Eliciting Responsibility for Negative Externalities: Theory and Experiment
How to cite
Guias, Horia. 2026. "Eliciting Responsibility for Negative Externalities: Theory and Experiment." AEA RCT Registry. AEARCTR-0019297. https://doi.org/10.1257/rct.19297-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
Do individuals remediate negative externalities out of forward-looking concern for outcomes or backward-looking feelings of personal responsibility? I design a lab experiment that disentangles these mechanisms by holding present restoration consequences constant while exogenously varying the extent to which past harm was self-caused or externally generated. By systematically varying the unit price of restoration, I identify individual willingness to pay to restore self-caused versus other-caused harm. Finally, I estimate a structural random utility model via Maximum Likelihood to characterize preference heterogeneity across selfish, purely altruistic, purely clean-slate, and altruistic types who exhibit a responsibility premium for self-caused harm.
Intervention
A participant selects a charity at the beginning of the experiment from a given list and receives instructions for the decisions she will make in this experiment. Then he faces a sequence of decision situations with varying final consequences: 7 decision situations in the first stage and 58 decision situations in the second stage (random order). The first stage serves to generate the negative externality which can then be restored in the second stage. In addition, the participant observes existing harm whose origin she does not know ex-ante.
Experimental design
Participants make incentivized choices affecting their monetary earnings and a real-world charity selected at the beginning of the experiment. Negative externalities are operationalized as the cancellation of charitable donations.
The experiment proceeds in three main parts:
Stage 1 (Harm Generation): Participants complete a series of choices between increasing personal monetary earnings and preventing negative externalities (cancelled charitable donations).
Stage 2 (Harm Restoration): Participants face choices where they can pay to remediate cancelled donations. The origin of the harm (self-caused in Stage 1 versus externally generated pre-existing harm) and the unit price of restoration are systematically varied to identify willingness to pay across harm types.
Post-Experiment Survey: Participants complete a survey eliciting personal moral norms regarding harm restoration, self-reported decision reasoning, and baseline attitudes toward the selected charity.
Design details
Cited by
No citation data available for this registration yet.