The census is racist?

[mr friendly guy]

I have seen people react in a similar manner to the Australian census. They say I am Australian and prefer to put that rather than <insert ethnic group here>. In reality I just think of it as them asking for you ethnicity rather than your nationality. Because all citizens are Australian but we are a nation of immigrants and as such are composed of multi ethnic groups.

I haven't really seen anyone self-describe as an "African Australian" though. I should call myself "English Australian", having both an English birth cert and Australian citizenship.

I don't mean in the sense they want to put things like Chinese Australian or African Australian. They want to put Australian in the ethnicity box period.

When I am overseas I say I am Australian, if people (or the census) asks me my ethnicity I put Chinese in the box. Its not a biggie, ethnicity and nationality are not necessarily synonymous.

[Netko]

Sorry if I stirred a hornets nest with my little list. I was mostly using the common way (to me) American subethnicities are defined. Anyway, the point wasn't the specific ethnicities, it was more that only in the US is the dominant ethnicity on the census "White". Its a purely racial designator telling you very little about the person's cultural, ethnic or even geographical background (compare the listed options about Asian ethnicities and the big "White" blob that covers Europe, Russia and the 'stans).

[irishmick79]

If you want to argue that the census is racist you could probably do that, but this is not the way to go. Since a lot of the census work is based on making contact by going door to door or by direct mail or telephone, you could make an argument that statistically speaking, poorer people would be underrepresented since they might lack those of contact. Since poorer people in America tend to be minorities, you could argue that the census discriminates against poor minorities. I'm not saying it's a good argument, just saying that there's some better grounds to stand on if you chose to give it a go.

[Broomstick]

If you want to argue that the census is racist you could probably do that, but this is not the way to go. Since a lot of the census work is based on making contact by going door to door or by direct mail or telephone, you could make an argument that statistically speaking, poorer people would be underrepresented since they might lack those of contact.

Poor people don't have doors?

Yes, some lack telephones, but most poor people do have a mailing address. I spent two months stomping around Lake County, Indiana to specifically update the Census mailing lists precisely so poor people weren't missed. The Census also makes an effort to locate people who don't, in fact, having mailing addresses in order to hand deliver Census forms or interview them face to face. No joke, one choice under "housing type" on the forms used by Census workers is CAVE. As in "lives in a hole in the ground". So while counting such people is a concern the Census does make an effort locate and count even the homeless.

Since poorer people in America tend to be minorities,

The Census actually documents that poor people in America actually, in fact, tend to be white by a large margin.

That said, minorities have a higher portion of their respective groups being poor, but the vast majority of poor in the US are, in fact, white. Which shouldn't be that shocking as whites are still the majority in most states of the union. (Hawaii being a notable exception, and I think California might fall into that group now).

The issue of race/ethnicity has been sticky from the start - if I recall, the initial few Censuses the only choices for race were "white" and "black" - totally ignoring the Natives, some of whom had become US citizens, and the existence of significant numbers of people of mixed heritage even back then. Stephen J. Gould in his essay "The Politics of Census" (I have a copy in my edition of Hen's Teeth and Horses' Toes) not only details some of the problems in collecting information for the 1840 Census that shows clear racism on the part of data collectors but also some of the nasty uses to which that misinformation was put.

As the US developed a more nuanced understanding of heritage more categories were added, although in reality the lines between many groups, if not most, are blurred and there will always be exceptions. As time goes by more and more people will identify as "mutts" as the individuals who can say "my ancestry is entirely X" will become the minority. Cultures, and cultural definitions, do not remain static over time. We're seeing that, not only in looking at historical data but also by looking around us as we live our lives.

[irishmick79]

If you want to argue that the census is racist you could probably do that, but this is not the way to go. Since a lot of the census work is based on making contact by going door to door or by direct mail or telephone, you could make an argument that statistically speaking, poorer people would be underrepresented since they might lack those of contact.

Poor people don't have doors?

Yes, some lack telephones, but most poor people do have a mailing address. I spent two months stomping around Lake County, Indiana to specifically update the Census mailing lists precisely so poor people weren't missed. The Census also makes an effort to locate people who don't, in fact, having mailing addresses in order to hand deliver Census forms or interview them face to face. No joke, one choice under "housing type" on the forms used by Census workers is CAVE. As in "lives in a hole in the ground". So while counting such people is a concern the Census does make an effort locate and count even the homeless.

Are you looking at the long form for the US census? The sample form that I saw for the 2010 census that I found here doesn't feature that at all. The form that's going out to every household is only about 10 questions long, while the long form that goes into more in depth questioning about economic status and such will only go to about 3 million households.

