Essays in Labor Economics
MetadataShow full item record
I model a hiring process in which a candidate is evaluated sequentially by two agents of a firm. Each agent observes an independent signal of the candidate's productivity. I show that if the second agent values a non-productive attribute of a given candidate, that candidate may be less likely to be hired than a candidate lacking the preferred non-productive attribute due to the first agent adjusting their own quality threshold to compensate. I go on to empirically explore the behavior of prisoners in Oregon based on exogenous shocks to the status quo. These shocks include changes in the generosity of sentence reductions available to certain prisoners and the implementation of a variety of policies that have made it less costly for prisoners to communicate with the outside world. I find that prisoners respond to behavioral reviews with improved behavior on the days immediately before and after a review, but increasing available sentence reductions awarded for good behavior does not reduce misconduct rates among inmates. Furthermore, I find that increasing the ability of prisoners to communicate with friends and family using technology has not led to the decrease in in-person visitation that many have predicted. Instead, total communication seems to have increased in Oregon prisons. Given the extensive literature that suggests increased communication with the outside world reduces a prisoner's likelihood of recidivating, this result may indicate that introducing communication technology and making it more affordable may be a cost effective policy to prevent future crimes. This dissertation includes unpublished co-authored material.