Lottery incentives and carbon pricing rebates

How to cite

Drichoutis, Andreas; Palma, Marco. 2025. "Lottery incentives and carbon pricing rebates." AEA RCT Registry. AEARCTR-0015002. https://doi.org/10.1257/rct.15002-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Andreas Drichoutis* , Marco Palma

* Primary investigator

Registry ID
AEARCTR-0015002
DOI
https://doi.org/10.1257/rct.15002-1.0
First registered
Last updated
Status
In development
Trial dates
2025-01-10 – 2025-02-28
Countries
United States of America ()
JEL codes
Q48, Q50, Q58, C90
Registry record
http://www.socialscienceregistry.org/trials/15002

Abstract

Carbon pricing implementation is gaining attention in the literature as an instrument to reduce carbon emissions to meet environmental targets (Narassimhan et al. 2018; Byce 2018). There are several market systems that have been proposed with carbon pricing or using a market-based approach to assign a price on carbon dioxide (CO2) emissions as a way to reduce greenhouse emissions. The implementation of such a system depends on whether there is sufficient public support for it. A key policy-relevant question for carbon pricing is whether using a carbon pricing rebate would increase public support. One of the potential drawbacks of such a policy is that it is expensive to implement as it requires to return a sizable portion of the amount of carbon price tax collected.

Recent literature in behavioral and experimental economics proposes that consumers' behavior towards a lottery is similar to a certain outcome if the level of the lottery is reasonably attainable. More specifically, Ahles, Palma and Drichoutis (2024) find that the incentives produced by a 10% lottery, produce the same outcomes in an incentivized real-market than a 100% realization. In this study, we implement lottery rebates to observe if the consumer reaction from these lottery incentives produces similar benefits than using the full rebate scheme. This is very useful as a policy instrument as implementing a rebate with 10% likelihood of implementation results in a significant reduction of the implementation cost of such a policy.

References:
Ahles, Amelia, Marco A. Palma, and Andreas C. Drichoutis. "Testing the effectiveness of lottery incentives in online experiments." American Journal of Agricultural Economics (2024).

Boyce, James K. "Carbon pricing: effectiveness and equity." Ecological Economics 150 (2018): 52-61.

Narassimhan, Easwaran, Kelly S. Gallagher, Stefan Koester, and Julio Rivera Alejo. "Carbon pricing in practice: A review of existing emissions trading systems." Climate Policy 18, no. 8 (2018): 967-991.

Intervention

The study is composed of two parts.

*Part 1*: Subjects will participate in an open-ended incentivized choice experiment (Corrigan et al. 2009) akin to a multiple price list task where they’ll have to indicate how many units of a fictitious good they’d like to purchase. Subjects will go through a list of 10 prices and for each price they will indicate how many units of the good they want given that the first unit gives them a value of 4, the second a value of 3 and the third a value of 2. The payoff associated with a price level and units is the sum of the values of the units purchased minus the price of the units. The prices are progressively increasing in a way that a person should choose to buy 3 units for low prices, 2 units for medium prices, 1 unit for higher prices and no units for very high prices. Subjects will be told that researchers have committed upon completion of the study on buying pollution rights via Compensators (https://www.compensators.org/) a charitable non-profit organization with the purpose of reducing surplus carbon emission allowances, and permanently remove them from the market. By doing so industries are forced to reduce their CO₂ emissions and the scarcity of emission allowances leads to price increases, which in turn makes low-emission technologies more competitive. However, each unit they purchase reduces researcher’s contribution by 100 lbs, so if a person chooses to buy all 3 units we will not buy any pollution rights. If a person chooses to purchase 2 units of the good the researchers will purchase 100 lbs of pollution rights, if they purchase 1 unit of the good the researchers will purchase 200 lbs of pollution rights and if they do not purchase any unit, the researchers will purchase all 300 lbs of pollution rights. Subjects will be told that one decision will be randomly selected for payment at the end of the experiment from Part 1.

*Part 2*: Subjects will be asked to select whether on top to any earnings from Part 1, they would like to have earnings from an additional decision added to their payoff. More specifically, subjects will be asked whether they want to be paid for (a) one randomly determined decision for low price levels (incentives are to purchase 3 units of the item for these price levels) or (b) if they’d prefer to be paid for one randomly determined decision for medium and high price levels (incentives are to purchase <3 units for these price levels).

- CONTROL treatment: Subjects go through Part 1 and Part 2 in sequence as described above.
- INTERVENTION treatment: Like the CONTROL treatment, but the following rule is applied to Part 2: If subjects select they prefer to be paid for one randomly determined decision for medium and high price levels, they receive $1 on top to all other earnings (cash back).
- HYPOTHETICAL treatment: Subjects are told that any earnings from Parts 1 and 2 are hypothetical and do not count.
- 100% treatment: Subjects are told that any earnings from Parts 1 and 2 will be paid with certainty.
- 10% treatment: Subjects are told that any earnings from Parts 1 and 2 will be paid with 10% chance.

Corrigan, Jay R., Dinah Pura T. Depositario, Rodolfo M. Nayga Jr., Ximing Wu, and Tiffany P. Laude. "Comparing Open-Ended Choice Experiments and Experimental Auctions: An Application to Golden Rice." American Journal of Agricultural Economics 91, no. 3 (2009): 837–853.

Experimental design

The study has a 2x3 design resulting in 6 cells:
1. CONTROL + HYPOTHETICAL
2. CONTROL + 100%
3. CONTROL + 10%
4. INTERVENTION + HYPOTHETICAL
5. INTERVENTION + 100%
6. INTERVENTION + 10%

Design details

Randomization method
All randomizations are performed by the software program (Qualtrics)
Randomization unit
individual
Clusters
no clusters
Observations
870 subjects
Arms
at least 145 subjects/treatment
Primary outcomes
- Binary variable of whether subjects choose option (b) in Part 2 vs. option (a). Option (b) includes price levels that maximize payoff for a lower number of units than option (a). - Number of units purchased in Part 1.

Cited by

No citation data available for this registration yet.

Keywords