Climate Levy, Informational Messages, Acceptability, and Treatment Effects of Framing Effectiveness and Earmarking (CLIMATEFEE)
Last registered on October 23, 2018


Trial Information
General Information
Climate Levy, Informational Messages, Acceptability, and Treatment Effects of Framing Effectiveness and Earmarking (CLIMATEFEE)
Initial registration date
October 23, 2018
Last updated
October 23, 2018 9:33 PM EDT
Primary Investigator
Georgia State University
Other Primary Investigator(s)
PI Affiliation
CICERO International Center for Climate Change Research
PI Affiliation
Appalachian State University
PI Affiliation
Georgia State University
Additional Trial Information
In development
Start date
End date
Secondary IDs
There is a general consensus among economists that carbon pricing represents the most cost-effective way to tackle climate change. Yet, only a relatively limited number of jurisdictions have implemented either a carbon tax or a carbon trading scheme. Most often, the main barrier to the implementation of carbon taxes is lack of public support. The literature in economics and other disciplines has pointed to information asymmetries driving this shortfall in public support. People tend to underestimate the benefits of carbon taxes and their incentive effect, and overestimate their drawbacks. Since people underestimate the effectiveness of carbon taxes, earmarking revenues for additional abatement tends to be relatively popular. Public ballots on carbon taxes are rare. We are aware of only three instances in which people voted on a carbon tax: in 2000 and 2015 (both times in Switzerland) and in 2016 (in Washington State). Inference in the literature is based on population surveys or lab experiments. No carbon tax has so far been accepted at the ballot box. In November 2018, for the second time in two years, the population of Washington State will be asked to vote on a carbon tax proposal (Initiative 1631). We plan to combine a population survey, measuring people’s preferences for the 2016 and 2018 carbon tax designs, with field interventions, providing valuable information about the functioning of carbon taxes and the design of Initiative 1631 to a randomly selected number of precincts.
External Link(s)
Registration Citation
Carattini, Stefano et al. 2018. "Climate Levy, Informational Messages, Acceptability, and Treatment Effects of Framing Effectiveness and Earmarking (CLIMATEFEE)." AEA RCT Registry. October 23.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
Potential voters in randomly-selected precincts in Washington State are exposed to informational ads providing details about the functioning of carbon taxes and the design of Initiative 1631, the proposal on the ballot box in 2018. The 2016 and 2018 designs are also compared. Potential voters clicking on the ads are asked to fill a survey. The survey to which potential voters are redirected provide information consistent with the ad that they were exposed to. The experiment will be supplemented with data from a statewide survey of voters, administered by a professional marketing company. This will provide additional evidence, based on stated preferences and reported behavior, and will also allow us to test the external validity of the field experiment.

A statewide survey of voters, administered by a professional marketing company, is used to provide additional evidence, based on stated preferences and reported behavior, and to test the external validity of the main intervention.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Clicks on the advertisements; stated yes-no votes for individuals recruited via the ads, stated yes-no votes for individuals recruited by the professional marketing company, actual participation and voting behavior in treated (versus control, versus rest of the State) precincts. Yes-no votes are measured conditional on (expected) participation.
Primary Outcomes (explanation)
Potential voters are exposed to different messages. Clicks on the advertisements, to collect additional information, represent a first outcome (revealed preferences). Intentions to vote, as provided in the surveys, represent an additional outcome. Actual voting patterns, as measured at the precinct level in administrative data, provide another set of primary outcomes.
Secondary Outcomes
Secondary Outcomes (end points)
Additional variables, as collected through Facebook and the online surveys, related to preferences towards carbon taxes and environmental policy in general, leveraging also the different designs between Initiative 1631, on the ballot box this year, and Initiative 732, on the ballot box in 2016.
Secondary Outcomes (explanation)
On top of analyzing the effect of randomized informational treatments, the objective of this research design is also to improve our understanding of people’s preferences for carbon taxes and how different policy designs may match such preferences.
Experimental Design
Experimental Design
Potential voters will be approached through online advertisement. Potential voters will be exposed to different informational treatments about carbon taxes and Initiative 1631. Information consistent with the treatments in the ads will be provided on the surveys, for the individuals who click on the online ads. For these individuals, additional variables will be gathered through an online survey. To supplement this, an identical online survey will be administered, by a professional marketing firm, to a representative sample of Washington State voters. Follow-up surveys may be administered after the ballot.
Experimental Design Details
Randomization Method
Randomization of potential viewers among the two treatments and the baseline groups, done by the researchers or the marketing firm administering the survey to the representative sample.
Randomization Unit
Precinct; individual in the survey administered by the professional marketing firm
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
We are target precincts and we observe data at both the precinct and the individual level.
Sample size: planned number of observations
1,000 observations in the representative survey administered by the marketing firm. Randomization across hundreds of precincts in the State of Washington, with more than 7,000 precincts overall to provide additional sources of comparison. Online advertisement campaign reaching out to potentially 500,000 individuals (order of magnitude). Clicks as a fraction of reach. Participation in the second survey as a fraction of clicks.
Sample size (or number of clusters) by treatment arms
Equal weight is given to each group.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB Name
Institutional Review Board (IRB), Appalachian State University
IRB Approval Date
IRB Approval Number
Post Trial Information
Study Withdrawal
Is the intervention completed?
Is data collection complete?
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers