How regional predictions of heat events in the coming summer and their forward-attribution to climate change affect support for adaptation and mitigation measures as well as attribution of guilt across generations.

Last registered on June 24, 2024
View Trial History

Pre-Trial

Trial Information

General Information

Title
How regional predictions of heat events in the coming summer and their forward-attribution to climate change affect support for adaptation and mitigation measures as well as attribution of guilt across generations.
RCT ID
AEARCTR-0013668
Initial registration date
June 06, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
June 24, 2024, 12:20 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country
Germany
Region

Primary Investigator

Name
Enrico Longo
Affiliation
University of Hamburg
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Mrunali Damania
PI Affiliation
Univesity of Hamburg
Contact Investigator
PI Name
Andreas Lange
PI Affiliation
Univesity of Hamburg
Contact Investigator
PI Name
Grischa Perino
PI Affiliation
Univesity of Hamburg
Contact Investigator
PI Name
Moritz Drupp
PI Affiliation
Univesity of Hamburg
Contact Investigator
PI Name
Johanna Baehr
PI Affiliation
Univesity of Hamburg
Contact Investigator

Additional Trial Information

Status
Completed
Start date
2024-06-06
End date
2024-06-17
Keywords
Behavior, Environment & Energy
Additional Keywords
Climate Change, Adaptation, Mitigation, Attribution
JEL code(s)
Q54, Q58, D91
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
How do regional predictions of the number of heat events (tropical nights) for the coming summer (June to August 2024) impact citizens’ willingness to engage in voluntary adaptation efforts, support for mitigation measures and the attribution of blame to the currently elder generation? How is this impact affected if the anthropogenic contribution in the predicted number of heat events is made transparent exa nte (´forward attribution´)? We study these fundamental questions with a large-scale survey experiment that involves 4,200 adult members from the general population in Germany.

There are three experimental conditions: Control, Prediction, and 'Prediction with Forward Attribution' ('Attribution' for short). All participants are informed about the health threats associated to heat events and tropic nights in particular. The Prediction and the Attribution groups receive regional (100km x 100km) predictions of the number of heat events (nights when temperature does not drop below 20°C, 75percentile from an ensemble of predictions with 30 members). The Attribution group on top of that learns the number of heat events (defined the same way as above) to be expected in a typical year (such as 1850) in a world without anthropogenic climate change. Predictions presented are conditioned on participants' postcodes.

Prior to treatment, we elicit core socio-demographic variables, post codes, attitudes to climate change, science and government as well as perceptions of the frequency of local summer heat events and holiday plans. Post treatment, we elicit emotional reactions to treatment, several core outcome variable (stated and revealed preference) on attribution of blame across generations, support for adaptation and mitigation measures. The survey concludes with further socio-demographic and attitutional questions.

In a second wave to be conducted in September 2024 we want to assess the impact that experienced weather has had on the same individuals in terms of support for adaptation and mitigation measures.
External Link(s)

Registration Citation

Citation
Baehr, Johanna et al. 2024. "How regional predictions of heat events in the coming summer and their forward-attribution to climate change affect support for adaptation and mitigation measures as well as attribution of guilt across generations.." AEA RCT Registry. June 24. https://doi.org/10.1257/rct.13668-1.0
Sponsors & Partners

Sponsors

Experimental Details

Interventions

Intervention(s)
Intervention 1: Prediction
We provide participants with a prediction for the number of heat events (75percentile of number of tropical nights, i.e. with temperatures always above 20°C, from an ensemble of predictions with 30 members) during the months June to August 2024. They are computed by research in our team (University of Hamburg) in cooperation with Germany's National Meteorological Service (Deutscher Wetterdienst). Predictions are region-specific (100km x 100km) and presented to participants based on the postcode of their home address elicited at the beginning of the survey.

Intervention 2: Prediction + Forward Attribution ('Attribution' for short)
Same as intervention 1 plus the prediction for the number of heat events (75percentile of number of tropical nights, i.e. with temperatures always above 20°C) during the months June to August in a typical summer without anthropogenic climate change (i.e. for the year 1850) as a reference point.
Intervention Start Date
2024-06-06
Intervention End Date
2024-06-17

Primary Outcomes

Primary Outcomes (end points)
1. Attribution of blame and financial burden between generations (two stated preference questions)
2. Support for adaptation measures (two revealed preference measures, one for protectinng oneself against heat, one to protect vulnerable groups against heat)
3. Support for mitigation measure (one revealed preference measure)
Primary Outcomes (explanation)
Attribution of blame and financial burden between generations (two stated preference questions)
a) How much is the generation of those 60 years or older to be blamed for global climate change? 5-point Likert scale from 'not at all' to 'very much'
b) Should the level of pension entitlements of current pensioners be reduced in order to finance climate protection? 5-point Likert scale from 'not, under no circumstances' to 'yes, under all circumstances'

Support for adaptation measures (two revealed preference measures, one for protectinng oneself against heat, one to protect vulnerable groups against heat)
a) Participation in a lottery (chances of winning 1:200). Participants have to choose between a given monetary prize (randomly drawn from values between 10€ and 100€ in steps of 10€) and a state-of-the-art air fan suitable for bedrooms with a market value of more than 100€. (revealed preference)
b) Participation in a lottery (chances of winning 1:200). Participants have to choose how much of a monetary prize of 100€ (in steps of 10€) they would donate to buy reusable water bottles distributed by a charity to homeless people in Germany. (revealed preference)

Support for mitigation measures (one revealed preference measure)
Vote between the purchase and permanent retirement of 10 emission rights from the European Union Emission Trading System (i.e. 10 tons of CO2) and 50cents. Vote among all participants of a treatment. Yes/No/Abstain decision (revealed preference)

For more details, see questionnaire.





