Using audit data from the 2000 Housing Discrimination Study and a fixed-effects Poisson model, this paper examines discrimination in the number of houses shown to homeseekers. Unlike previous studies, it considers auditors' actual socioeconomic characteristics to minimize the estimation bias....