Intervention (Hidden)
Description of research question, experimental design, sample, and ex-ante hypothesis:
We conduct a survey and experiment with individuals who have stock-market experience to study the effect of uncertainty about future CO2-prices on green investment decisions. Prior to our main survey, we surveyed experts about their expectations about future CO2-prices. The participants of our main survey are exposed to some of the results of this prior survey with experts.
All participants are informed about the experts' average/median CO2-price expectation for the year 2030. Participants in different treatment arms are exposed to the probability estimates of different groups of experts that the realized CO2-price will indeed be in a bandwith around the average expectation. The objective of the experimental part is to randomly generate high or low uncertainty about future CO2-prices (i.e., the second moment), while keeping average expectations (i.e., the first moment) constant. Our survey-experimental design is inspired by Coibion et al. (2023, ECMA), Coibion et al. (2023, AER) and Mikosch et al. (forthcoming, AEJ Macro).
We will have three randomized groups: a control group that is only exposed to the average/median expectation of the experts; group "low uncertainty" that is exposed to the average/median expectation, plus information that one group of experts believe that there is a 70% probability that the CO2-price in 2030 will be in a bandwith around the median expert expectation; group "high uncertainty" that is exposed to the average/median expectation, plus information that one group of experts believe that there is a 15% probability that the CO2-price in 2030 will be in a bandwith around the median expert expectation.
The main outcome variable is a question in which survey participants are asked to decide whether they invest an hypothetical amount of money in a conventional fund or a fund that is certified to be sustainable. We also elicit prior and posterior beliefs about CO2-prices and the likelihood that certain CO2-prices will realize, and have several questions on financial literacy and experience with green finance.
Sample:
Type of survey participants that we want in the sample: German retail investors who invest in stocks or funds and currently live in Germany
Sample size: 2500
40% female
60% male
Age group distribution: about 1/3 each for 18-35 years old, 36 – 50 years old, and 51 - 65 years old population group
Two attention checks: First one, at least select two correct answers out of three in three multiple choices. Second one, at least select two correct answers out of three in five multiple choices. If not passed, they will be screened out. The attention check questions are rephrased neutrally.
Subgroups we are particularly interested in: Low (sustainable) financial literacy and high (sustainable) financial literacy; understanding of CO2 price
Other sources of heterogeneity: Social demographic variables, such as, age, gender, region (Bundesland), education, income.
Aside from analyzing the treatment effect, we are also interested in reporting descriptives, for example with respect to the following variables: (sustainable) financial literacy, green investment behavior, sustainable investment and climate change perceptions, expectation of the CO2 price, perception of fees, returns and risk between sustainable funds and conventional funds.
Main ex-ante hypothesis:
We expect that we see less green investment if uncertainty about the 2030 CO2-price is high, relative to a situation with lower uncertainty about the 2030 CO2-price. Our ex-ante hypothesis with a clear sign (i.e., green invest|high uncertainty < green invest|low uncertainty) allows for one-sided statistical tests to compare the two main treatment groups.