What the evidence says
Across a set of preference studies, the evidence consistently shows that when people are asked how they would like tax or policy revenue used, environmental earmarking is the most preferred option in the absence of additional information. Focus group research conducted in five countries — Denmark, France, Germany, Ireland, and the United Kingdom — found that earmarking energy tax revenue for further emissions reductions was the most preferred option for participants in every country studied, followed by social cushioning for vulnerable groups such as low-income and remote households, with revenue-neutral redistribution (cutting other taxes) ranked least preferred. Separately, cross-sectional evidence on forced trade-offs between environmental earmarking and socially progressive redistribution finds the same ordering: people tend to prefer environmental earmarking over progressive redistribution when the two are presented as a direct trade-off.
That baseline ranking is not stable once information is introduced. A discrete choice experiment on Swiss energy taxes found that providing respondents information about the environmental effectiveness of different carbon tax designs reduces stated preference for environmental earmarking. In the same experimental setting, when respondents were also given information about the distributional effects of each design, the most favored options for revenue use became redistribution through lump-sum transfers and social cushioning — placing these above earmarking rather than below it.
A separate strand of evidence addresses the distributional properties of lump-sum transfers directly, independent of stated preference. Compensation delivered as a lump-sum transfer is progressive in its incidence: because the transfer is a fixed amount, it accounts for a larger share of income in low-income households than in higher-income households. No quantified income share or dollar figure is given in the sources on this page; the claim establishes direction of incidence, not magnitude.
Taken together, the evidence describes a two-layer pattern: an uninformed baseline preference favoring earmarking, and an information-sensitive shift toward lump-sum and cushioning transfers once respondents are told about a policy's effectiveness or its distributional consequences. No source here measures actual political-economy outcomes — voting behavior, program uptake, or revenue allocation under implemented policy — all of the evidence is preference or incidence data from focus groups, cross-sectional work, and choice experiments.
Where the evidence splits
The clearest split is between the uninformed and informed preference findings. Focus-group and cross-sectional evidence (five-country focus groups; Baranzini & Carattini, 2017; Sælen & Kallbekken, 2011) finds environmental earmarking preferred over redistributive alternatives, including progressive redistribution. The Carattini et al. (2017) discrete choice experiment on Swiss energy taxes finds the opposite ordering once information is added: providing information about a carbon tax design's environmental effectiveness reduces preference for earmarking, and providing information about distributional effects makes lump-sum transfers and social cushioning the most favored options, above earmarking. The claims themselves identify the difference in design that explains this split: the choice-experiment condition explicitly provided information on tax effectiveness and distributional effects, while the focus-group and cross-sectional studies did not describe an equivalent information treatment. The direction of preference is not fixed — it depends on whether respondents are informed about effectiveness and distributional incidence before stating a preference.