CODING PRIVILEGE, AUTOMATING INEQUALITY: A SYSTEMATIC REVIEW OF SOCIOECONOMIC SORTING

Abstract

Automated Decision-Making (ADM) systems have rapidly displaced human discretion in critical high-stakes domains such as hiring, lending, and education. While Science and Technology Studies (STS) scholarship has extensively documented algorithmic biases regarding race and gender, the specific mechanisms of class reproduction and the intergenerational transmission of socioeconomic advantage remain significantly undertheorized. To address this gap, this study presents a systematic review of peer-reviewed literature published between 2015 and 2025, adhering to the PRISMA protocol to synthesize data from Scopus, Web of Science, and the ACM Digital Library. The analysis identifies three primary modes of socioeconomic sorting: (1) proxy discrimination, where latent variables like zip codes or device usage serve as digital class markers; (2) predatory inclusion, which targets low-income populations for extractive financial products; and (3) automated gatekeeping, which invisibly filters access to premium labor and educational opportunities. Ultimately, this review argues that ADM systems function as "engines of conservation". By digitizing historical class dispositions into predictive risk scores, these technologies effectively freeze social mobility, systematically coding privilege and automating inequality.