. Are you in favor or against that parties receive more public money from the Electoral Fund for the votes obtained by their women candidates than for the votes obtained by their men candidates in elections?
The set of possible answers will be that they are: in favor, against, indifferent or don’t know.

Finally, we ask whether they already knew about the reform we informed them about before the survey. We ask this question at the end because we do not want to give away the information or prime them before the treatment and it might serve as a sanity check in some instances.

Q7. Before this survey, were you already aware that Brazilian political parties have the obligation to spend a part of the money they receive from the Electoral Fund in the candidacies of women? (IF YES) And would you say you are well informed, more or less informed or badly informed about this reservation of money for the candidacies of women?

1. Yes, and you are well informed
2. Yes, and you are more or less informed
3. Yes, and you are badly informed
4. Not informed
99. Don’t know

Design details

Randomization method
Randomization of individual respondents into treatment groups. Among the 2,000 people interviewed, 500 will be randomly assigned to the control group, 500 will be randomly assigned to each of the three treatment groups. The survey will take the form of personal, physical interviews with trained professional interviewers stopping individuals on the street. An interviewer will approach an individual and if they accept to be interviewed and answer the first set of questions that are common to everyone, they will randomly allocate each individual respondent to a treatment group.
Randomization unit
Individual respondent
Clusters
1
Observations
2000 individual respondents
Arms
500 respondents control, 500 respondents treatment 1, 500 respondents treatment 2, and 500 respondents treatment 3.
Primary outcomes
1. Citizens' preference for gender equality in representation in elected office 2. Citizens' reform awareness about campaign financing rules (Electoral Fund) 3. Attitudes towards electoral gender quotas 4. Attitudes towards campaign finance quotas (CFQ) 5. Attitudes towards public funding aspect of CFQs 6. Attitudes towards performance-based aspect of CFQs

Cited by

No citation data available for this registration yet.

Keywords