The Role of the Source of Power in the Public Bad Game

Last registered on January 09, 2026

Pre-Trial

Trial Information

General Information

Title
The Role of the Source of Power in the Public Bad Game
RCT ID
AEARCTR-0017032
Initial registration date
January 08, 2026

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
January 09, 2026, 9:01 AM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Region

Primary Investigator

Affiliation
IMT School for Advanced Studies Lucca

Other Primary Investigator(s)

PI Affiliation
IMT School for Advanced Studies Lucca
PI Affiliation
University of Bologna

Additional Trial Information

Status
In development
Start date
2026-01-08
End date
2026-01-15
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
Collective decisions in the real world rarely occur in isolation but within systems where authority shapes norms, expectations, and outcomes. While extensive research has examined leadership in cooperative behavior, the distinct influence of authority, in particular how its legitimacy affects compliance remains underexplored. This gap is especially critical in environmental dilemmas where technical solutions require deeper understanding of human behavioral responses to different sources of power.


This study investigates how different sources of authority influence rule compliance in a public bad game, and how the authorities themselves perceive the sources of power. Authorities are randomly assigned to one of four sources of power (Consensus, Merit, Random, or Win-the-Struggle) and set a rule r ∈ {0,1,…,10}. Group members make conditional compliance choices using the strategy method and are paid according to the rule actually chosen by their Authority. Since the rule is payoff-neutral, any systematic variation in compliance identifies non-monetary motives.

Experimental setup

Our experimental design consists of two distinct experiments conducted in the following order:
(i) the Authority Side (Part_one), where a series of tasks are performed and rules are collected; and (ii) the Group Member Side (Part_two), where participants play a Public Bad Game (PBG) in groups of five using the strategy method. More specifically, we run treatments that vary the source of power associated with each Authority.
We study how different sources of power affect rule compliance. The treatments are defined as follows:
T1. Random Selection: Group members receive rules from an Authority randomly selected.
T2. Consensus Selection: Group members receive rules from an Authority elected by majority rule.
T3. Merit Selection: Group members receive rules from an Authority who won a selection based on a combination of an IQ test and an effort task.
T4. Win-the-Struggle Selection: Group members receive rules from an Authority who won an all-pay auction.

Hypotheses:

H_1: The distribution of rule choices in Part_one is identical across all sources of power (Random, Consensus, Merit, Win-the-Struggle).
Test: We will compare the distribution of rules chosen by Authorities across treatments using an appropriate non-parametric test.

H_2: The group members’ absolute deviations in Part_two are distributed in the same way across all sources of power.
Test: We will compare the distribution of deviations ( D_i=| x_i − r | ) across the four power sources using an appropriate non-parametric test.

H_3: The behavioural types in Part_two are distributed in the same way across all sources of power.
Test: We will classify subjects into behavioural types according to the shape of their conditional compliance function (investment decisions as function of rule levels). Then, we will compare the distribution of these behavioural types across treatments using an appropriate non-parametric test.

Auxiliary hypotheses:

AH_4: We test whether lower deviations are associated with higher (unconditional) intrinsic respect for rule, as measured by compliance in the Y Task used in Bicchieri, C., Gächter, S., Molleman, L., & Nosenzo, D. (2025).

AH_5: We test whether lower deviations are associated with higher descriptive and normative beliefs.

AH_6. Conditional on the rule level r, investment decisions in the Public Bad Game do not differ across sources of power.
External Link(s)

Registration Citation

Citation
Bilancini, Ennio, Veronica Pizziol and Filippo Scarparo. 2026. "The Role of the Source of Power in the Public Bad Game." AEA RCT Registry. January 09. https://doi.org/10.1257/rct.17032-1.0
Experimental Details

Interventions

Intervention(s)
Intervention (Hidden)
Intervention Start Date
2026-01-08
Intervention End Date
2026-01-15

Primary Outcomes

Primary Outcomes (end points)
Part One:
The main variables from the first part of the experiment are:
(i) the rule chosen by the Authority (r∈{0,…,10}).

Part Two:
The main variables from the second part of the experiment are:
(i) the Absolute Deviation of each participant from the rule (D_i= | x_i − r | );
(ii) the behavioural type for each participant based on their conditional investment decisions;



Primary Outcomes (explanation)
In our setting, compliance is defined as the degree to which an individual’s decis align with the established rule r. For each possible rule r∈[0,10], participants indicate their intended investment x(r) using the strategy method. The individual deviation from the rule is measured as representing the absolute distance between the prescribed behavior and the actual decision.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
The experiment consists of two sequential parts: an Authority side (Part_one) and a Group Member side (Part_two), conducted with separate samples of participants.
In the Authority part, participants are randomly assigned to one of four sources of power — Random, Consensus, Merit, and Win-the-Struggle. Each participant acts as an Authority responsible for setting a rule. Authorities are informed that their decision will later be used with a different group of participants (the Group Members), but will have no direct payoff consequences for themselves or others in this stage.
Each Authority is asked to (i) choose a rule r∈{0,1,…,10} indicating the recommended investment level for group members who will play a Public Bad Game; (ii) report their own preferred investment level; (iii) state beliefs about average investment decisions of others (beliefs about compliance); and (iv) state beliefs about the rule choices of other Authorities exposed to the same source of power. These additional measures capture the Authority’s expectations and preferences beyond the chosen rule.

In the Group Member part, a separate sample of participants—who never interact directly with the Authorities—are randomly matched in groups of five and assigned to one of the same four power-source conditions. Participants at the beginning complete the Y task and later make conditional investment decisions for each possible rule between 0 and 10 (strategy method). The actual rule is implemented for payment. After the Public Bad Game, Group Members report beliefs about others’ deviations and beliefs about what others think one should do.
Final payoffs consist of a fixed participation fee plus a variable bonus determined by the realized rule and the participant’s actual decision in the Public Bad Game.
In addition, another group of participants plays the Public Bad Game under the same parameters but without Authority and any rules (no treatment- Standard PBG). Decisions are one-shot and unconditional, and all other procedures are identical to Part Two.
Experimental Design Details
Randomization Method
All randomization is pseudo-randomly generated using a library random on oTree software. (Treatments are assigned iteratively to participants based on their order of arrival.)
Randomization Unit
The unit of randomization is individual and each participant is randomly assigned a treatment
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
Part One: 90 groups of 2 participants for treatment,
Part Two: 90 groups of 5 participants for treatment plus 10 groups of participants for Standard PBG.
Sample size: planned number of observations
Part One: 720 participants in total. Part Two: 1850 participants in total.
Sample size (or number of clusters) by treatment arms
Part One: 180 participants for treatment.
Part Two: 450 participants for treatment plus 50 participants for Standard PBG.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
Comitato Etico Congiunto per la ricerca della Scuola Normale Superiore, Scuola Superiore Sant’Anna e Scuola IMT Alti Studi Lucca
IRB Approval Date
2025-06-30
IRB Approval Number
24/2025

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials