Using Statistics (John Speaking)
						MAIN PATH: Fred
				Fred

Suppose a company has nearly all white employees and the management prefers to hire on the basis of a recommendation from a current employee. Now suppose the white employees have mainly white friends, neighbors, and relatives. They will recommend white people and the company will end up hiring white people as a result. This will happen even if no one thinks about color!!

Is this process color-blind? By Fred's definition it is. What I'm suggesting is that we should make a distinction between this kind of color-blindness (which I would call "surface color-blindness") and the kind of color-blindness that eliminates practices like the one in the example. I call it "deep" color-blindness.

Surface color-blindness is an attempt to eliminate intentional, explicit discrimination against blacks. Deep color-blindness operates on the assumption that some hiring practices can filter out black people without anyone intending to do so AND without any consideration of the things that are usually thought of as related to doing a job well. It attempts to find and eliminate these hidden 'filters' that operate like intentional discrimination.

I think there are a lot of mechanisms that work like the recommendation system I mentioned, and so I think that we should strive for deep rather than surface color-blindness. We could call affirmative action to eliminate these practices "corrective." It doesn't have to involve preferences, but it might require substantial changes in the way things are done.

Ariadne Home	Affirmative Action Bibliography	Affirmative Action Home