Parties, Opinions and Actions: COVID-19 and Climate Change in Germany

There will be two online survey experiments conducted at the same time. In one of them, the respondents get to read a statement that is either labeled to be from a specific party or not labeled at all and is either in favor of or against more environmental and climate protection during the COVID-19 crisis. In the other survey experiment, the respondents get to read statements from all parties that are either all in favor of, all against or have mixed opinions on more environmental and climate protection during the COVID-19 crisis.

In the first survey experiment, the respondent will be provided with a single statement that she is supposed to read. There are either two or three randomization steps that lead to the specific statement each respondent gets. First, it is determined whether the statement is to be labeled as being the prevailing opinion of one specific party or if it is just an assessment of the urgency to take action protecting the environment and climate during the COVID-19 crisis. Next, if the statement is supposed to be labeled as a party statement, it is randomly determined which party out of the seven parties in the German parliament should appear. Finally, in both cases, i.e. if there was a party label assigned or not, it is randomly specified what the statement is supposed to say. There are two possible statements, one that is in favor of more action to be taken protecting the environment and climate during the current pandemic, i.e. the pro statement, and one that is against this, i.e. the con statement. Thus, there are in total 16 different possibilities for the statement and framing text the respondent can be provided with.

In the second survey experiment, the respondent is provided with an overview of seven statements one from each of the seven parties in the German parliament, respectively. For each party two different statements could potentially be shown in the overview. Similar to the first study, the pro statement is in favor of more action to be taken protecting the environment and climate during the current pandemic and the con statement is against this. There is one randomization step that determines for each party whether the pro or con statement is to be read by the respondent. In this randomization step one of three possible treatments is selected. The first one would be an all party consensus to take less action during the COVID-19 crisis meaning that for every party the con statement is shown to the respondent. The second and opposite case would be the all party consensus to take more action, i.e. the pro statement is shown for all parties. Finally, in the party disagreement treatment, four parties are shown with the pro and three with the con statement. The state of the statements is predetermined taking into account the actual opinion each party would most likely have at the moment and always stays the same for this treatment.

2021-04-20

2021-05-04

There are two main sets of outcomes that measure the importance the respondent attaches to environment and climate protection. The two sets are each composed of two questions concerning either the action taken by the German government or oneself to protect the environment and the climate in terms of the adequacy and urgency of these actions at the moment. Each of these four questions will be used separately to construct four dependent variables. Additionally, a question concerning the donation decision of the respondents is also used to construct another dependent variable.

The dependent variables are composed of the differences within the answers to each of these five questions before and after the treatments. Thus, there are five dependent variables.

The respondents have to fill out an online survey. For both survey experiments, they first have to answer some general questions on personal data and knowledge and believes about politics and the environment. After that they have to answer the questions on the outcome variables, then get to read the statements described above and answer the questions on the outcome variables again.

All respondents first have to answer a set of questions on their personal data, political orientation and knowledge, pro-environmental behavior, climate change knowledge and environmental concern which will later on be used as control variables in the analysis. This is followed by the questions on the outcome variables described above. Next, the respondent is provided with the statement (in the first experiment) or statements (in the second experiment) that was/were chosen randomly as described in the intervention description. In a last step the respondent again has to answer the five outcome questions.

Randomization is done by a computer, i.e. via the online surveying platform Qualtrics.

Randomization is done at individual level.

No

3000 respondents in total, with about 2530 partaking in the first and 470 in the second survey experiment.

3000 respondents in total, with about 2530 partaking in the first and 470 in the second survey experiment.

About 160 per treatment group, with 16 groups ((7 party labels/no label x pro/con statement) in the first and 3 groups (pro consensus, con consensus, overall disagreement) in the second survey experiment.

This paper provides insight into the impact of party cues on the public’s desire for climate protection. In particular, the effects of cues from one or multiple parties as well as the mechanisms behind these effects are analyzed. Utilizing the case of Germany’s multi-party system, two online survey experiments with a representative sample of the German voting population are conducted. Despite finding rather small effect sizes overall, results show that a party statement in favor of more climate protection is effective in changing participants’ opinions towards the same direction. People appear to be even more impressionable when they receive unexpected cues or are lead to believe that all parties work together to fight climate change. Finally, respondents that do not care about or oppose climate protection are more easily persuaded.

Stöhr, Valentina, Climate Protection in Germany: Party Cues in a Multi-Party System. Available at SSRN: https://ssrn.com/abstract=4330958 or http://dx.doi.org/10.2139/ssrn.4330958

http://dx.doi.org/10.2139/ssrn.4330958