Customer satisfaction and American Idol (9 June 2009)

Robin Hanson, an economist who frequently writes about signaling -- how cultural capital is deployed -- notices this WaPo account of the flaws in customer-satisfaction surveys. It turns out that people are systematically biased toward giving people who are (in Hanson's interpretation) perceived to have higher status a better evaluation.

Hekman [the study's lead author] found that these objective measures of performance correlated with patient satisfaction reports only when the doctors were white men. For women and minorities, extra quality, accessibility and diligence not only did not result in better evaluations by patients -- they produced worse evaluations.

As someone who, as a college instructor, was frequently rated by my "customers," the study's findings ring true to me. I could coast on my white maleness where my female colleagues couldn't; they, meanwhile, were being told how unfashionable their clothes were and how they should smile more. I always resented the way universities would rely on evaluations in their decision-making, and the idea that teachers' pay could be affected by it seemed like a great reason to get out of the profession altogether. Some evaluators will do their best to be objective but have their opinions colored perhaps unknowingly by their desire for what Hanson calls status affiliation; others will gleefully regard the survey as an opportunity to vote in a popularity contest.

Hanson seems eager to differentiate status affiliation from outright racism:

People usually invoke two explanations for such behavior:

1) Irrational or ideological racism or sexism.

2) Rational stereotyping that just happens to go wrong in these cases.

But a third explanation seems to me more plausible:

3) We prefer to affiliate with higher status folks. If female doctors, black or female sales clerks, or latino golf club employees are considered lower status, then customers will be less satisfied with them even if they do exactly the same things.

That seems like splitting hairs to me -- racism, sexism, income inequality, class bigotry, and so on are better justified if they can be displaced and relabeled as status concerns. Ideological racism is precisely what this study describes, the systematic association of status and better performance with otherwise irrelevant characteristics. The close association of status with race and gender and so on is what makes racist ideology seem perfectly rational, excusing us for our prejudice. And that association is justified, in a tautological sort of way, through surveys and such that seem to instantiate democratic participation. Hey, we're voting! We matter! But in that spirit -- that voting is about boosting our own self-esteem -- we vote as a way to express what we would like to be affiliated with, not what we have decided about the matter at hand. If we have no comprehension of the matter at hand, so much the better; our ability to vote our ego becomes that much easier to countenance.

Democracy is all well and good, but it seems to be evoked at times to justify and glorify something altogether different, when uninformed people are invited to rate the performance of those whose work they aren't that qualified to evaluate. Such surveys, popularized by shows like American Idol, end up having the function of negating the idea that objective standards are relevant, and promote the idea that status and popularity are always trustworthy proxies for quality. This durable species of capitalist ideology is a close cousin of the kind of market-think that views price or brands as always-reliable signals of quality. This shifts the responsibility for perpetuating status-quo inequities onto ordinary people, making it seem the natural order of things and an expression of the people's choice. So, any time we cast an ignorant vote, fill out some comment card or cast a vote on So You Think You Can Dance out of some vague and unquestioned impulse about their "talent", we strengthen the grip of this ideology.