Cancelled: Conformity or Silence?

How to cite

Samahita, Margaret; Morales, Juan S.. 2021. "Cancelled: Conformity or Silence?." AEA RCT Registry. AEARCTR-0007905. https://doi.org/10.1257/rct.7905-4.3

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

Margaret Samahita* , Juan S. Morales

* Primary investigator

Registry ID
AEARCTR-0007905
DOI
https://doi.org/10.1257/rct.7905-4.3
First registered
Last updated
Status
Completed
Trial dates
2021-08-19 – 2021-08-26
Countries
United States of America ()
JEL codes
C90, D83, D91, P16
Registry record
http://www.socialscienceregistry.org/trials/7905

Abstract

Social norms, social image concerns, and social stigma are all factors which can shape the public expression of opinion. With increased political polarization and the rise of social media, many prominent voices have recently argued that some of these social norms have become too strict and that fear of social backlash has led to a stifling of freedom of expression. We study how perceived social pressure affects the public expression of opinion. How sensitive are individuals to current concerns of social backlash? To what extent may public views differ from private opinion? Do these dynamics vary depending on individual characteristics like political affiliation or social media use? To investigate these questions, we first outline a simple model that formalizes these ideas and helps to frame our study. We highlight how social pressure can affect public opinion either through a change in publicly stated views towards a norm (conformity) or by inducing self-censorship (silence). We then propose to conduct an online survey experiment that allows us to disentangle these two effects and to better understand these social dynamics in the context of two current debates: Gender and LGBTQ issues, and Race.

Intervention

Our intervention is priming participants with a short text about "cancel culture" which includes two examples of individuals who have experienced negative backlash on social media as a result of their posts.

Experimental design

Participants are randomised into 1 of 5 treatments:
T1: prime text, elicitation of stated opinion, control text, elicitation of willingness to publish
T2: control text, elicitation of stated opinion, primetext, elicitation of willingness to publish
T3: control text, elicitation of stated opinion, control text, elicitation of willingness to publish
T4: prime text, elicitation of stated opinion AND elicitation of willingness to publish
T5: control text, elicitation of stated opinion AND elicitation of willingness to publish

Design details

Randomization method
Randomisation is done on the Qualtrics program.
Randomization unit
Individuals are randomised into 1 of the 5 treatments. We allocate 27.5% of participants to each of T1-2, and 15% to each of T3-5. This enables us to compare behavior when primed at the start vs later with greater power. Individuals are also randomised into either the Gender or Race topic. While we show both statements to elicit stated opinion and willingness to publish, we only ask one of them for the questions about willingness to pay for publication and perceived norm.
Clusters
900 individuals (300 each of self-reported Republicans, Democrats and Independents)
Observations
900 individuals (300 each of self-reported Republicans, Democrats and Independents)
Arms
27.5% in T1-2, 15% in T3-5
Primary outcomes
First, we elicit participants' stated opinion on two controversial statements about gender and race (1-7 strongly disagree - strongly agree). Second, we elicit participants' willingness to publish their stated opinion on a public social media site anonymously and with their name (willing/unwilling). For a randomly selected topic, we also elicit participants' willingness to pay for the publication (how many lottery tickets participant is willing to give up/accept).

Cited by

No citation data available for this registration yet.

Keywords