Knowledge that Transforms

Firms are starting to replace humans with algorithms in important screening decisions, but there are potential spillovers of human biases contained in datasets to subsequent algorithmic predictions. When these biases are motivated by human prejudices, there are risks of algorithms perpetuating discrimination. I prove that when datasets are generated by a sufficiently discriminatory human, firms are more profitable when training discriminatory algorithms. If instead enough affirmative action is instituted in favor of a disadvantaged group, firms are more profitable when training algorithms that inflate scores for this group, but this effect diminishes with excess affirmative action.

This paper analyzes the effects of firms posting multiple but varying numbers of vacancies, hence differing in their market power, in professional labor markets. I find that strategic wage posting does, in general, not result in an efficient assignment of workers to firms. This is because firms with a larger number of vacancies pay on average lower wages than their competitors due to a lack of within-firm rivalry. If highly productive firms hire more, the resulting welfare loss due to mismatch may be substantial. Moreover, I provide a potential explanation why firms post uniform wages, missing out on more skilled workers.

We provide the first evidence that the employers of network contacts from college can influence job search success. In contrast to the peer effects literature that emphasizes peer characteristics, we examine how the hiring rate of in-network firms affects reemployment probabilities. To avoid endogeneity from voluntary transitions, we focus on individuals who experience mass layoffs. Using administrative data and a cross-cohort design, we find that network connections with actively hiring employers increase the reemployment rate. This result is driven by reemployment at contacts’ firms. These results suggest that college can improve employment outcomes beyond improved human capital and signaling.