We study how shifting the weight placed on different hiring criteria in an applicant recommendation algorithm shifts hiring and post-hiring outcomes for applicants and firms, using a field experiment on a large online jobs platform.
Experimental design
See attached pre-analysis plan.
Design details
Randomization method
See attached pre-analysis plan.
Randomization unit
Job postings
Clusters
We are targeting a sample of at least 700 job postings for our study.
Observations
We are targeting a sample of at least 700 job postings for our study. The exact number will depend on the rate of posting on the platform, which we do not control. Based on the average length of applicant shortlists, this will result in a sample of at least 17,500 individual applicants.
Arms
The job postings in our sample are randomized into one of two treatment arms, targeting a sample of at least 350 job postings for each of our control group and treatment group.
Primary outcomes
Job search outcomes (e.g., measures of job search effort), job search expectations (e.g., expected search duration), hiring process outcomes (e.g., number of job offers), post-hiring outcomes (e.g., total wage bill), labor market outcomes (e.g., employment)
Cited by
No citation data available for this registration yet.