If 80% of the doctors in an geographical area belong to race B then about 80% of doctors should be of race B. Assuming everyone is roughly of equal ability/expe

If 80% of the doctors in an geographical area belong to race B then about 80% of doctors should be of race B. Assuming everyone is roughly of equal ability/experience and no hiring biases are at play. Imbalances need to be measured against the population of doctors, not the general population. The question should be why few people of a particular race go to medical school. The biases need to be addressed at the root. If you focus your bias search at the end-leaf points you are making a blunder. A blunder that can even hurt people for the opposite reasons because there are "too many" of race X relative to local population. Some minorities (Asians) are negatively affected by this lazy anti-bias action.