Encouraging Protest Participation: A Field Experiment to Study the Consequences of Political Experience and Peer Effects in Protest Participation
Last registered on July 12, 2017

Pre-Trial

Trial Information
General Information
Title
Encouraging Protest Participation: A Field Experiment to Study the Consequences of Political Experience and Peer Effects in Protest Participation
RCT ID
AEARCTR-0002301
Initial registration date
July 08, 2017
Last updated
July 12, 2017 9:40 PM EDT
Location(s)
Region
Primary Investigator
Affiliation
Stanford University
Other Primary Investigator(s)
PI Affiliation
University of California, Berkeley
PI Affiliation
University of Munich
PI Affiliation
University of Chicago
PI Affiliation
Hong Kong University of Science and Technology
Additional Trial Information
Status
On going
Start date
2017-06-08
End date
2017-08-15
Secondary IDs
Abstract
Participation in political movements has the potential to shape political institutions, public policy, and economic outcomes. In this project, we study the causal effects of a “nudge” that induces individuals to turnout to a high-stakes, anti-authoritarian protest. Studying the effects of this sort of nudge, we have two broad areas of inquiry: first, if an individual is randomly nudged to experience a political protest, how does that affect her political attitudes and subsequent political behavior? Are there short-run effects? Persistent effects? Second, does nudging individuals to participate in a protest have spillover effects on their peers’ decisions to protest? Determining whether there are peer effects in the decision to turnout to protest is important for understanding the dynamics of protest participation and protest size.
External Link(s)
Registration Citation
Citation
Bursztyn, Leonardo et al. 2017. "Encouraging Protest Participation: A Field Experiment to Study the Consequences of Political Experience and Peer Effects in Protest Participation." AEA RCT Registry. July 12. https://www.socialscienceregistry.org/trials/2301/history/19382
Experimental Details
Interventions
Intervention(s)
In a second part to our survey (where recruitment occurs through emails to participants who have completed the first part), we randomly assign student participants in the survey into three groups (the details of treatment assignment are described in the following subsection): (i) the protest encouragement treatment group; (ii) a “subway” control group; and, (iii) a “pure” control group.

The treatment condition is intended to “nudge” individuals on the margin of protest attendance into choosing to attend the July 1 protest March by paying them a meaningful amount of money for a small commitment of time while at the March.

The “subway” control condition provides a very similar nudge, and similar payment, for a similar time commitment; however, rather than nudging toward attending the protest March on July 1, the nudge is toward going to central Hong Kong (near the location of the March) on the following weekend. This comparison group will help address concerns about experimenter demand effects and income effects when interpreting differences between a pure control group and the treatment group.

The pure control condition provides another benchmark against which to compare the treatment group.
Intervention Start Date
2017-06-17
Intervention End Date
2017-08-15
Primary Outcomes
Primary Outcomes (end points)
In the submitted pre-analysis plan document, we include the 23 questions asked in the survey after treatment intervention. These questions are numbered, and we now describe how they are considered in terms of outcomes of interests. Outcome variables of interest for our study of the effects of protest participation on individuals’ beliefs and attitudes: * Question 1: is there a first stage? * Questions 8-9: is there learning about the protest and about politics? * Questions 10-13: first set of “primary outcomes of interest” – is there an effect of the treatment on political beliefs? * Questions 14-17: second set of “primary outcomes of interest” – is there an effect of the treatment on political attitudes? * Questions 18-21: third set of “primary outcomes of interest” – is there an effect of the treatment on second order political beliefs? * Questions 22-23: fourth set of “primary outcomes of interest” – is there an effect of the treatment on political attitudes as expressed in real donations to political organizations? Outcome variables of interest for our study of peer effects in the decision to protest: * Question 1: this is the primary outcome of interest in the peer effects study. * Questions 3-7: set of questions suggesting the mechanisms potentially explaining an effect with respect to question 1.
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
The sample from which we draw is recruited from the Hong Kong University of Science and Technology. Students are recruited via a mass email sent to the undergraduate student population. The email will invite students to participate in an online survey in return for payment for their participation.

All subjects were asked to complete a baseline survey in mid-June of 2017, which included a range of questions used in our experimental analysis. These questions included: cohort, major, gender, housing (on campus or off), whether they were born in Hong Kong, and whether they attended high school in Hong Kong. We also ask about subjects’ political attitudes, and plans to attend the pro-democracy March on July 1, 2017.

In a second part to our survey (where recruitment occurs through emails to participants who have completed the first part), we randomly assign student participants in the survey into three groups (the details of treatment assignment are described in the following subsection): (i) the protest encouragement treatment group; (ii) a “subway” control group; and, (iii) a “pure” control group.

Outcome data will be collected via another online survey conducted between mid-July and mid-August 2017, and the detailed outcome questionnaire will be described in the attached pre-analysis plan.
Experimental Design Details
Randomization Method
Randomization done in office by a computer.
Randomization Unit
To identify the causal effect of protest participation, we will randomly assign the treatment condition across individuals, and use this variation to identify the causal effect of the treatment (encouraging protest participation) on individuals’ political beliefs and attitudes.

To identify peer effects in the protest decision (and the possibility of non-linear peer effects depending on the fraction of peers protesting), our treatment will also randomly vary at the “social group cell” level, as we will vary treatment intensity (i.e., the fraction of individuals treated) across cells.

We discuss the two levels of randomization in details in the attached pre-analysis plan document.
Was the treatment clustered?
Yes
Experiment Characteristics
Sample size: planned number of clusters
98 social group cells.
Sample size: planned number of observations
Approximately 1650 university students participants.
Sample size (or number of clusters) by treatment arms
Cells are randomly assigned to receive either 0 individuals in the treatment group (approximately 15% of cells), 1 individual in the treatment group (around 20% of cells), 50% of the individuals in the treatment group (approximately 30% of cells), or 75% of the individuals in the treatment group (approximately 35% of cells).

Cells are also randomized into receiving 0 individuals in the “subway” control group (approximately 40% of cells), 1 individual in the treatment group (around 30% of cells), 50% of the individuals in the treatment group (approximately 25% of cells), or 75% of the individuals in the treatment group (approximately 5% of cells). The “subway” control group assignment at the cell level was cross randomized, subject to the constraint that we could not have cells with both 75% treatment and either 50% or 75% “subway”, nor could we have cells with 50% treatment and 75% “subway.”

Within cells, assignment to the treatment group and to “subway” control group status across individuals is random. This leaves us with approximately 45% treatment group subjects overall, 20% “subway” control group subjects, and 35% pure control group subjects.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
Hong Kong University of Science and Technology
IRB Approval Date
2017-06-26
IRB Approval Number
147
IRB Name
Stanford IRB
IRB Approval Date
2017-05-30
IRB Approval Number
38481
Post-Trial
Post Trial Information
Study Withdrawal
Intervention
Is the intervention completed?
No
Is data collection complete?
Data Publication
Data Publication
Is public data available?
No
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers