Are Assessed Risks and Optimism Bias Affected by Partisan Information Cues

How to cite

Wolaver, Amy; Doces, John; Goldberg, Jack. 2021. "Are Assessed Risks and Optimism Bias Affected by Partisan Information Cues." AEA RCT Registry. AEARCTR-0008547. https://doi.org/10.1257/rct.8547-1.0

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

Amy Wolaver* , John Doces , Jack Goldberg

* Primary investigator

Registry ID
AEARCTR-0008547
DOI
https://doi.org/10.1257/rct.8547-1.0
First registered
Last updated
Status
In development
Trial dates
2021-11-15 – 2021-12-31
Countries
United States of America (National)
JEL codes
Z18 Public Policy
Secondary IDs
I18 Government Policy • Regulation • Public Health ; o K32 Energy, Environmental, Health, and Safety Law
Registry record
http://www.socialscienceregistry.org/trials/8547

Abstract

Objective: To determine whether assessed risks to oneself, assessed risks to the general population, and optimism bias about a variety of policy topics are affected by partisan informational cues. Half of the respondents will be randomized to receive a politically neutral version of the risk area and half to a version including a partisan clue. For example, risk of infection from COVID versus risk of infection from the flu; risk from a natural disaster with no attributed cause versus risk from a natural disaster caused by climate change, etc... Respondents will then be asked for their level of support for policies related to each of the topical areas used in the risk assessement.

Intervention

Views about the risk from a variety of policy areas have become increasingly polarized in the United States. Prior research has documented that Democrats rated the risks associated with COVID-19 as much higher than Republicans. Optimism bias, the rating of one’s own risk as lower than a peer’s, is an important correlate with a variety of behaviors and views. We propose to test whether partisan framing cues affect the assessed risk to self, the assessed risk to a peer, and the level of optimism bias for a number of policy issues. For each policy, we randomize between a liberal, conservative or neutral cue (see figure for examples). We include a neutral traffic accident scenario as a control. Following the assessment of risk, we ask for levels of support for policies related to each risk scenario and test for impacts on both COVID-19 and flue vaccination status/intentions.
We will test for the following hypotheses:
COVID-19 will raise the risk assessment for Democrats/liberals/non-Trump voters and depress the risk assessment for Republicans/conservatives/Trump voters.
On the other issues:
Liberal cues will raise the risk assessment for the scenario for Democrats/liberals/non-Trump voters relative to the neutral or conservative cues.
Conservative cues will raise the risk assessment for the scenario for Republicans/conservatives/Trump voters relative to the neutral or conservative cues.
Higher levels of risk assessment will increase support among the partisan group for the related policy action.
Research Questions:
Will the levels of risk-to-self versus risk-to-a-peer be affected differently by the partisan cues (thus affecting optimism bias)?
What effect, if any, will optimism bias have on support for policy action?

Experimental design

Different topical scenarios will be randomized between liberal, conservative and neutral partisan cues in questions assessing the respondent's personal risk and their assessed risk of harm to another, similar person from the scenario.

Design details

Randomization method
Randomization by computer in the online survey questions.
Randomization unit
Individual
Clusters
1000-1200 individuals
Observations
1000-1200 individuals
Arms
500-600 for most interventions, 300-400 for one intervention with 3 randomization levels.
Primary outcomes
Level of assessed personal risk and risk to another, similar person, level of optimism bias, level of support for policy, and vaccination status and intentions.

Cited by

No citation data available for this registration yet.

Keywords