The US census incensed us: I sense a lack of consensus

TTTE loves surveys. As a student of survey design, I saw the chance to participate in America’s decennial census as quite an honour! Well it turned put to be a dubious kind of a honour, tinged with confusion and dismay. I do not consider myself particularly precious, but I cannot abide these seemingly pedestrian errors:

Confusing instructions

The census arrived saying ‘Fill it out on April 1.’ … Inside the envelop was an accompanying letter from the Director of the Census Bureau asking that it be completed and mailed back, Today! Why do they ask me to fill in details for a date ten days into the future?

Surely if they want an accurate snapshot of the population on April 1, shouldn’t they ask for the form to be mailed back on April 2?

If that wasn’t confusing enough…

Question 1: How many people were living or staying in this house, apartment or mobile home on April 1 2010?

The list of qualifications is long and intriguing: Count all people, including babies, who live and sleep here most of the time. Do not count people at college, armed forces, nursing home, in jail.

Eh? That’s the path to a bunch of useless outcomes! Do they want the people sleeping in the house on the night in question, or the people who usually sleep in the house? Both?

Question 6: The census is fixated on race.

The census asked me if I am white, black, asian or one of eleven other options (see below).  I guess my skin is, in certain places, whitish. But I’ve never had to describe myself as such…

But to focus on race misses the bigger picture. Don’t they want to know if I’m an immigrant? A citizen? What is my original nationality? Why so obsessed with skin colour?

Asking about race means non-significant differences are amplified and significant differences are masked. A Sudanese refugee would mark the same box as Barack Obama.

The Asian kids at Berkeley are mainstream American. Much more so than any transplanted Australian. But if you had one Burmese grandparent, would you describe yourself as white? Probably not. And so their race is counted, to mark them as ‘different’.

Why put such an important question in such uninformative terms? Asking about language spoken, country of birth and parents’ country of birth gives a much more nuanced view of patterns of ethnicity, linguistic diversity and possible subsequent disadvantage.

A Silent Protest

To add to the fun, the census website has no useful FAQ. All of these problems likely explain the fact that one in three American households don’t bother to send their form back.

Meanwhile the Australian census has about 40 pages of useful information on health, wealth, education, transport choices, etc.  That’s why I ♥ the ABS. Shout out to my homies in Belconnen.

Politicisation

At home, the census is not controversial outside the tea-room at the ABS. Statisticians probably get up in arms over question-wording. But in the US, oh boy.

Census time heightens privacy concerns
Census stalkers
Census: An unconstitutional invasion of privacy

When people like these say ‘the Government’, they mean ‘the cabal of bankers and industrialists who secretly run the world‘. They want a fictionally powerful body to blame things on. But ‘the Government’ is of and by us. The government is us. Fearing it is an embrace of powerlessness, and there’s a whole strain of American politics that makes that fear an unimpeachable talisman.

Anyway, the end result of all the fighting and incompetence is one-and-a-half measly pages of confusing malarkey.

Which informed the alternative title for this piece, which, sadly, got voted down in committee: The US census: a shit census!


Published by

thomasthethinkengine

Thomas the Think Engine is the blog of a trained economist. It comes to you from Melbourne Australia.

5 thoughts on “The US census incensed us: I sense a lack of consensus”

  1. When working in the States, I found I often had to mark my race on various forms. I’d never thought about being ‘caucasian’ before – prefering to identify as a tighty-whitey or neo-colonialist. Anyway, I think race is pretty taboo in Australian culture and pretty out there in American. Think about how US politics is so heavily influenced by raical blocks of voters. Even Bush courted the Latino vote. Although Australia’s pretty multicltural, this hasn’t been a feature of Australian polictics since the Cold War, and then it was about country of origin more than race. I can’t imagine Abbott courting the ‘Asian’ vote in the next election. He’ll be out courting the mortgage belt, with maybe a few women thrown in. So maybe race is a valid question to ask in the US census, since its a contruct Americans seem to identify with enough to change voting patterns. That’s something every Government wants to know about.

    Like

  2. This obsession with race has always bothered me (reading any US demography/social science/economics/political writing at all, race is central). However, you assume that they are trying to get at “patterns of ethnicity, linguistic diversity and possible subsequent disadvantage” with the question about race, whereas it’s not that at all. It’s apparently about monitoring racism…

    2010.census.gov tells me that questions about race have been asked since 1790, is key to implementing many federal laws (civil rights and the like) and the data are used to determine congressional, state and local voting districts. Strange indeed, but not really the fault of the census.

    Apparently, the data are also used to monitor fairness in employment practices, and racial disparities in health and education. This seems like a good idea to me, if indeed race is a good predictor of these kinds of things.

    If this is so, however, there is a cultural problem of monumental proportions that could be addressed partly by the changes to the question you suggest. For a 10 question census, though, (WTF! 10 questions!?) it seems they have stuck with the one required by law, without bothering with the detail.

    Like

Leave a Comment