Carbon information, pricing, and bans. Evidence from a field experiment

All students and employees of HEC Paris get an invitation to participate in the survey about the HEC Paris community’s attitudes toward different policies that could reduce carbon emissions associated with food consumption at HEC's canteen.

2023-12-11

2024-02-10

The outcomes of the experiment are individuals' survey responses to a set of questions. Of primary interest are: 1) the choice among "do nothing", "information", "pricing", "ban" policies from policymaker's perspective; 2) the choice among "do nothing", "carbon footrpint information", "carbon pricing", "high-carbon footprint dish supply reduction" in referendum setting; 3) the choice among "no info" vs. "info" from individual perspective; 4) the choice among "carbon pricing" or "carbon ban" conditional on requirement to implement one of them.

In a prior field experiment, we tested the effectiveness of three policies aimed at the reduction of food carbon footprint: 1) provision of clear and reliable information about the carbon footprint of dishes offered on the menu; 2) adjusting dish prices with a pricing system where dish prices and the carbon footprint of the dishes are positively correlated; 3) regulating supply by replacing high-carbon dishes with equally nourishing low-carbon dishes. The information provision treatment was introduced on 21 November 2022 and lasted until 5 May 2023. The pricing treatment was conducted in four non-consecutive treatments within the information phase between 13 March 2023 and 21 April 2023.
For the supply regulation treatment, we exploit the no-red-meat Thursday policy implemented by the HEC administration on 9 September 2022 which lasted throughout the academic year 2022-2023.

In the follow-up survey that will be launched on 11 December 2023, all students and employees of HEC Paris will get an invitation to participate in the survey about the HEC Paris community’s attitudes toward different policies that could reduce carbon emissions associated with food consumption at HEC's canteen.

Not available

Randomization is done by the randomization algorithm of Qualtrics Survey Software.

The unit of randomization is individual (student or employee of HEC Paris who participates in the survey).

No

Since the treatment is administered at the individual level and each individual receives only one survey, the number of clusters is equal to the number of observations (see below).

The survey will be sent to all students and employees of HEC Paris who were enrolled at / affiliated with HEC Paris during the academic years 2022-2023 and 2023-2024 (as of November 2023). This constitutes about 4000 individuals. Assuming a response rate of 40%, the final sample will consist of about 1600 individuals (variation in response rate might lead to a different exact number of final observations).

50% of the survey respondents will be assigned to the "policymaker" treatment, while the other 50% will be assigned to the "referendum" treatment. Taking the assumption made above about a 40% response rate and a total of 4000 individuals being offered to participate in the survey, we obtain approximately 800 individuals in the "policymaker" treatment, and 800 individuals in the "referendum" treatment.