Changing Beliefs, Changing Bribes
Last registered on August 17, 2015


Trial Information
General Information
Changing Beliefs, Changing Bribes
Initial registration date
April 24, 2014
Last updated
August 17, 2015 10:44 AM EDT
Primary Investigator
University of Southern California
Other Primary Investigator(s)
PI Affiliation
Yale University
PI Affiliation
Massachusetts Institute of Technology
PI Affiliation
Yale University
PI Affiliation
Massachusetts Institute of Technology
PI Affiliation
Massachusetts Institute of Technology
PI Affiliation
Massachusetts Institute of Technology
Additional Trial Information
On going
Start date
End date
Secondary IDs
This project investigates the effect of changing legal penalties, and citizens' beliefs about these penalties, on corruption. We focus on a single, clear case: the law against riding a motorcycle without a helmet in India, a setting where many citizens pay bribes instead of a formal ticket, and are unsure of the true legal fine. We randomly inform motorcyclists of the true legal penalty using private information campaigns. The intervention spans a major policy change, during which the fine for riding without a helmet will increase by 500%, allowing us to capture time series variation in how bribes respond to legal changes. Our results, both on the process of bribery as well as on the degree of law-breaking and police enforcement seek to provide new insights into the underlying mechanism of corruption, as well as direct policy implications for reducing graft.
External Link(s)
Registration Citation
Banerjee, Abhijit et al. 2015. "Changing Beliefs, Changing Bribes." AEA RCT Registry. August 17.
Sponsors & Partners

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

Request Information
Experimental Details
Indian drivers generally do not know the true value of fines for traffic infringements. If they overestimate what the fine would be, which previous data had indicated was the case, then they may elect to pay bribes, and even bribe amounts larger than legal fine. In response, our intervention informs drivers of the true fine amounts for one particular, observable offense: driving a motorcycle without a helmet.

To do this, we call all respondents from our baseline survey to ask if they would like to take our followup survey, and while doing this, inform half about the true fine value for riding a motorcycle without a helmet. This was done twice: once about a month after they completed the survey, and once again three months after that. We also sent a text message to both control and treatment drivers at those times, where both were reminded of the dangers of riding without a helmet, while treatment drivers were also told the true fine value for doing so.

After the information campaign, and approximately 9 months after the baseline survey, phone respondents were called for an endline survey and incentivized by an offer of mobile airtime in exchange for answering questions. This panel approach is necessary for us to examine how the effect of information interacts with drivers' priors on fine amounts. The survey also provides insight into the response of the police to greater driver knowledge of fine amounts. We can, for example, observe whether public information on the legal fine influences police to shift their enforcement to other traffic violations on which citizens have no additional information.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
The amount that drivers report paying to the police for helmet violations and whether fine amounts were paid as an official traffic ticket or a bribe
Primary Outcomes (explanation)
Other outcomes of interest include -
* Drivers' knowledge of fine amounts for riding without a helmet
* Drivers' knowledge of fine amounts for other traffic offenses, to see if they updated their subjective beliefs in response to being informed about the helmet fine
* Overall fraction of drivers who elect to ride without a helmet (checking whether the campaign affected selection into being pulled over)
* Changes observed in how both police and drivers negotiate (length of negotiations, response to police demands,
* Probability of paying more than the true fine amount for an offense)
* Hypothetical Driver willingness to pay bribes
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
The first phase of the project randomizes the provision of information at the individual level, with stratification based on geographic location. Our baseline survey was conducted in 41 towns and 4 cities in Gujarat, 27 towns and 7 cities in Haryana, and 42 towns and 6 cities in Rajasthan. We randomly selected 1/3 of drivers in each town or city to be informed about the true fine for riding a motorcycle without a helmet. These drivers were informed of the fine via two individual phone calls and text messages, alongside a message of driving safely on a motorcycle. The control drivers were also called and sent text messages, but with only the generic message of driving safely on a motorcycle.

The second phase of the study will be conducted after the Indian Parliament passes pending legislation to increase fine amounts for traffic offenses (Motor Vehicles Act amendment). The design for this component of the study is currently being finalized, and a separate preanalysis plan will be submitted for that component of the project. It will closely resemble the design in the first phase.
Experimental Design Details
Randomization Method
Randomization in office using Stata
Randomization Unit
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
Sample size: planned number of observations
20,000 drivers
Sample size (or number of clusters) by treatment arms
6,666 treatment drivers, 13,333 control drivers
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
.27 SD for the measure of amount paid to the police for the violation, at 80% power, with an estimated 420 stops for helmet violations and .448 correlation between baseline and followup
IRB Name
Institute for Financial and Management Research
IRB Approval Date
IRB Approval Number
IRB Name
MIT Committee On the Use of Humans as Experimental Subjects (COUHES)
IRB Approval Date
IRB Approval Number
IRB Name
Human Subjects Committee - Yale University
IRB Approval Date
IRB Approval Number
Analysis Plan

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

Request Information
Post Trial Information
Study Withdrawal
Is the intervention completed?
Is data collection complete?
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers