Can the court bridge the gap? Public perception of economic vs. generational inequality in climate change mitigation policies.

Between-subjects survey experiment with vignettes, in which respondents are primed to consider economic vs. generational inequality concerns relating to climate change mitigation policies.

We implemented a 2*4 between-subject experiment research design. Each
group of respondents was shown a vignette consisting of three main parts: the first part of
the vignette presented the respondent with a graphical representation of the rise in the Earth’s
temperature. This is included with the aim to bring the severity of climate change to the
respondent’s mind, without biasing the topic in any inequality-related direction by including
text that would otherwise explain related consequences of climate change. After the image,
the respondent is given the following factual text: ”In order to reduce CO2 emissions and
thus combat climate change, the German federal government has decided to set a price per
ton of CO2. This is currently €30 and is to be increased to €55 by 2025. This means that
many everyday expenditures will gradually become more expensive.”. The second part of the
vignette introduced the first experimental variation (the court treatment) in two dimensions.
Respondents were either assigned to a treatment where we highlighted the court’s ruling, orto a control group with no further information (with equal probability): ”Germany’s highest
court, the Federal Constitutional Court, ruled in 2021 that the state is obliged to take more
comprehensive climate protection measures.”. The third part of the vignette introduced the
second experimental variation - the inequality treatment. Before being randomly assigned into
the four dimensions, all four groups receive the additional information that: ”A reduction in
CO2 emissions is necessary if living conditions on our planet are not to deteriorate further.”.
This additional information equally highlights the necessity for a reduction in CO2 emissions
across the groups. The aim of this is to make the treatment groups somewhat more comparable,
so that the emphasis is laid rather on the need for climate change action, than on the potential
inequalities.3 The four randomization groups: 1) a generational inequality treatment group
which highlighted how the carbon taxation policy would benefit future generations (”Thus, the
additional costs incurred now will benefit younger as well as future generations.”); 2) an economic
inequality treatment group, highlighting the likely increase in income inequality following such
carbon taxation policies (”However, people with low incomes have to spend a larger share of
their income on CO2-taxed goods and are thus burdened more than those with high incomes.”);
3) an economic and generational inequality treatment group that combines the two primes from
the first two dimensions. This group is further randomized so that half of the group receives the
economic inequality prime, and then the generational inequality prime, and the second group
receives the same information in the reversed order. The aim of this combined group is to
combat the problem of our rather incomparable treatment groups (generational and economic
inequality). Lastly, dimension 4) is a control group that receives no further priming information.

2022-09-01

2022-10-31

After the survey experiment, respondents are first asked: ”Do you support such a carbon taxation
policy?”. This is on a Likert-scale from 1 (”Do not agree at all”) to 5 (”Fully agree”). In order
to check whether respondents’ disagreement with the policy might be due to it not aiding enough
in climate change mitigation, we additionally ask respondents: ”Do you think the suggested CO2
taxation is:” with the answer options ”Too high”, ”exactly right”, ”too low”, and lastly ”Don’t
know”

We implemented a 2*4 between-subject experiment research design. Each
group of respondents was shown a vignette consisting of three main parts: 1) an introduction text about a current climate change mitigation policy in Germany, 2) a randomisation into an elite-cue treatment group (vs. control group with no extra information), 3) a randomisation into one of 4 treatment groups (3 inequality treatment groups and one control group).

Randomisation by computer.

Randomization at the individual level.

No

Estimated to be 6000 individuals.

6000 individuals.

6000 individuals.