I don't doubt the desire of the Census bureau to get an accurate count - according to various news outlets like this san francisco chronicle article, door to door follow-up visits for households who didn't respond to the mailings would cost roughly $80 to $90 million per 1% of the population. As thorough as the bureau wants to be, I imagine that if there's a place where they could keep costs down if that was an objective would be in the door to door operations.

The Census actually documents that poor people in America actually, in fact, tend to be white by a large margin.

That said, minorities have a higher portion of their respective groups being poor, but the vast majority of poor in the US are, in fact, white. Which shouldn't be that shocking as whites are still the majority in most states of the union. (Hawaii being a notable exception, and I think California might fall into that group now).

The issue of race/ethnicity has been sticky from the start - if I recall, the initial few Censuses the only choices for race were "white" and "black" - totally ignoring the Natives, some of whom had become US citizens, and the existence of significant numbers of people of mixed heritage even back then. Stephen J. Gould in his essay "The Politics of Census" (I have a copy in my edition of Hen's Teeth and Horses' Toes) not only details some of the problems in collecting information for the 1840 Census that shows clear racism on the part of data collectors but also some of the nasty uses to which that misinformation was put.

As the US developed a more nuanced understanding of heritage more categories were added, although in reality the lines between many groups, if not most, are blurred and there will always be exceptions. As time goes by more and more people will identify as "mutts" as the individuals who can say "my ancestry is entirely X" will become the minority. Cultures, and cultural definitions, do not remain static over time. We're seeing that, not only in looking at historical data but also by looking around us as we live our lives.

Absolutely true - I think I slightly misspoke, although you made the point that I was driving at, namely that minorities have a larger portion of their populations living in poverty than whites.

The census definitely tries to get things right, but how the data gets used is really where most problems lie, and that's not really the fault of the bureau. The one thing I do question is their counting of prisoners - they're allowing states to count inmates where they're currently being held. That could allow for a lot of manipulation if sate governments decide to screw around and ship prisoners to certain areas where they would like to see a higher population count.

[Broomstick]

If you want to argue that the census is racist you could probably do that, but this is not the way to go. Since a lot of the census work is based on making contact by going door to door or by direct mail or telephone, you could make an argument that statistically speaking, poorer people would be underrepresented since they might lack those of contact.

Poor people don't have doors?

Yes, some lack telephones, but most poor people do have a mailing address. I spent two months stomping around Lake County, Indiana to specifically update the Census mailing lists precisely so poor people weren't missed. The Census also makes an effort to locate people who don't, in fact, having mailing addresses in order to hand deliver Census forms or interview them face to face. No joke, one choice under "housing type" on the forms used by Census workers is CAVE. As in "lives in a hole in the ground". So while counting such people is a concern the Census does make an effort locate and count even the homeless.

Are you looking at the long form for the US census?

No. I have emphasized the relevant portion of my quote. I was referring to forms used by Census employees, not the "Census forms" mailed to households. Please note the distinction. People living in caves would not get the standard Census forms because they don't have mailing addresses.

The form that's going out to every household is only about 10 questions long, while the long form that goes into more in depth questioning about economic status and such will only go to about 3 million households.

Correct. However, the Census is not counting JUST traditional households - the Census also attempts to count people living in weird or illegal situations, or who are just plain homeless. As I said, such people do not get the standard form since they don't have normal mailing addresses. Such people are visited by Census employees utilizing different tools. I'll be finding out what, exactly, those tools are in about two weeks because - YAY! - the Census called me last night and that's when I go for training.

As thorough as the bureau wants to be, I imagine that if there's a place where they could keep costs down if that was an objective would be in the door to door operations.

Yes, hence the current advertising blitz on TV, radio, print, mailings to households.... all in an effort to increase household compliance so that follow up to them is minimized and face-to-face work is predominantly for those people who can't be reached by mail.

The census definitely tries to get things right, but how the data gets used is really where most problems lie, and that's not really the fault of the bureau. The one thing I do question is their counting of prisoners - they're allowing states to count inmates where they're currently being held. That could allow for a lot of manipulation if sate governments decide to screw around and ship prisoners to certain areas where they would like to see a higher population count.

My understanding is that counting prisoners where they are held has been standard operating procedure since the beginning, but prisoners used to be a lot more dispersed and thus it was less of an impact. There is also the issue of, if you don't count them where they live, where do you count them? And do you distinguish between those with short sentences (under a year, for example) who arguably are only temporarily relocated vs. those with lengthy sentences (10, 20, 30 years, life, etc.).

