Correcting mistaken beliefs about democracy, populism and economic performance

How to cite

Ajzenman, Nicolas; Cruces, Guillermo; Fiszbein, Martin; Garcia-Zavaleta, Gaston. 2023. "Correcting mistaken beliefs about democracy, populism and economic performance." AEA RCT Registry. AEARCTR-0012309. https://doi.org/10.1257/rct.12309-1.1

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

Nicolas Ajzenman* , Guillermo Cruces , Martin Fiszbein , Gaston Garcia-Zavaleta

* Primary investigator

Registry ID
AEARCTR-0012309
DOI
https://doi.org/10.1257/rct.12309-1.1
First registered
Last updated
Status
On going
Trial dates
2023-10-16 – 2023-10-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/12309

Abstract

In the last few years, the world has been experiencing a decline in support for democratic regimes. This fact is evident from surveys (such as the World Values Survey) and electoral results across the globe. As documented by Guriev and Papaioannou (2022), the proportion of populist regimes with autocratic tendencies has grown heavily since the beginning of the century. A recent paper by Acemoglu et al. (2021) shows that democracies breed their support only when they are successful, specifically regarding specific socio-economic outcomes (growth, equality, peace, low levels of corruption, public goods). Interestingly, in the last years pre-COVID, the world has also experienced relative prosperity, which poses a puzzle: why is democracy losing support if it is relatively successful?

We hypothesize that, in the modern world, misinformation (which is widespread) could soften the relationship between successful democracies and support for democracy because many people wrongly perceive that an autocracy (or a populist government with certain autocratic tendencies) could outperform their current democratic regimes. Acemoglu et al (2019) show that, indeed, democracies cause growth and, recently Funke et al. (2020) show that populism reduces growth, but it is likely that people ignore these relationships and thus have a mistaken belief about which system has a better economic performance. If this is true, an information intervention based on scientific facts, showing that democracies that are not populist have on average better performances, could trigger an increase in support for democracy.

We propose implementing a survey/questionnaire experiment in Argentina before the upcoming presidential elections. Survey shows that Argentinians are tired of the apparent economic failures of the system (GDP per capita is virtually stagnated since 2011) and thus seem to be supporting populist candidates with clear autocratic tendencies.

Our survey experiment will have five arms:

a) A short video describing Acemoglu et al's (2019) result in a simple way, explaining that on average democracies grow faster than autocracies. It will also describe (first) the characteristics of a democracy and autocracy.
b) A short video describing Funke et al's (2020) result in a simple way, explaining that on average right wing populist governments have lower growth rates. It will also describe (first) the characteristics of a right-wing and a left-wing populist government.
c) A short video describing Funke et al's (2020) result in a simple way, explaining that on average left wing populist governments have lower growth rates. It will also describe (first) the characteristics of a right-wing and a left-wing populist government.
d) A short video describing (first) the characteristics of a right-wing and a left-wing populist government, without mentioning their economic performance. (CONTROL 1)
e) A short video describing (first) the characteristics of democracies and autocracies without mentioning their economic performance. (CONTROL 2)

We will elicit beliefs about political system and economic performance (pre and post-treatment), about attitudes towards democratic and populist (right and left) governments and, finally, voting preferences for the incoming election in Argentina.

Intervention

In the last few years, the world has been experiencing a decline in support for democratic regimes. This fact is evident from surveys (such as the World Values Survey) and electoral results across the globe. As documented by Guriev and Papaioannou (2022), the proportion of populist regimes with autocratic tendencies has grown heavily since the beginning of the century. A recent paper by Acemoglu et al. (2021) shows that democracies breed their support only when they are successful, specifically regarding specific socio-economic outcomes (growth, equality, peace, low levels of corruption, public goods). Interestingly, in the last years pre-COVID, the world has also experienced relative prosperity, which poses a puzzle: why is democracy losing support if it is relatively successful?

We hypothesize that, in the modern world, misinformation (which is widespread) could soften the relationship between successful democracies and support for democracy because many people wrongly perceive that an autocracy (or a populist government with certain autocratic tendencies) could outperform their current democratic regimes. Acemoglu et al (2019) show that, indeed, democracies cause growth and, recently Funke et al. (2020) show that populism reduces growth, but it is likely that people ignore these relationships and thus have a mistaken belief about which system has a better economic performance. If this is true, an information intervention based on scientific facts, showing that democracies that are not populist have on average better performances, could trigger an increase in support for democracy.

We propose implementing a survey/questionnaire experiment in Argentina before the upcoming presidential elections. Survey shows that Argentinians are tired of the apparent economic failures of the system (GDP per capita is virtually stagnated since 2011) and thus seem to be supporting populist candidates with clear autocratic tendencies.

