Matching Algorithms that Reward Patience Increase Social Inequalities in Higher Education Admissions

Understanding socioeconomic differences in patience is crucial, because time preferences influence various life outcomes, including education and earnings. Previous research consistently shows that people from lower socioeconomic backgrounds (SES) display less patient behaviors, and are more likely to prioritize immediate over delayed rewards. However, most studies rely on laboratory tasks or monetary trade-offs to measure patience, which limits real-world generalisability. This study investigates SES-related differences in waiting behaviour in the context of French higher education admissions. To do so, we leveraged a unique national database that recorded the admission behaviour of every high schooler in France over a period of three years. During the admissions process, students faced a trade-off: they could either accept an immediate offer from a university or wait in the hope of receiving a preferred offer later. Results collected among 274,316 students reveal that low-SES students were significantly less likely to wait for preferred offers, even when controlling for academic performance, aspirations, and costs such as relocation. This led low-SES students to ultimately enroll in programmes they liked less and that were less prestigious. These results highlight how admission algorithms rewarding patience may inadvertently exacerbate educational inequalities. Such systems, while efficient, can end up favouring students who wait more. This research contributes to current discussions on algorithmic fairness by illustrating that seemingly neutral algorithms can in fact deepen social inequalities if their design rewards patience. It also provides a real life example of socioeconomic differences in patience that are not related to budget constraints.