Is anti-Black discrimination concentrated among a discriminatory few, or widespread across many decision-makers? The handful of studies that have addressed this question have reached divergent conclusions, with some suggesting that discrimination follows the 80/20 rule (i.e., a Pareto distribution) and others suggesting that discrimination is normally distributed. This paper explores the distribution of discrimination in hiring, housing, and judicial decisions. Study 1 examined the distribution of anti-Black discrimination in judges' repeated sentencing decisions. The distribution of discrimination was more consistent with a normal distribution than a Pareto distribution. In Study 2, meta-analyses of hiring and housing field experiments revealed anti-Black discrimination in more than 80% of studies. Simulations of widespread discrimination using a normal distribution were more consistent with these experimental data than were simulations of concentrated discrimination using a Pareto distribution. These findings suggest that discrimination is not concentrated in the behaviors of a few highly biased individuals.
Keywords: field experiments; hiring; housing; judicial decisions; meta-analysis; racial discrimination; racism; simulations.