Framing Winners and Losers: Distributional Outcomes and Support for Carbon Pricing

We conduct an online survey experiment with German households living in rental housing, randomly assigning participants to receive information about the upcoming EU Emissions Trading Scheme 2 (EU-ETS2), which will take effect in 2027, framed either to emphasize negative consequences (costs and burdens) or positive outcomes (fairness and improved living conditions) for disadvantaged households.

In total, we generate four experimental groups:
(1) Control - Neutral, factual information about the policy without any distributional framing or political attribution. This message is seen by all experimental groups.
(2) Negative Frame, Anonymous - Quote by a politician that emphasizes costs or burdens of the policy (e.g., higher expenses, unfairness, social division).
(3) Negative Frame, with Politician Cue - Identical to the negative frame, but explicitly attributed to a named politician including party affiliation. This tests whether political source cues amplify or polarize reactions along partisan lines— i.e., evidence of affective polarization.
(4) Positive Frame - Factual message that emphasizes benefits of the policy (e.g., reducing inequality, fairness, or broader social gains).
Please refer to the Pre-Analysis Plan for further details.

2025-10-13

2025-11-02

The study focuses on three primary outcomes. We seek to measure the extent to which alternative framings of the upcoming EU Emissions Trading Scheme 2 (EU ETS2) shape (1) public attitudes toward the policy (e.g. general support, perceived social justice, perceived effectiveness, perceived cost burden), (2) political alignment, specifically voters’ willingness to support politicians/political party who endorse/oppose the policy, and (3) the role of affective polarization, specifically whether people’s attitudes toward the policy diverge depending on their partisan alignment.

Throughout our analysis, we elicit stated support and further attitudes towards the EU ETS 2 using a 5-point Likert scale. We will test whether the politician cue in the third treatment arm amplifies the negative frame along partisan lines. That is, we will capture whether partisan support moderates the effect of the cue, which is the essence of affective polarization.

The study investigates three secondary outcomes. First, how the framing influences tenants’ willingness to accept higher rents in exchange for lower energy costs. Second, the personal burden mechanisms through which framing may shape support for the EU ETS2, focusing on perceptions of relative economic position and whether respondents themselves live in an unrenovated dwelling. Third, whether respondents can identify the evaluative direction of the framing (i.e., whether the policy is presented in a more favorable or unfavorable light) using a 5-point Likert scale, and how more informed respondents (those with prior knowledge of the EU ETS29 are better able to recognize framing directions.

Please refer to the Pre-Analysis Plan for further details.

All respondents first receive a neutral baseline message with factual information about the policy (EU ETS2). They are then randomly assigned to one of three framing conditions: a negative frame emphasizing costs and inquality using an anonymous quote, identical negative frame but with explicit attribution to a politician/party affiliation, or a positive frame highlighting benefits such as fairness and inequality reduction.

The experiment begins with an introduction to the upcoming EU Emissions Trading Scheme 2 (ETS2), which from 2027 will place a price on CO₂ emissions from heating fuels and transport across Europe, replacing the existing German CO₂ price. Respondents are informed about the initial price, with future prices determined largely by supply and demand but subject to regulation, while Germany’s CO₂ cost-sharing law between tenants and landlords will remain in effect. Participants are also asked about their prior awareness of the ETS2, with response options ranging from full knowledge to never having heard of it.

They are then randomly assigned to one of three framing conditions:
(1) Negative frame: Message highlighting costs and burdens of the policy (e.g., higher expenses, unfairness, social division).
(2) Negative frame with politician cue: Identical to the negative frame, but explicitly attributed to a named politician to test the role of political source cues.
(3) Positive frame: Message emphasizing benefits of the policy, such as fairness, inequality reduction, and broader social gains.

Randomization done in office by a computer.

Randomization unit is at the survey respondent level.

No

None

The experiment will involve approximately 5,700 tenants who are heads of their households across Germany.

Approximately 1400 household heads per treatment arm.