[CaptainChewbacca]

Well, I got notified in the mail that I'll be getting the long census form. I'll let you know whats on it.

[Questor]

Well, I got notified in the mail that I'll be getting the long census form. I'll let you know whats on it.

Broomstick, are the long forms randomly distributed to utilize sampling, or are they targeted somehow?

[irishmick79]

No. I have emphasized the relevant portion of my quote. I was referring to forms used by Census employees, not the "Census forms" mailed to households. Please note the distinction. People living in caves would not get the standard Census forms because they don't have mailing addresses.

Um, yeah, my bad. I completely missed that when I first read it.

Correct. However, the Census is not counting JUST traditional households - the Census also attempts to count people living in weird or illegal situations, or who are just plain homeless. As I said, such people do not get the standard form since they don't have normal mailing addresses. Such people are visited by Census employees utilizing different tools. I'll be finding out what, exactly, those tools are in about two weeks because - YAY! - the Census called me last night and that's when I go for training.

How much does the census expect the foreclosure crisis to impact their contact rates, especially in the hardest hit states? I was trying to find some sort of analysis of contact rates and really couldn't find much. Is there a report out there that covers that sort of thing?

My understanding is that counting prisoners where they are held has been standard operating procedure since the beginning, but prisoners used to be a lot more dispersed and thus it was less of an impact. There is also the issue of, if you don't count them where they live, where do you count them? And do you distinguish between those with short sentences (under a year, for example) who arguably are only temporarily relocated vs. those with lengthy sentences (10, 20, 30 years, life, etc.).

What I was reading was suggesting that a few districts in New York would have to be redrawn if the prison population wasn't counted. This Washington Post article describes those NY districts as being rural and conservative, and if they had to redraw those districts they might change their political composition a bit. From my understanding there isn't much of a distinction made regarding sentence length, with the census counting prisoners where they live on that particular day.

A few groups are trying to get it changed so that the census counts a prisoner at their last known address before going into prison, although that seems to be an imperfect solution as well.

[Broomstick]

Well, I got notified in the mail that I'll be getting the long census form. I'll let you know whats on it.

Broomstick, are the long forms randomly distributed to utilize sampling, or are they targeted somehow?

Don't know for sure, but I think it's random.

Correct. However, the Census is not counting JUST traditional households - the Census also attempts to count people living in weird or illegal situations, or who are just plain homeless. As I said, such people do not get the standard form since they don't have normal mailing addresses. Such people are visited by Census employees utilizing different tools. I'll be finding out what, exactly, those tools are in about two weeks because - YAY! - the Census called me last night and that's when I go for training.

How much does the census expect the foreclosure crisis to impact their contact rates, especially in the hardest hit states? I was trying to find some sort of analysis of contact rates and really couldn't find much. Is there a report out there that covers that sort of thing?

Well, it's certainly likely to affect home ownership rates. On the other hand, a lot of people remain in their homes during the foreclosure process, which can take up to a year (more, sometimes) though it's usually shorter than that. If there's a report of some sort out there talking about that I'm not aware of it. It may be that such a report can't be written until the information is collected.

My understanding is that counting prisoners where they are held has been standard operating procedure since the beginning, but prisoners used to be a lot more dispersed and thus it was less of an impact. There is also the issue of, if you don't count them where they live, where do you count them? And do you distinguish between those with short sentences (under a year, for example) who arguably are only temporarily relocated vs. those with lengthy sentences (10, 20, 30 years, life, etc.).

What I was reading was suggesting that a few districts in New York would have to be redrawn if the prison population wasn't counted. This Washington Post article describes those NY districts as being rural and conservative, and if they had to redraw those districts they might change their political composition a bit. From my understanding there isn't much of a distinction made regarding sentence length, with the census counting prisoners where they live on that particular day.

A few groups are trying to get it changed so that the census counts a prisoner at their last known address before going into prison, although that seems to be an imperfect solution as well.

Not just New York - prisons tend to be located in more rural/conservative areas although there are, of course exceptions. The Census is interested in where people are actually living, not where they used to live, so in that sense the current method is correct. The issue about counting prisoners where they used to live is politics. That doesn't invalidate the concerns, just pointing that out. Handling the details on going by where somewhere used to live is problematic.

An alternative solution is to let prisoners vote - something many Americans find upsetting on a visceral level although I'm told it's common practice in some other countries. My, my, my - wouldn't a bunch of prisoners voting skew the results in some of those otherwise low-population conservative districts? Hm....