Algorithmic Bias and Access to Opportunities

The chapter discusses the problem of algorithmic bias in decision-making processes that determine access to opportunities, such as recidivism scores, college admission decisions, or loan scores. After describing the technical bases of algorithmic bias, it asks how to evaluate them, drawing on Iris Marion Young’s perspective of structural (in)justice. The focus is in particular on the risk of so-called ‘Matthew effects’, in which privileged individuals gain more advantages, while those who are already disadvantaged suffer further. Some proposed solutions are discussed, with an emphasis on the need to take a broad, interdisciplinary perspective rather than a purely technical perspective. The chapter also replies to the objection that private firms cannot be held responsible for addressing structural injustices and concludes by emphasizing the need for political and social action.

Modeling College Admission as a Signaling Game

This paper models the college admission process as a signaling game between the admissions office and a pool of heterogeneous applicants characterized by academic abilities and demonstrated interest. In the screening and selection process, applicants attempt to signal their private information through their performance on a standardized test (for academic abilities) and interview (for demonstrated interest). We show, under general conditions, that a separating equilibrium exists in which it is incentive-compatible for applicants to reveal their characteristics truthfully, and thus the admissions office does not fall victim to the problem of information asymmetry. Furthermore, we delineate how this equilibrium can be induced by setting appropriate “scoring” thresholds associated with standardized tests and interviews.

College Admission with Multidimensional Privileges: The Brazilian Affirmative Action Case

In 2012, Brazilian public universities were mandated to use affirmative action policies for candidates from racial and income minorities. We show that the policy makes the students’ affirmative action status a strategic choice and may reject high-achieving minority students while admitting low-achieving majority students. Empirical data shows evidence consistent with this type of unfairness in more than 49 percent of the programs. We propose a selection criterion and an incentive-compatible mechanism that, for a wider range of similar problems and the one in Brazil in particular, is fair and removes any gain from strategizing over the privileges claimed. (JEL I23, O15, I28, J15, D82, H52)