By "making a decison based on race" I mean making a conscious decision in which race is a significant factor. If you want me to be more specific, I'll restate that as: in my estimation, any racial bias that may exist among the scientists I've met is not significant enough to in aggregate result in a difference greater than 1% in the ratio of people of any particular ethnicity hired compared to people of other ethnicities. I'd suspect other biases have far greater effect, e.g. against people from no-name universities.