Join 3,564 readers in helping fund MetaFilter (Hide)


Stand up! Stand up and be counted...
December 23, 2007 2:40 AM   Subscribe

Census sensitivity. The Economist takes a look at the politics of enumeration.
posted by goo (14 comments total) 3 users marked this as a favorite

 
Interesting. The Census is somewhat controversial every time it takes place in the UK, but this is shaping up to be a transatlantic storm (in a teacup? We'll see).
For some reason, the government body which selects the census takers is considering appointing arms manufacturer and US-based multinational LockheedMartin.

As BBC Radio 4 puts it, there are concerns that the giant company is subject to the USA-Patriot Act, and that therefore data collected in the UK will be vulnerable to requests from the US government, or data loss, fraud and leaks.

I don't know why this isn't considered sensitive enough to make a proviso that the contract be awarded to a UK owned company. What other nation would contract out such a task?
posted by dash_slot- at 3:58 AM on December 23, 2007


Interesting that the UK is looking at Lockheed Martin considering the kerfuffle in Canada last year.
posted by splice at 5:17 AM on December 23, 2007


What was the kerfuffle splice? I couldn't find anything on a quick google.
posted by srboisvert at 5:32 AM on December 23, 2007


Interesting article, thanks.
posted by languagehat at 5:37 AM on December 23, 2007


Fascinating. Thanks! I'm Census Liaison for my state and the one thing I've learned is that there's an ironic proportion between the importance of taking an accurate and complete count and the interest and energy of the people who should be helping out.
posted by mmahaffie at 5:46 AM on December 23, 2007


srboisvert: What was the kerfuffle splice? I couldn't find anything on a quick google

Really?

Not being obtuse, but I can't/won't actually speak about it because of employment related concerns.
posted by splice at 5:49 AM on December 23, 2007


kerfuffle
posted by dash_slot- at 5:49 AM on December 23, 2007


Agencies in homeless services have more or less abandoned the idea of point in time counts of the population. Whether or not acurate point in time counts are even possible has actually been a pretty hot button debate in the field for twenty years, and has included a lot of statisticians that developed the idea of taking counts over time, so that the population can estimated at X number of people from point in time A to point in time B were homeless.

I've seen how mysterious the migrations of this population can be doing street outreach. There was a three week period last year during the height of the summer when the population on the street should have been at its heaviest and we couldn't find anybody. Our outreach crews walked all the known routes and then started ambling off the grid and still found almost nobody. We still don't know where everyone went during those weeks, there was probably some new service (a good meal spot, maybe) that opened in a different part of town that spread by word of mouth and took everyone away before we could get in the loop. If a point in time count had been taken during this window even by experts in the field it would have appeared as if the street homeless population in Philly was negligible. Right after this couple week blip it was back at its highest point in ten years.

Mystifying.
posted by The Straightener at 7:25 AM on December 23, 2007


"...the one thing I've learned is that there's an ironic proportion between the importance of taking an accurate and complete count and the interest and energy of the people who should be helping out."

Could you explain that a little more? I'm interested.
posted by atchafalaya at 8:24 AM on December 23, 2007


i did a census canvasser gig in 2000 just to see what it was like. the american census machine is a wreck. the regional contractors were apparently selected as some kind of patronage reward, not out of merit. after going unpaid for two months, i made a big enough stink that they finally paid me - in travelers checks!

the issue of carte blanche access to rural properties is also troubling. with no background check, it would have been a perfect job for a burglar, casing over 100 properties a day under government aegis. i felt sorry in particular for the blind woman alone on her 20 acres; it was unreasonable for me to be able to disturb her privacy and create apprehension, even though it was perfectly lawful. i live in an even more sparsely populated area now, there's a locked gate at the foot of my driveway and nobody gets through that gate without my permission.
posted by bruce at 8:56 AM on December 23, 2007 [1 favorite]


atchafalaya, I was being a bit cynical about my experience trying to get local mayors and county administrators to devote time, resources, and/or their own energy to doing the grunt work of Census prep. That is, checking address data, suggesting census geography changes, and helping to talk their constituents into responding to the Census.

Eventually, they do. But it can be a challenge to get things started. The next US Census will be in 2010, but we're already well into the prep work.
posted by mmahaffie at 10:15 AM on December 23, 2007


“GO, NUMBER Israel from Beersheba even to Dan; and bring the number of them to me, that I may know it.”

And we all "know" what he meant by "know:" that number was going to be fucked.
posted by three blind mice at 10:25 AM on December 23, 2007 [2 favorites]


I understand all the worry about censuses, but the historian in me so wishes they could have started the British one in 1601 instead of 1801. That would have made my life so much easier. Or 1301, or 1101, or 601...
posted by jb at 1:46 PM on December 23, 2007


Very interesting. I've always believed, and continue to believe even after reading this, that the quality of decisions people make, whether individually or collectively, is dependent on the quality of information that they have. It seems to me to be almost tautological that we can't govern better by knowing less.

All of the arguments given in these articles for not taking censuses, obviously, depend on examples of abuse of the results and of the process. It seems to me that the problem there is with the abuse, rather than the process, and this is always a danger with any government no matter how or what it does. Stalin shooting people over census processes is no more an argument against censuses, than Stalin shooting people over farming processes would be an argument against agriculture.

Also, the question of whether or not to take a census as such is becoming increasingly moot. Firstly, censuses naturally arise out of government databases from nothing more complicated than counting entries for various things. Secondly, commercial interests already collect so much data that they (collectively) have far more and better information than governments do. Thirdly, as new information is added to any system, it can be used for more and different things, much more than the sum of the old information's use and the new information's use. These uses will be as good, or evil, as the entity using them.
posted by aeschenkarnos at 2:35 AM on December 24, 2007


« Older After I posted this article, many people asked me ...  |  You can't always get what you ... Newer »


This thread has been archived and is closed to new comments