Cultural Determinants of Data Acquisition Costs

I saw a fascinating piece in the New York Times awhile back.  It was about a trend at American universities, asking applicants if they were gay or not.  Apparently, these institutions believe that by asking students this question, they are sending a message that they are a gay-positive environment.

Interesting.

Americans think that transparency about identity is the path to utopia.  Enrolment statistics by race?  They’ve got them.  Indeed, they are required to keep such statistics, because of a clutch of laws designed to monitor whether or not Blacks (and, to a lesser extent, Latinos and America Indians) are being discriminated against.

In Canada, the rule of thumb is simple: on forms used for administrative purposes, you can’t compel anyone to reveal data about identity, beyond what is strictly necessary to achieve the purpose for which the information is being collected.  So, on applications to universities and colleges, asking people’s names and addresses is about as far as you can go (provinces have different standards on whether you can ask gender – some say you can’t).  Asking about ethnicity, or aboriginal status?  Totally verboten.  Whereas in the US it’s mandatory.

What that means is that, in Canada, acquiring any data about students – other than raw numbers – requires voluntary surveys.  And those can get expensive: done centrally through StatsCan (and its levels of quality standards) they cost millions; even if you just get a decent-sized consortium together to do something, it will run into hundreds of thousands once you count everyone’s labour costs.  You can get it down into the tens of thousands if you go with an electronic survey, but then there are response bias issues (you can correct for them, but it requires someone to have already done a decent survey to begin with – and with the loss of the census long form, it’s not clear that we have such a survey).

Of course, even Canada is at least somewhat ahead of, say, France.  There, the local conception of nationalism means that state agencies are forbidden from classifying citizens as anything other than citizens.  Blanc, beur, noir: they’re all French according to the government, and its socially unacceptable to classify them as anything else.  A morally attractive stance, perhaps, but what it means is that the French have real trouble measuring social inequality in ways that matter.

All of this is simply to say, if you’ve ever wondered why we don’t have statistics on ethnicity the way the Americans do, it’s this: they assume racial bias exists and keep stats to measure it.  We assume that racial bias exists, and so try to mask parts of individuals’ identities to prevent it.

Posted in

3 responses to “Cultural Determinants of Data Acquisition Costs

  1. An interesting footnote to this (at least I think it is interesting) . . . when I was on sabbatical in France a few years ago, the reserachers there were amazed at the detail on ethnicity, etc. in the Canadian data. I asked why they didn’t have this in France. Apparently they stopped gathering these data following WWII. During the war, the Nazis and their Frenach collaborators used these detailed data to locate and deport the Jewish population and any others they deemed undesirable.

    1. That is interesting – I had always understood this as an issue in Germany, but not France. In France I understood the reluctance to be much more about Republican ideals, no hyphenated Frenchmen/women, etc.

Leave a Reply

Your email address will not be published. Required fields are marked *

Search the Blog

Enjoy Reading?

Get One Thought sent straight to your inbox.
Subscribe now.