I took at as them normalizing for a known risk factor. In other words, if white people are armed at a higher percentage than blacks, they wouldn't want that to skew the statistics.