Seminarios Académicos

Abstract: In many hiring processes, job candidates are evaluated by intermediaries, such as HR personnel and/or external recruiters. These intermediaries evaluate the candidates based on CV and other information, and preselect the candidates which are most likely to be hired by the hiring manager. However, to do so, intermediaries should not only take the skills of the candidate into account, but also potential biases of the hiring manager such as discrimination of a certain group. In this paper, we study how intermediaries form these beliefs and how they ultimately decide which candidates to forward to the hiring manager. Using two incentivized experiments, we ask participants to either predict how a manager will rate a job candidate or choose which job candidate to forward to the manager for a hiring decision. Participants observe both information about the job candidate's application—such as gender and test scores measuring aptitudes and personality—and also certain information of the manager, including gender. In our experiments, we consistently find that participants expect managers to favor job candidates of their gender, and do not observe a clear gender bias against either males or females. Using process-tracing methods, we also measure how participants allocate their attention to different types of information about the job candidates. We find that participants exhibit distinct attentional patterns favoring specific information, and these patterns are predictive of their choices. Notably, these patterns do not seem to be driven by whom the participants are hiring for, but rather by their own personal beliefs.