Monday, November 29, 2010

Why analysts appear racist

The following experiment caught my eye: Professor Stephen Sauer, from Clarkson University,
together with two colleagues recruited 101 analysts to review information to value the stock of currently privately held company. All analysts were given the exact same information with two “minor” adaptations: In some of them the CEO had gone to a prestigious university; in some of them to a second-tier school. In some of them the CEO was white; in some of them the CEO was black, thus basically creating four groups (prestigious & white; prestigious & black; second-tier & white; second-tier & black). Then the compared the analysts’ valuations…

In spite of it being “the exact same (fictitious) companies” there were major differences in valuation. By far the highest value the analysts assigned to companies whose CEO was white and from a prestigious university, followed by those from second-tier schools who were black, and second-tier schools who were white. Rock bottom were those companies headed by a CEO from a prestigious university who was black…

Stephen and colleagues found this a rather scary find…

So they decided to run yet another experiment: They recruited another 62 analysts and gave them the same company information. However, this time on half of them, they explicitly stated the following about student recruitment at the CEO’s alma mater:
“The university [had used] a double-blind procedure with no special consideration for gender, race, or ethnicity”, so that it was unambiguous that the CEO had made it into the (prestigious) school purely based on merit, and nothing else.

Now the results changed spectacularly: The companies with CEOs from a prestigious university who were black received the exact same valuation as companies with CEOs from a prestigious university who were white. Apparently, people do not just devalue a company because its CEO is black – phew! – there is something else at play.

Why did they then in the first experiment assign the lowest value to those companies with a CEO of a prestigious company who is black? Well, apparently, whenever people see a black person from a prestigious university, they are inclined to assume that he might have been admitted there not based on merit but because of some sort of affirmative action…: positive discrimination. And hence, that it is likely that he is not as good as his credentials might suggest. And of course, statistically they are right…; black males are more likely to have been admitted to B-school due to affirmative action than white males (for the simple reason that the number of white males admitted due to affirmative action is quite certain to be zero). Take away that possibility and analysts don’t care about the colour of someone’s skin.

But that of course does not mean that all black males are admitted to a prestigious school due to affirmative action. It doesn’t even mean that most black males are admitted to a prestigious university for that reason. So analysts do get it wrong rather often, and that must be one darn annoying thing, if you’re a black male at a prestigious university who simply made it in based on merit, and nothing else.