Apparent Algorithmic Discrimination and Real-Time Algorithmic Learning in Digital Search Advertising
35 Pages Posted: 12 May 2020 Last revised: 15 Apr 2021
Date Written: April 12, 2020
Digital algorithms try and target ads to consumers who would be interested in these ads. They do this by swiftly learning whether or not targeted consumers engage with a particular ad. In this paper, we explore whether this learning process can lead to disparate outcomes for minority groups, potentially reinforcing inequality. We examine this question in the context of ads displayed following searches on Google for Black or White names. Previous research documented that searches for Black names were more likely to return ads that highlighted the need for a criminal background check than searches for White names. We implement a set of search advertising campaigns that target ads to searches for Black and White names and use the data that Google provides to advertisers to analyse the outcome. Our ads are indeed more likely to be displayed following searches for Black names. We document the process of algorithmic learning as one explanation for this finding. If an algorithm receives in real time less data about one group, it will learn at different speeds. Since Black names are less common, the algorithm learns about the quality of the underlying ad more slowly, and as a result an ad, including an undesirable ad, is more likely to persist for searches next to Black names even if the algorithm judges the ad to be of low-quality. We confirm similar patterns in a second study that targets ads towards searches for religious groups. Our results suggest that the process of real-time algorithmic learning can lead to differential outcomes across those whose characteristics are more common and those who are rarer in society.
Keywords: Algorithmic Bias, Advertising, Inequality, online advertising, algorithmic learning, digital discrimination
JEL Classification: M21, M31, M15, M38
Suggested Citation: Suggested Citation