Experimental Design Details
Each participant in the main experiment goes through a sequence of two tasks, only one of them being (randomly) selected for implementation / payment.
The first and main task is a real effort task where subjects successively press "a" and "b" on their keyboard as many times as possible, each pair earning one point. In treatment, each point corresponds to a pre-specified amount of CO2 being removed from the atmosphere. After a 30 seconds training period, this task lasts for 5 minutes although participants can elect to finish early.
In the main task, all participants face an increasing marginal abatement cost curve. This implies that the amount of CO2 abatement per point scored declines with the cumulative number of points achieved. Moreover, to study how alternative external interventions affect the distribution of abatement effort accross heterogenous agents, participants can be in either of two groups (randomly selected). In group A, the "low abatement cost" group, each point scored by a participant implies a relatively high abatement. In group B, the "high abatement cost", the amount of CO2 removed is eight times smaller. See hereunder the abatement schedules for both groups:
Group A (number of pairs: abatement in kg of CO2 per pair of keys)
0-200: 2
201-400: 1
401-600: 0.5
601-800: 0.25
801-1,000: 0.125
1,001+: 0.06
Group B (number of pairs: abatement in kg of CO2 per pair of keys)
0-200: 0.25
201-400: 0.125
401-600: 0.06
601-800: 0.03
801-1,000: 0.015
1,001+: 0.01
Participants are randomly assigned to either a control group with two conditions or to one of 12 experimental treatments. The latter combine six possible conditions for each abatement schedule (groups A and B).
Respondents in the control groups face a neutral frame, meaning that the information they receive does not refer to CO2. In the first control group, participants simply receive an explanation about the key-pressing task. It is explicitly said that their score does not affect their CO2 abatement nor their monetary payment. This group allows to identify the baseline willingness to exert an effort. In the second control condition, participants are incentivized with direct monetary incentives. Specifically, they receive the same information about the key-pressing task and receive additional information about the payment received for each points they score. The payment schedule is following (in cents per point):
0-200: 2.5
201-400: 1.25
401-600: 0.6
601-800: 0.3
801-1000: 0.15
1001+: 0.1
This condition allows to quantify how effort is affected to monetary incentives in a neutral frame.
Each of the following 12 conditions (6 conditions per abatement schedule) refers to CO2.
In the first treatment condition, named "Voluntary CO2 abatement", participants receive information about the key-pressing task and additional information about the CO2 abatement schedule (either group A or B). We also explain that they will receive a proof of CO2 certificates corresponding to the abatement effort achieved in the experiment. In addition, we provide some basic background information about CO2 emissions in Switzerland.
The second treatment condition is named "Moral suasion". In this condition, participants receive information about the key-pressing task and CO2 emissions (as in condition voluntary CO2 abatement). Additional text motivates abatement effort by calling upon moral justice. This condition is intended to affect intrinsic motivation for voluntary abatement effort. The message explicitly links carbon abatement, moral duty and climate justice.
The following treatment condition is called "Monetary incentives". In this treatment, participants receive the same information as in the voluntary CO2 abatement condition, and we incentivize CO2 abatement using two different carbon prices. First, we consider CHF12.5 per tonne CO2, which gives rise to following per point payment schedule (in cents per point):
Group A / Group B:
0-200: 2.5 / 0.3
201-400: 1.25 / 0.15
401-600: 0.6 / 0.1
601-800: 0.3 / 0.05
801-1000: 0.15 / 0.02
1001+: 0.1 / 0.01
The second price of CO2 is set at CHF100 per tonne CO2, which gives rise to a second payment schedule (in cents per point):
Group A / Group B:
0-200: 20 / 2.5
201-400: 10 / 1.25
401-600: 5 / 0.6
601-800: 2.5 / 0.3
801-1000: 1.25 / 0.15
1001+: 0.6 / 0.1
The last two treatment conditions combine the textual message incentivizing moral behavior and one of the two financial incentives of the monetary incentives condition. This treatment allows to test whether monetary incentives crowd-out intrinsic motivation activated with the moral suasion message.
A third task, applied to a separate, determines the perceived moral duty associated with CO2 abatement. In this task, subjects are asked to rate the choice of choosing CHF100 over abating 0.5 tonne of CO2 as "socially acceptable and consistent with moral and appropriate behavior" from "socially inacceptable and inconsistent with moral and appropriate behavior". Importantly, respondents are asked to rate what other subjects would think about the statement rather than own preferences. Participants who match the modal response elicited in the pilot receive CHF 10.