If whites drive more than blacks or Hispanics, and we show with population comparisons that whites are less likely to be pulled over, then it means that whatever disparity we document using population numbers would be an under-, not an overestimate of the true degree of disparity.
