Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Sharing with Friends: Job Information and Social Networks
Last registered on April 16, 2021


Trial Information
General Information
Sharing with Friends: Job Information and Social Networks
Initial registration date
April 16, 2021
Last updated
April 16, 2021 11:18 AM EDT
Primary Investigator
University of Virginia
Other Primary Investigator(s)
PI Affiliation
World Bank
PI Affiliation
American University
Additional Trial Information
On going
Start date
End date
Secondary IDs
Social networks are recognized to be important for labor market outcomes as job seekers rely upon their networks to find jobs. However, there is limited evidence about specifically what kind of job information flows through networks and how agents choose to disclose information with their connections. This project seeks to investigate these mechanisms through an experiment that varies the availability of job information and competition for the job among graduating college students in India. We then track how this information flows among student cohorts.
External Link(s)
Registration Citation
Chiplunkar, Gaurav, Erin Kelley and Gregory Lane. 2021. "Sharing with Friends: Job Information and Social Networks." AEA RCT Registry. April 16. https://doi.org/10.1257/rct.7564-1.0.
Experimental Details
For a period of six weeks, we design a small work-task (a “gig”) that takes approximately 30-45 minutes each week. The task involves finding five policy relevant articles on a particular topic for a researcher. The applicant must then summarize each article. Each week we change the type or article they have to find. For example we ask them to evaluate government policies in health one week, education the next, agriculture the next, or women’s empowerment.

While the task is identical across all groups in a given week, we experimentally vary two aspects: first, the wage i.e. for some batches/colleges, the task offers USD 5, whereas for others it offers USD 10. Second, we vary the "rivalry" in application for the task i.e. either make recruitment a rival or a non-rival activity. For non-rival recruitment, we tell a treated student that: "Here is a small task that will take 30-45 mins of your time and will pay you INR XX. This position is guaranteed to you if you decide to apply for it. There are limited spots and we are considering hiring others in their batch if you’d like to spread the word to others in your batch and encourage them to apply." For the rival recruitment, we tell a treated student that: "Here is a small task that will take 30-45 mins of your time and will pay you INR XX. There are limited spots and you can apply for this position as well as spread the word to others in your batch and encourage them to apply."
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Job applications, knowledge of job openings, information sharing decision
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Performance on the job, heterogeneity by identify of recipient / provider of information
Secondary Outcomes (explanation)
We will focus heterogeneity of impacts based on the identity of the provider and recipient of the job / CV information. More specifically we will examine whether the gender, field of specialization, or baseline academic or cognitive ability matter for the type and degree of information flows.
Experimental Design
Experimental Design
For each week, each batch will be allocated to one of four treatment groups or a control group. As mentioned before, we vary two aspects about the job, namely the wage (High and Low) and the number of students who will be recruited for the job (Rival and Non-Rival). Each week we select a new subset of students within the batch to hear about the job. Selected treatment job-seekers from each college are therefore assigned to one of four groups each week: {High Wage, Non-Rival}, {High Wage, Rival}, {Low Wage, Non-Rival}, {Low Wage, Rival}. Treated students are provided with a unique referral code which they can share with their friends. These friends must enter in this referral code when they apply, which enables us to track the spread of information within the social network.
Experimental Design Details
Not available
Randomization Method
Randomization Unit
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
23 courses across 6 colleges
Sample size: planned number of observations
500 students
Sample size (or number of clusters) by treatment arms
20-30% of students within each batch will be selected for treatment, while the others will be in the control group.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB Name
Institutional Review Board for Social and Behavioral Sciences, University of Virginia
IRB Approval Date
IRB Approval Number