Often the characterization of the impact of a variable as 'appropriate' or 'inappropriate' is controversial, and consequently there is wide disagreement as to the meaning of disparity. This article presents a method of clarifying this debate by developing aggregate measures of disparity. The value of such aggregate measures is illustrated using data from California. The exemplary models suggest that, given relevant data sets, it is possible to develop viable measures of disparity and that an estimation of the extent of disparity is highly dependent on normative beliefs about disparity. (Publisher abstract)