Message bias and consumer attitudes towards antibiotics in beef and dairy

How to cite

Bernedo, Maria; Lhermie, Guillaume; Lansing, David. 2023. "Message bias and consumer attitudes towards antibiotics in beef and dairy." AEA RCT Registry. AEARCTR-0011819. https://doi.org/10.1257/rct.11819-1.0

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

Maria Bernedo* , Guillaume Lhermie , David Lansing

* Primary investigator

Registry ID
AEARCTR-0011819
DOI
https://doi.org/10.1257/rct.11819-1.0
First registered
Last updated
Status
Completed
Trial dates
2022-11-22 – 2023-01-25
Countries
United States of America (Online)
JEL codes
C9, D91, Q1
Registry record
http://www.socialscienceregistry.org/trials/11819

Abstract

In this experiment we examine how an information intervention affects people’s attitudes towards food attributes. Specifically, we focus on antibiotic use in food production. First, we test whether one-sided and two-sided messages affect consumers’ willingness to pay for the antibiotics free premium. Second, we examine the kinds of consumers that are most receptive to either of the messages. We examine factors potentially affecting a consumer's receptivity to the different messages, such as existing subjective and objective knowledge about antibiotics, their scientific literacy, their education level and their cultural attitudes. Finally, we study a potential mechanism through which the messages change people’s preferences. We hypothesize whether a person that delivers a balanced message is perceived as more trustworthy than a person that delivers a biased message and therefore the consumer is more open to contrary positions.

Experimental design

Our survey will be administered by Qualtrics, an online survey platform. The survey approximately takes 15 minutes to complete.

Sample recruitment criteria
Participants are a population-representative sample that matches the national U.S. composition in terms of the age, gender, income, race, and rural/urban distribution. They are required to be 18 years old or older and eat meat and drink milk. Participants are considered inattentive and dropped from the sample if they meet one or more of the following criteria: (i) they spent 45% of the median time or less on the survey, (ii) provided pattern-based responses in the two matrix shaped cultural worldview questions and (iii) entered impossible height and weight values.

Survey
We will pose the survey questions before and after the information treatment. Before the treatment we will ask some socio-demographic questions (age, gender, race, education and income) and a subjective knowledge question. After these questions we will ask people for their willingness to pay (WTP) for products with the antibiotic-free label and without it. The products are ground-beef and milk. The WTP questions are hypothetical and open-ended. After the treatment we will repeat the WTP questions. In total each participant will provide eight WTP values. Additionally, we randomize the order in which the beef and milk questions appear before and after the video.

After the second round of WTP questions we ask additional questions that we believe correlate with the premium or could moderate the treatment effect. We ask these questions at the end to avoid priming participants (Folk and Zimmerman, 2013; Grewenig et al. 2020).

Information treatment
We randomly assign participants to one of the three videos about the use and effects of antibiotics (the positive message, the negative message or the balanced message) or to a control video (see Appendix). The “control video” is about a different topic entirely.
At the beginning of each video we added the same opening chyron that introduced the speaker as a professor of a renowned university (see Appendix). No other reference to sources is mentioned in the videos. By harmonizing the source of the information, we expect that the effect of the message will depend only on its content and not on people’s perception about the quality of the information.

Design details

Randomization method
Randomization done by Qualtrics.
Randomization unit
We randomize by individual.
Clusters
We will not cluster observations.
Observations
1,500 individuals
Arms
375 individuals per treatment.
Primary outcomes
1. Premium for antibiotic free products at baseline, i.e. the difference between the willingness to pay (WTP) for the antibiotic free product and the WTP for the product raised using antibiotics 2. Change in the antibiotic free premium: Difference between the antibiotic free premium after the information is provided and the one at baseline

Cited by

No citation data available for this registration yet.

Keywords