Social bias blind spots: Attractiveness-biased outcomes are seemingly tolerated because people fail to notice the bias

Discrimination remains a key challenge for social equity. A prerequisite for effective individual and societal responses to discrimination is that instances of it are detected. Yet, prejudice and discriminatory intent are rarely directly observable and the presence of discrimination has to be inferred from circumstantial evidence, such as the over- or underrepresentation of certain individuals (i.e., statistical bias). Here, we study how people judge outcomes that are statistically biased along different dimensions. Six primary and two supplemental studies with Dutch and U.S. participants (total N = 3,591, six preregistered) show that gender- and race-biased outcomes are perceived as much less fair than unbiased outcomes, but we do not observe the same for attractiveness-biased outcomes. Our results suggest that this occurs because people spontaneously pay attention to a few salient dimensions (e.g., gender and race) when scrutinizing decision outcomes. Statistical biases along less salient dimensions (e.g., attractiveness) are more likely to go undetected. Thus, people may show muted responses to some biased outcomes not because the bias is tolerated or seen as legitimate, but because people fail to notice the bias.