Risk as Excuses for Selfishness in Social Decisions: Evidence from a Laboratory Experiment

How to cite

DONG, Wanxin; Zheng, Jiakun. 2021. "Risk as Excuses for Selfishness in Social Decisions: Evidence from a Laboratory Experiment." AEA RCT Registry. AEARCTR-0008697. https://doi.org/10.1257/rct.8697-2.0

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

Wanxin DONG* , Jiakun Zheng

* Primary investigator

Registry ID
AEARCTR-0008697
DOI
https://doi.org/10.1257/rct.8697-2.0
First registered
Last updated
Status
In development
Trial dates
2022-02-27 – 2022-04-30
Countries
Private
JEL codes
D71, D80, D91
Registry record
http://www.socialscienceregistry.org/trials/8697

Abstract

Exley (2016) has examined charitable giving behavior under risk, i.e., how much one is willing to improve others’ welfare while sacrificing one’s own welfare. She has found that individuals use risk as excuses not to give when there is self-partner tradeoff. People become more averse to partners’ risk and less averse to self risk. However, many real-life situations are about how much one is ready to benefit from sacrificing others’ welfare or to sacrifice one’s own welfare to benefit others. We investigate the role of risk in these types of situations. In our laboratory experiment, participants make a series of decisions between risky and riskless gains or losses for themselves or their partners. Their altered risk responses between no self-partner tradeoff and self-partner tradeoff will reflect the existence of excuse-driven risk preferences. We expect a reverse pattern of individual risk attitudes in the loss domain, i.e., more seeking to partners’ risk and less seeking to self risk. Moreover, individuals tend to exhibit more excuse-driven risk preferences in the loss domain.

Intervention

We measure excuse-driven risk preferences in the gain domain and loss domain. Considering the setting of payoffs, we adopt a between-subject design. In each experiment, there are four blocks of decision tasks including 30 multiple price lists
and a moral wiggle room including three decision games. Price lists in each block only differ in 7 different probabilities, i.e. {95%, 90%, 75%, 50%, 25%, 10%, 5%}, and four blocks are presented in a random order for each participant.

Experimental design

Our experimental design is mainly based on Exley (2016)’s methods, but we modify and apply in the gain domain and loss domain. We adopt a between-subject design. Participants are required to complete thirty price lists respectively in the gain and loss experiment. Thirty price lists include one “normalization” price list, one buffer price list and 28 valuation price lists. Each price list consists of a series of binary decisions. Through these price lists, we could measure individuals' responses to different risk situations.
In each experiment, participants first complete a normalization price list that determines partner-specific X values. Partner-specific X values are used later to determine the non-zero gains or losses of partner lotteries and partner certain amounts in valuation price lists. The second one is a buffer price list similar to the normalization one. After completing the first two price lists, participants complete 28 valuation price lists, which are used to estimate individuals’ lottery valuations. In this part, participants face four types of decision tasks respectively in the gain and loss experiment – {self gain lottery, partner gain lottery} X {self gain certain amount, partner gain certain amount} and {self loss lottery, partner loss lottery} X {self loss certain amount, partner loss certain amount}.
After participants complete the price lists, they need answer three moral wiggle room questions. The first one is a choice-to-reveal question, the second one is a revealed-unaligned-state question, and the third one is a revealed-aligned-state question. Through moral wiggle room, we plan to examine the correlation between excuse-driven risk preferences and excuse-driven types of participants.

Design details

Randomization method
randomization done by a computer.
Randomization unit
individual.
Clusters
100 undergraduate students from Renmin University of China in each of the gain and loss experiments, that is 200 students in total
Observations
100 undergraduate students from Renmin University of China in each of the gain and loss experiments, that is 200 students in total
Arms
Due to the between-subject design, 100 subjects in the sample make the gain experiment, and 100 subjects make the loss experiment.
Primary outcomes
- partner-specific values in the gain and loss domain respectively - participants' valuations of different lotteries - excuse-driven types of participants

Cited by

No citation data available for this registration yet.

Keywords