1) Have any data been collected for this study already? No, no data have been collected for this study yet.
2) What's the main question being asked or hypothesis being tested in this study? 1) Are workers (mis)informed about the future employment impacts of decarbonization?
2) Does providing respondents with expert forecasts regarding the employment implications of the energy transition affect their support for the energy transition?
3) Describe the key dependent variable(s) specifying how they will be measured. 1. Self-reported support for the energy transition will be measured through an index based on the following questions:
- Do you agree or disagree with the following statement: "The U.S. government should produce less energy from fossil fuel sources such as oil, coal and natural gas and make more energy from renewable sources such as wind and solar". (strongly disagree/somewhat disagree/ neither disagree nor agree/ somewhat agree/ strongly agree)
- Do you support or oppose the following policies? (strongly oppose/somewhat oppose/ neither oppose nor support/ somewhat support/ strongly support)
- Subsidies or government funding for technologies that reduce carbon, like carbon capture and storage, or renewable energy
- Increasing government funding for green infrastructure programs like public transportation, renovating buildings, and building renewable energy facilities
- A tax on pollution or carbon emissions that companies produce (a carbon tax)
- A cap-and-trade system: a limit on total emissions which lets companies trade emission allowances. Each company is allowed to pollute up to a limit. If they pollute more, they must buy more permits. If they pollute less, they can sell their extra permits to others.
- A rule that requires companies to install specific technologies to reduce emissions
2) Trade-off question: We will build a dummy variable based on the following question:
With which one of these statements about jobs and the energy transition do you most agree?
-The energy transition should be prioritized, even at the risk of displacing some workers.
-Protecting workers should be prioritized, even if it means the energy transition moves more slowly.
3) Petition:
- We will ask respondents whether they are willing to sign two petitions, one asking the government to pass policies that advance the energy transition and one asking the government to slow down the passage of energy transition policies. We will create a dummy variable from each answer.
4) How many and which conditions will participants be assigned to? Participants are randomly allocated to one of two treatment groups.
Group 1 – Treatment OECD: Receives information about an expert (OECD) forecast on employment under energy transition policies in 2030.
Group 2 – Control: Does not receive any information about expert employment forecasts before the main outcomes are measured.
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. Main Specification:
We first estimate the following equation: Yi = α + β1Treat_i + ATX_i + e_i , where
Yi is the outcome of interest, either policy preferences or views towards the energy transition.
Treat is an indicator for the treatment group respondent i was allocated to.
Xi is a vector of individual socio-demographic characteristics. We also report results without any controls. e_i is an individual-specific error term.
Our coefficient of interest is β1, which measures the impact of the information treatment.
Then we exploit heterogeneity in the pre-treatment estimation error. The estimation error will be calculated as the difference between the respondent's beliefs about the number of jobs under the energy transition scenario in 2030 and the objective expert estimate.
In particular, we estimate the following:
Yi = α + β1EstError_i + β2Treat_i+ γTreat_i ×EstError_i + ATX_i + e_i
where EstError_i is a measure of the individual's estimation error. A negative value means that individuals underestimated the number of overall jobs relative to the expert estimate. A positive value means that individuals overestimated the number of overall jobs relative to the expert estimate.
Our coefficient of interest is γ, which measures the effect of the information treatment as a function of individual's initial estimation error.
We will also re-estimate the above equation replacing EstError_i with Overestimator_i a dummy variable taking value 1 for respondents that overestimate the number of jobs in the scenario with decarbonization policies (and 0 otherwise).
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. We will drop observations with nonresponse, those that fail attention checks, those that answer open ended questions with nonsensical responses, those that rush through the survey, and extreme data points in response to question about the number of jobs expected (e.g., trimming at the 1st and 99th percentile).
7) How many observations will be collected or what will determine sample size? No need to justify decision, but be precise about exactly how the number will be determined. We plan to recruit 3,000 to 4,000 employed workers in each country. The exact number of participants will depend on the exact response rate of panelists invited to our study. Our target sample size is based on the provider's best estimate.
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?) We will run descriptive analyses on the pre-treatment knowledge of the energy transition and its impact on jobs and support for policies addressing the consequences of job losses. We will assess perceptions, identify whether knowledge gaps are prevalent, and which socio-demographic characteristics correlate with them. We will also examine whether the energy transition treatment impacts respondents' job prospects (own occupation and regional employment) and climate change attitudes.
We plan on studying heterogeneity of impacts based on vulnerability to the energy transition and climate change impacts (considering region, occupation and industry, and other socioeconomic variables) as well as across ideology and country. This will be done via regressions interacting the treatment with measures of these characteristics.
