Animal welfare is a stronger determinant of public support for meat taxation than climate change mitigation in Germany

The study examines public support in Germany for a tax on meat, focusing on how specific policy attributes shape voter approval. Policymakers are debating meat taxes primarily to address climate change and animal welfare concerns. The authors investigate whether support differs depending on the tax’s stated purpose (animal welfare vs climate mitigation), whether the tax is uniform per kilogram or differentiated by product externalities (emissions or husbandry level), and whether making expected behavioral effects salient affects support. Based on prior evidence suggesting animal welfare motives resonate strongly with consumers, the authors hypothesize higher support for animal welfare framing, decreasing support with higher tax levels, and potentially higher support for differentiated (Pigouvian) taxes when their behavioral effects are salient.

Past surveys and choice experiments show that acceptance of (carbon) taxes is sensitive to policy design: avoiding the word “tax,” earmarking revenues, progressive features, and clear communication of impacts can raise support. Studies on food taxes and labels suggest animal welfare arguments often motivate consumers more than climate arguments in voluntary contexts. Modeling studies indicate meat taxes can reduce environmental and health externalities. Evidence on whether a differentiated Pigouvian design increases public support relative to a uniform tax is limited. Prior research also highlights that perceived consequentiality and trial periods can shape acceptance, and that earmarking is often critical for supporting stand-alone carbon taxes. This study extends that literature by directly comparing animal welfare versus climate justifications, uniform versus differentiated tax structures, and varying salience of expected behavioral effects in a referendum-style setup.

Design: A pre-registered (AEARCTR-0008507) online referendum choice experiment conducted in Germany between 30 November and 9 December 2021 with 2,855 adult participants recruited via a professional panel provider. The survey presented a proposal for a levy on meat products (fresh meat, sausages, cold cuts) with revenues earmarked to either animal welfare improvements or climate protection. Experimental factors (between-subjects; random assignment): - Tax justification: animal welfare levy vs climate levy. - Degree of differentiation: uniform per-kg levy vs differentiated levy (by husbandry level for animal welfare or by meat type for climate; higher external damage faced higher rates). - Salience of behavioral effects: low salience (belief elicitation after voting) vs high salience (belief elicitation before voting). Within-subject variation: Each participant voted in six consecutive referenda that differed only by the tax rate, increasing across tasks. Average tax levels ranged from €0.19/kg (≈€25/t CO2) to €1.56/kg (≈€200/t CO2). For climate-differentiated rates, levels were derived from FAO GLEAM v2.0 emissions and CO2 prices. Uniform rates were weighted averages using 2020 German meat consumption shares. Animal welfare treatments used the same rate schedules mapped onto husbandry levels (Haltungsform 1–4), enabling direct comparison of justification without confounding by amounts. Consequentiality: Participants were informed that results would be sent to relevant committees in the German parliament, and an example letter was shown. Abstention was allowed, mirroring real referenda. Belief elicitation: Participants stated expectations on overall meat consumption and by subcategories (beef/level 1, lamb/level 2, pork/level 3, poultry/level 4): decrease, remain, or increase. Sample handling: Of 3,169 completions, 314 were excluded for extreme durations (<5 min or >45 min), leaving 2,855. Vegetarian/vegan participants were not excluded. Final valid votes for analysis were 15,908 (abstentions excluded from support rate calculation). Statistical analysis: Ordinary least squares linear probability models estimated average marginal effects of tax level (continuous € per kg), justification (climate vs animal welfare), differentiation (differentiated vs uniform), salience (high vs low), and a differentiation×salience interaction on support (1 yes, 0 no), clustering standard errors by respondent. Robustness checks used logistic regressions and models with categorical tax levels and included controls (demographics, political views, consumption habits, consequentiality, attention, and social desirability). Exploratory generalized ordered logistic regressions analyzed belief outcomes (three-category ordinal) to estimate marginal effects of treatment variables on perceived consumption changes.