Secondary Outcomes

Secondary Outcomes (end points)
To identify channels of transmission, we measure the extent to which participants perceive the forecast reliable, ability to summarize the content of the information provided, and the emotions evoked when reading the forecast/intervention.
Also of interest are participants guesses on the number of heat events in the summer 2023 and 2025. The former allows to identify whether the prediction provided shifts beliefs upwards or downwards. The latter, whether they expect more or less heat events in the future.
We also ask prior to treatments, how likely is it that participants will buy a new fan for summer 2024. 5-point Likert scale from 'definitively not' to 'definitively yes' (stated preference) This serves as a reference point for the revealed-preference primary outcome fan lottery.
We also expect that gender and the number of days of planned holidays and whether participants travel to hotter or cooler places to affect responses to the predictions.
Secondary Outcomes (explanation)
The perceived number of heat events in the summer 2023 serves as a proxy for participants' beliefs about the expected heat events in 2024. In the Prediction and Attribution treatments, this belief is updated based on the predictions provided. The direction of belief updates (predicted heat events vs. remembered heat events and assuming them to be zero for the Control group) will be used as explanatory variable.

Experimental Design

Experimental Design
We run online surveys in Germany. We recruit participants through a professional survey company (Bilendi). To be eligible subjects must be adults (age 18+). The sample will use quotas for age brackets, gender and eduction (with and without a highschool diploma 'Abitur') to match the distribution in the general population. Moreover, we will exclude participants that complete the survey too quickly to have read the information provided and answers (<5min in total) or those showing clear signs of inattention (response times >60min in total).
All subjects answers questions about demographics and prior beliefs and attitudes towards climate change, climate policy and climate science.
Pre-treatment attitudes towards climate change.
Subjects then, in both the 'Prediction' and in the 'Attribution' treatments, receive a forecast on the number of heat events (days the temperature does not fall below 20°C, i.e. 'tropical nights') for the three summer months of June to August 2024.
All subjects also indicate how reliable they find the prediction, how they feel when they read it and they are asked to write down briefly how they would explain the prediction to a friend.
Only in the “Attribution” treatment, in addition to the prediction described above, participants also receive information on the number of heat events / tropical nights that would be expected in a typical summer without human-induced climate change.
Subjects in the “Attribution” treatment then indicate how reliable they find this second comparative information, how they feel when they read it and they are asked to write down briefly how they would explain it to a friend.
All subjects then answer questions about the key revealed and stated preference outcome variables on attribution of blame between generations, adaptation measures for oneself and for vulnerable groups, and a mitigation policy (details see above).
Finally, socio-demographic characteristics are collected, such as family composition, income and political attitudes.

Experimental Design Details
Hypotheses

Prediction vs Control
main
1. Support for adaptation and mitigation measures is increasing in the change in beliefs about the number of heat events induced in the PredictionOnly treatment using the Control group as a reference point.

supplementary
2. Treatment effects on fans are higher for women and for those expecting more heat in 2025. (women: because one quote says, that women have a higher risk of dying during heat events)


PredictionOnly vs Attribution
main
3. In the Attribution treatment more blame and financial burden is placed on the elder generation.
4. The propensity to buy and willingness-to-pay for a fan does not differ between treatments.
5. The donation rate for the protection of vulnerable people against heat and support for mitigation policies is higher in the Attribution treatment.

supplementary
6. Treatment effects are stronger among those comprehending the information provided and mediated by the emotional response to the attribution information.
Randomization Method
The randomization into the three experimental groups of equal size (planned N1 = N2 = N3 = 1,400) is done by the survey company by a computer. Ranomization of cash amounts in the willingness-to-pay for an air fan question is done by the survey company by computer (all amount with equal probability).
The randomization of winners of the lottery (one of the primary outcome variables) is done by the researchers (by computer) after completion of data collection.
Randomization Unit
We randomize individual participants in one control group and two treatment groups. All groups have the same a priori probability (planned N1 = N2 = N3 = 1,400).
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
there will be no clustering
Sample size: planned number of observations
4,200 individuals
Sample size (or number of clusters) by treatment arms
1,400 individuals
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials

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

Request Information
IRB

Institutional Review Boards (IRBs)

IRB Name
Ethics committee for the Faculty of Business, Economics and Social Sciences at Universität Hamburg
IRB Approval Date
2024-05-17
IRB Approval Number
2024-015
Analysis Plan

Post-Trial

Post Trial Information

Study Withdrawal

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

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials