The Politicization of Animal-Based Food Production and Consumption: Effects on Attitudes, Policy Preferences and Voter Mobilization

Participants are randomly assigned to one of four conditions. The placebo control condition presents neutral factual information on the instruments of direct democracy in Switzerland (popular initiatives and referendums), unrelated to the Ernährungsinitiative. The remaining three conditions present information about the Ernährungsinitiative that varies in the degree to which the topic is presented as salient and controversial: a condition presenting neutral factual information about the initiative (T1; salience treatment); a condition additionally presenting balanced pro and con arguments reflecting the existing controversy around the initiative, in neutral, factual language (T2; politicization treatment, neutral); and a condition presenting the same balanced arguments using authentic public campaign material, including campaign imagery (T3; politicization treatment, authentic).
The four conditions are designed to vary the perceived politicization of the topic – i.e., its perceived salience and polarization: the placebo control involves no exposure to salient or polarizing information about the initiative; T1 introduces salience without indicating that the topic is controversial; T2 and T3 additionally present balanced pro/con arguments reflecting the existing controversy, with T3 using authentic campaign material and imagery, which may elicit a comparatively higher degree of perceived polarization than T2's neutral language.

2026-09-01

2026-09-20

In the following, you can see the key dependent variables of interest in this experiment, corresponding to our main hypotheses (H1–H6):
• Attitudes toward animal-based food consumption: The perceived importance of reducing the production and consumption of animal-based products (7-point Likert scale from very unimportant to very important), measured in Wave 1 and Wave 2.
• Food policy preferences: Support for or opposition to seven specific food policy measures related to animal-based food production and consumption (e.g., meat taxation, subsidies for plant-based alternatives, mandatory climate labelling), each on a 7-point Likert scale (ranging from strongly oppose to strongly support), measured in Wave 1.
• Voting mobilization: Stated intention to vote on the Ernährungsinitiative (Wave 1) and voluntary self-reported participation in the vote (Wave 2), each measured as a binary variable indicating whether respondents planned to (or did) participate in the vote.
• Voting direction: Stated intended vote on the Ernährungsinitiative, measured on a 4-point scale ranging from definitely accept to definitely reject (Wave 1), and voluntary self-reported actual Yes/No vote (Wave 2).
• Satisfaction with democracy (as proxy for democratic legitimacy perceptions): Overall satisfaction with the way democracy works in Switzerland, measured on a 7-point Likert scale (ranging from not at all satisfied to very satisfied), measured in Wave 1.
• Populist attitudes: Agreement with a parsimonious 3-item scale measuring populist attitudes across three dimensions – popular sovereignty, Manichean worldview, and anti-elitism (Van Hauwaert et al., 2020) – each on a 7-point Likert scale (ranging from strongly disagree to strongly agree), measured in Wave 1.
• Affective/emotional reaction (mechanism check): Self-reported anxiety, anger, satisfaction, and enthusiasm following treatment exposure (7-point Likert scales from not strong at all to very strong), measured in Wave 1 and Wave 2.
• Activation of prior attitudes (mechanism check): The extent to which the treatment information prompted respondents to reflect on their own attitudes (7-point Likert scale from not strong at all to very strong), measured in Wave 1 and Wave 2.

Secondary Outcomes
• Intention to change consumption of animal-based products: Stated intention to reduce or increase consumption of animal-based products in the next two weeks (5-point scale from significantly reduce to significantly increase), measured in Wave 1.
• Perceived societal support: Respondents' estimate of how the Swiss population will vote on the Ernährungsinitiative and Neutralitätsinitiative (4-point categorical scale from will definitely be accepted to will definitely be rejected), measured in Wave 1.
• Attitude reinforcement: The extent to which the treatment information is perceived to have strengthened versus changed respondents' pre-existing attitudes (7-point bipolar scale from strongly changed to strongly strengthened), measured in Wave 1 and Wave 2.
• Perceived salience (manipulation check): Personal and perceived public importance of animal-based food production and consumption as a political topic (7-point Likert scales from very unimportant to very important), measured in Wave 1 and Wave 2.
• Perceived polarization (manipulation check): Personal and perceived public contestation of animal-based food production and consumption as a political topic (7-point Likert scales not contested at all to very contested), measured in Wave 1 and Wave 2.
• Trust in treatment information (mechanism check): Perceived credibility of the information presented (7-point Likert scale from not credible at all to very credible), measured in Wave 1 only.

This study employs a randomized controlled trial (RCT) using a two-wave online survey experiment fielded around a Swiss popular vote on the Ernährungsinitiative. In Wave 1, participants are randomly assigned to one of four conditions (placebo control, T1, T2, T3) that vary in the salience and polarization of information presented about the initiative (see Interventions). Following treatment exposure, participants report their attitudes, food policy preferences, and voting intentions. Wave 2, fielded shortly after the vote, reassesses these outcomes and collects participants' voluntarily reported voting behavior.

Not available

Randomization is conducted by the survey platform (Qualtrics) using its built-in randomizer feature with even presentation enabled, which randomly assigns participants to one of the four conditions while ensuring approximately equal group sizes across the course of data collection. Randomization is independent of all measured covariates; balance on key pre-treatment covariates (e.g., age, gender, education, political orientation) will be verified in the final data, and any imbalances will be reported in the analysis.

The unit of randomization is the individual respondent.

No

-

The target sample size for Wave 1 is 2,800 adult Swiss residents eligible to vote in federal popular votes. Participants are recruited through the commercial panel provider Bilendi over a fielding period of approximately three weeks (calendar weeks 36–38, September 2026), until the target sample size and quotas are reached. The final achieved sample size may vary slightly depending on response and completion rates during this recruitment period. Wave 2 is fielded shortly after the vote (calendar weeks 41–43, October 2026) among all Wave 1 respondents; based on expected attrition of 20–30% between waves, the anticipated Wave 2 sample size is approximately 2,000–2,200 respondents.

Participants will be randomly assigned to one of four treatment arms with the aim of achieving approximately equal group sizes:
• Placebo control: approximately 700 respondents
• T1 (salience treatment): approximately 700 respondents
• T2 (politicization treatment, neutral): approximately 700 respondents
• T3 (politicization treatment, authentic): approximately 700 respondents
Actual sample sizes may vary slightly due to randomization and participation patterns.

Assuming a small effect size (Cohen's d = 0.20) and 80% power, with a Bonferroni-corrected significance threshold accounting for four planned pairwise contrasts (α_corr = 0.0125), the target sample size of N = 2,800 (700 per arm) is powered to detect the specified minimum effect size for the primary outcomes, accounting for expected attrition of 20–30% between Wave 1 and Wave 2.