Our survey experiment will have five arms:

a) A short video describing Acemoglu et al's (2019) result in a simple way, explaining that on average democracies grow faster than autocracies. It will also describe (first) the characteristics of a democracy and autocracy.
b) A short video describing Funke et al's (2020) result in a simple way, explaining that on average right wing populist governments have lower growth rates. It will also describe (first) the characteristics of a right-wing and a left-wing populist government.
c) A short video describing Funke et al's (2020) result in a simple way, explaining that on average left wing populist governments have lower growth rates. It will also describe (first) the characteristics of a right-wing and a left-wing populist government.
d) A short video describing (first) the characteristics of a right-wing and a left-wing populist government, without mentioning their economic performance. (CONTROL 1)
e) A short video describing (first) the characteristics of democracies and autocracies without mentioning their economic performance. (CONTROL 2)

We will elicit beliefs about political system and economic performance (pre and post-treatment), about attitudes towards democratic and populist (right and left) governments and, finally, voting preferences for the incoming election in Argentina.

Experimental design

We hypothesize that the support for a particular political leader partly reflects an attribution of successful economic performance to his leadership style. We focus on two types of leaders -populist leaders and dictators- and design a survey experiment to assess whether the perception of their economic performance impacts the support they receive.

Our survey is divided into five different sections:
1. Introduction and consent
2. Demographic questions
3. Prior beliefs about the economic performance of populism and dictatorships and treatment (two control groups and three treatment arms)
4. Preferences for populism and dictatorships
5. Posterior beliefs about the economic performance of populism and dictatorships.

In section 3, each respondent is randomly assigned to one of five groups. Assignment to treatment is stratified by age group and geographical zone.
1. Control 1: This group must watch a video that describes populism and answer some questions about the economic performance of right-wing and left-wing populist leaders.
2. Control 2: This group must watch a video that describes dictatorships and answer some questions about the economic performance of dictatorships.
3. Treatment 1: This group must watch a video that describes populism and answer some questions about the economic performance of right-wing and left-wing populist leaders. Additionally, we show them empirical evidence about the economic performance of right-wing populist leaders.
4. Treatment 2: This group must watch a video that describes populism and answer some questions about the economic performance of right-wing and left-wing populist leaders. Additionally, we show them empirical evidence about the economic performance of left-wing populist leaders.
5. Treatment 3: This group must watch a video that describes dictatorships and answer some questions about the economic performance of dictatorships. Additionally, we show them empirical evidence about the economic performance of dictatorships.

Design details

Randomization method
It will be randomized in office by a computer (researches will not have access to this randomization, it will be conducted by a survey firm)
Randomization unit
Individuals (no clustering).
Clusters
We aim at 800 individuals per arm. However, we expect some attrition after the attention check.
Observations
We aim at 800 individuals per arm. However, we expect some attrition after the attention check.
Arms
We aim at 800 individuals per arm. However, we expect some attrition after the attention check.
Primary outcomes
Beliefs about political systems (democracy, autocracy, left wing and right win populism) and economic performance: we will ask if a right-wing populist, left-wing populist versus non-populist have better economic performance (Control 1 and two treatment arms related to populism). We will ask if autocrat versus democrats have better economic performance (Control 2 and one treatment arm related to autocracy). We expect citizens to update their beliefs according to the information we provided in the videos. Preferences for different political regimes: we will mention different characteristics typically associated with popoulism, nonpopulism, democracies and autocracies and ask if the individual agrees (from 1 to 5) with different statements related to how desirable/non desirable those characteristics are in a political system. We will ask at the end of the survey who they are planning to vote in the upcoming presidential elections. Important 1: we have a specific control for "autocracy versus democracy" and one specific control for "populism (right or left) versus nonpopulism". Important 2: we will show unpooled and pooled results for populism (that is, left+right populism). Important 3: in the treatment arms, we described a populist regime but called it "personalist"; that is because the word "populism" in Spanish is too much associated (in the public view) with specific politicians. Important 4: we will analyze the beliefs results controlling by pre-treatment beliefs (the idea is to measure the change on beliefs) and also will analyze heterogeneous effects by pre-treatment beliefs (on the relevant topic: populism or autocracy)
Secondary outcomes
Knowledge about populist/non populist/autocracies and democratic governments: we will ask specific characteristics that we mentioned in the videos to test if individuals recall what they saw. Demand for information: we will ask at the end of the survey if they would like to know more about democratic regimes and economic performance from scientific papers.

Cited by

No citation data available for this registration yet.

Keywords