Let me also add one thing I didn't phrase clearly last night - if it is true (and I, like you, don't believe it's true) that an Asian Stanford grad has the same

Let me also add one thing I didn't phrase clearly last night - if it is true (and I, like you, don't believe it's true) that an Asian Stanford grad has the same perspectives and experiences as a white Stanford grad, then that gives all the more validity to making sure you're not unnecessarily discriminating against one or the other group. Your ratio of Asian to white hires should be roughly on par with the fraction of, if nothing else, Asian and white Stanford grads. If it is wildly different, then it's an indicator that something is wrong. Perhaps you are not losing out on perspectives and experiences, per se, but perhaps the biased person running your initial phone screens is causing you to lose out on good candidates who are from a race they don't like, i.e., you're limiting the number of candidates of this perspective and experience that you're willing to interview. That would be a problem worth being aware of! Completely ignoring race means you don't have the data. It doesn't prevent the problem from happening. (See also, "Why is the net wired randomly?", asked Minsky. "I do not want it to have any preconceptions of how to play", Sussman said.*)