- Support declines with tax level: Each €0.10/kg increase reduces support by 2.6 percentage points (95% CI: −2.78 pp, −2.49 pp). Average support is 62% at €0.19/kg and 23% at €1.56/kg. Linear interpolation indicates 50% support at ~€0.39/kg overall. - Animal welfare framing increases support: Animal-welfare-justified taxes receive 11.1 pp (95% CI: 8.3, 14.0) higher support than otherwise identical climate-justified taxes across all levels. - Differentiation has negligible effect on support: The differentiated vs uniform design shows at most a small, statistically insignificant impact on support (e.g., β ≈ 0.024; CI includes zero; reported CI −1.6 pp to 6.4 pp), contradicting the hypothesis that Pigouvian design boosts acceptance. - Salience modestly increases support: High salience (belief task before voting) raises support by 4.0 pp (95% CI: 0.0, 8.1). The interaction between salience and differentiation is near zero and insignificant (β ≈ −0.007; CI −6.5 pp, 5.0 pp), indicating that emphasizing behavioral effects does not preferentially increase support for differentiated taxes. - Referendum pass points (50% support, via linear interpolation): animal welfare uniform €0.4443/kg; animal welfare differentiated €0.5291/kg; climate uniform €0.2598/kg; climate differentiated €0.2885/kg; all schemes €0.3734/kg. - Beliefs about impacts: - Participants expect differentiated taxes to shift consumption toward lower-impact products (higher probability of ‘decrease’ for highly taxed categories; reverse for least-taxed), while expecting little change in overall consumption attributable to differentiation. - Climate framing is perceived as more likely to reduce consumption within subcategories than animal welfare framing, even under uniform taxes, though overall consumption effects are not perceived as different by framing. This suggests support for animal welfare framing is not driven by beliefs about stronger consumption reduction.

The findings indicate that how a meat tax is justified matters substantially for public support: animal welfare framing yields much higher acceptance than climate framing. This remains true even though participants tend to believe climate-framed taxes reduce subcategory consumption more, implying support is driven by values or perceived co-benefits (e.g., direct animal welfare improvements) rather than expectations about reduced meat consumption. The tax’s structural design (uniform vs differentiated) does not materially affect acceptance, despite respondents understanding that differentiated taxes better steer consumption composition; this suggests limited public appreciation for Pigouvian differentiation per se. Higher salience of behavioral effects modestly increases support, consistent with evidence that experience or reflection on policy effects can enhance acceptance. Policy implications include starting with lower tax rates (ratcheting over time), clearly earmarking and communicating the purpose of revenues, emphasizing animal welfare benefits, and transparently explaining intended behavioral effects. Since support is most sensitive to framing and level rather than differentiation, policymakers may face little political downside in adopting the more effective differentiated design if desired, while recognizing it is unlikely to increase support relative to a uniform approach.

This study shows that in Germany, public support for a meat tax depends more on the stated purpose and the tax level than on whether the tax is uniform or differentiated. Animal welfare framing substantially increases acceptance compared to climate framing, and emphasizing behavioral effects yields a modest boost in support. The results suggest that policymakers seeking to implement meat taxation should consider low initial rates with potential future increases, earmark revenues, and communicate animal welfare benefits and intended behavioral changes clearly. Future research should disentangle the relative contributions of framing versus earmarking, examine dynamic support for ratcheting tax paths, assess combined justifications (and split revenue uses), and compare support across countries and political contexts.

- Hypothetical voting context: Although consequentiality was enhanced by informing participants that results would be shared with parliament committees, real-world behavior may differ from stated preferences. - Snapshot in time: Support reflects contemporaneous economic conditions, debates, and incomes; generalizability over time may be limited. - Mapping of animal welfare rates: Animal welfare tax levels were derived from climate-based calculations and mapped to husbandry levels for comparability, which may not reflect true welfare-related damages. - Framing vs earmarking: The design does not isolate the independent effect of framing from revenue earmarking; both were bundled. - Online panel and representativeness: While the sample matches key demographics reasonably well, it is an online adult sample; education levels are somewhat higher than the general population. - Self-reports and attention: Despite exclusions for extreme durations and attention checks, residual measurement error or social desirability bias may remain. - Exploratory belief analysis: Belief-related findings were not pre-registered and should be interpreted cautiously.