The incentive compatibility condition, firm culture, and social norms under moral hazard

Last registered on May 26, 2024


Trial Information

General Information

The incentive compatibility condition, firm culture, and social norms under moral hazard
Initial registration date
November 30, 2023

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
December 06, 2023, 8:24 AM EST

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

Last updated
May 26, 2024, 8:36 PM EDT

Last updated is the most recent time when changes to the trial's registration were published.



Primary Investigator


Other Primary Investigator(s)

PI Affiliation
University of Leicester

Additional Trial Information

In development
Start date
End date
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
We consider a principal-agent model under moral hazard and examine, theoretically, and empirically, the psychological and social motivations of the agent. In particular we examine the effect of such motivations on the incentive compatibility condition (ICC) of the agent. We show that even when the ICC is violated in the classical analysis, if firm culture and social norms are effective, then the worker will prefer to exert high effort levels. If this is the case, then the classical analysis overstates the informational rents to workers, as well as the nature of the moral hazard problem.
External Link(s)

Registration Citation

Dhami, Sanjit and Mengxing Wei. 2024. "The incentive compatibility condition, firm culture, and social norms under moral hazard." AEA RCT Registry. May 26.
Experimental Details


4 treatments:
(1) Baseline treatment: classical principal-agent analysis
(2) Treatment 1: Firm culture, workplace norms, and guilt-aversion
(3) Treatment 2: Social norms and shame-aversion
(4) Treatment 3: Firm culture, workplace norms, punishment and guilt-aversion
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
worker's effort choices in different treatments, second order belief (worker's belief of the firm's expected effort), normative expectation and empirical expectation of the effort, sanction, etc.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
demographic variables, etc.
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
There are 3 treatments. It's a between-subjects design.
1. Baseline treatment (classical principal-agent analysis)
Subjects are randomly assigned to two roles fi rms and workers. Each subject is assigned to one and only one of these two roles. There is one firm and one worker in each group.
Subjects in the role of firms make one of the following two choices. (i) Offer the contract, which gives them a positive profi t, or (ii) exit the experiment with only the participation fee.
Subjects in the role of workers make one of the following two choices. (i) Choose the contract that is offered, in which case they also need to choose the effort level, which is either `high' or `low', or (ii) choose to exit the experiment with their participation fee.
The chosen effort level is only privately observable to the worker, but it is never observed by the fi rm. The fi rm cannot observe the effort level of the worker, and hence cannot punish or impose any sanctions on the worker.
Each effort level leads to one of two possible states of the world - a good state and a bad state. (i) A good state of the world, in which the fi rm receives a high profi t, or (ii) a bad state of the world in which the firm receives a low pro fit.
Information on the level of profi ts in each state of the world, good state and bad state, is shown only to the firms but not to the workers.
A high (low) effort level by the worker makes the good (bad) state and high (low) profi ts of the fi rm, relatively more likely.
Workers make their effort choices for two different cases, both run in a random order.
Variable wage case: In the first case, workers are paid a wage that depends only on the state of the world.
Fixed Wage case: In the second case, workers are paid a fi xed wage independent of the state of the world.
The two cases will run only once. After completing responses for one case, subjects cannot know any result fron the completed part before starting the second case. After completing both cases, they will get to know the results of both cases. Subject's income in tokens is calculated separately in each case and it depends on their decisions and the decisions of their matched fi rm or worker. After the experiment, only one case will be randomly chosen to pay the subjects. The identity of subjects stays anonymous.
2. Treatment 1 (Firm culture, workplace norms, and guilt-aversion)
The only difference than baseline treatment is that in Treatment 1 the workplace norms announcement makes workers aware of the firm's expectations of worker's (high) effort.
3. Treatment 2 (Social norms and shame-aversion)
The only difference than baseline treatment is that in Treatment 2 the workers are given information that their social group expect they ought to choose a `high' effort level. Subjects also receive information on the effort level chosen by other members of the social group in similar experiments. If a worker falls short of the effort expectations of their social group, they could be sanctioned by the social group. Such sanctions take the form of social disapproval of `low' effort that falls below the expectations of the social group.
4. Treatment 3 (Firm culture, workplace norms, punishment and guilt-aversion)
The only difference, as compared to Treatment 1, is that in Treatment 3 the worker's effort choice is observable to the firm but the firm cannot produce a verifiable signal of the effort of the worker that can be produced as evidence in a court of law (i.e., observable but non-verifiable effort). The firm can disapprove worker's low effort but cannot impose any monetary punishments.
Experimental Design Details
Not available
Randomization Method
randomization done by a computer
Randomization Unit
I. individual-level randomization: (i) which session the subjects will be in, and (ii) which role (firm or worker) they will be in a random way
II. group-level randomization: the firm subject and the worker subject will be matched randomly
III. Task-level randomization: (1) the two experimental tasks will be in a random order and (2) one of the two tasks will be randomly chosen to pay
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
4 treatments
Sample size: planned number of observations
around 400 subjects
Sample size (or number of clusters) by treatment arms
around 100 subjects in each treatment
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Institutional Review Boards (IRBs)

IRB Name
School of Economics at Nankai University Research Ethics Committee
IRB Approval Date
IRB Approval Number
Analysis Plan

Analysis Plan Documents

[Mechanism Design] PAP_27 May.pdf

MD5: e6c94261a60c3c991be7dbbc08509e7d

SHA1: fee46a2beb51ef750255cf0d30223af0dd46d5f3

Uploaded At: May 26, 2024