Replication of “Propensities to engage in and punish corrupt behavior: Experimental evidence from Australia, India, Indonesia and Singapore” in America

How to cite

Khan, Shayan; Michler, Jeffrey. 2024. "Replication of “Propensities to engage in and punish corrupt behavior: Experimental evidence from Australia, India, Indonesia and Singapore” in America." AEA RCT Registry. AEARCTR-0014506. https://doi.org/10.1257/rct.14506-1.0

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

Shayan Khan* , Jeffrey Michler

* Primary investigator

Registry ID
AEARCTR-0014506
DOI
https://doi.org/10.1257/rct.14506-1.0
First registered
Last updated
Status
In development
Trial dates
2024-10-21 – 2024-11-15
Countries
United States of America (Arizona)
Registry record
http://www.socialscienceregistry.org/trials/14506

Abstract

This study will be replicating “Propensities to engage in and punish corrupt behavior: Experimental evidence from Australia, India, Indonesia and Singapore” (Cameron et al., 2009) in America. The original study found support for the claim that the level of corruption of a country given by corruption indices is correlated with the willingness to engage and punish corrupt behavior of the country's citizens. So, this study will see if American citizens are more willing to engage or punish corrupt behavior and compare that to the country's corruption indices. Corruption will be modeled through a one-shot, sequential corruption game based on bribery taken from Cameron et al. (2009) where there are three players: a firm, a public official, and a citizen. The firm offers a bribe to the public official where they can accept or reject the bribe. If the bribe is rejected the game is over. But if the bribe is accepted, the citizen loses money and they will have the opportunity to punish the firm and the public official financially. We hypothesize that America will have a lower propensity to engage in corruption due to its good standing against corruption as shown in indices,

Intervention

There will be two possible treatments of either welfare-reducing or welfare-enhancing payoffs. Welfare-enhancing refers to the total payoff gains from bribery are greater than the total payoff loss. Welfare-reducing refers to the opposite of welfare-enhancing. The entire sample size will be split into these two treatments.

Experimental design

The experiment design mirrors the study Cameron et al. (2009). In the lab experiment, there will be three players per group playing a one-shot sequential game that focuses on bribery, a common occurrence of corruption. These groups will be divided between welfare-enhancing and welfare-reducing payoffs.The game starts with Player 1 who acts as a firm. They can choose to offer a bribe or not offer a bribe. If they choose to not offer a bribe, the game ends. If they choose to offer a bribe, they can pick the amount B ∈ [4, 8]. Once a bribe is offered, they will incur a cost of a transfer fee of 2. Next, Player 2 who acts as a public official can choose to accept or reject the bribe. If the bribe is accepted, the earnings of Players 1 and 2 will increase by 3B and the earnings of Player 3 will decrease by B. If the bribe is rejected, the game is instantly over. Finally, Player 3 who acts as a citizen can choose to punish or not punish Players 1
and 2. If they choose to punish, they can pick the punishment amount P ∈ [2, 12]. P will be deducted from the earnings of Player 3 and the amount 3P will be deducted from the earnings of Players 1 and 2. If Player 3 chooses to not punish the other players, the game
will be over.Once all the decisions have been made, the payoffs are calculated. If this group receives
welfare-enhancing payoffs, the payoffs are the following: •Firm: 3B − 2 − 3P • Public Official: 3B − 3P • Citizen: −B − P
If this group receives welfare-reducing payoffs, the payoffs are the following: • Firm: 3B − 2 − 3P • Public Official: 3B − 3P • Citizen: −7B − P
We conduct the experiment in the Economic Science Laboratory (ESL) at the University of Arizona. Participants will be recruited through email and will consist of undergraduate students. We expect to have 21 or 30 participants per session. Since three participants are required per game, we will ensure that the number of participants per session is a multiple of three. After reading and signing the consent form, the participants will receive the instructions of the experiment both written and verbally. Then, they will fill out a demographic survey on their computer that asks the following: if they have participated in an ESL experiment before, class standing, age, gender, and nationality. Once the survey is completed, groups will be randomly assigned either welfare-enhancing or welfare-reducing payoffs. After the treatment, the aforementioned bribery game will be played. The participants will be made aware of their payoffs, the effects of their decision, and the payoffs of others. Once the game is completed, the participants will fill out a survey asking the motives behind their decisions in the game. After this survey is completed, participants can collect their earnings from the game.

Design details

Randomization method
Randomization is done by the website in which the game is played
Randomization unit
Individuals will be randomized into groups of 3 to play the game. Each group is randomized to either receive welfare-enhancing or welfare-reducing payments.
Clusters
34 groups
Observations
102 participants
Arms
17 welfare-enhancing groups, 17 welfare-reducing groups
Primary outcomes
If participant offers a bribe, if participant accepts a bribe, if participant punishes bribery, if participant is an American citizen
Secondary outcomes
The effect of welfare-reducing/enhancing treatment on willingness to engage and punish corruption

Cited by

No citation data available for this registration yet.

Keywords