Census Answers to Race

[Uncayimmy]

You're making enough for everyone. You should slow down a bit and give the other protesters a chance. Be fair.

Stop bullying me!

It doesn't exactly work like that, not at the local level. How it works at the regional or central level, I wouldn't know, not being party to those levels.

Thanks for taking the time to tell me you don't know.

That's the purpose of the April 1 date.

Really? Wow! I'm so glad you told me that.

But most dupes are simply address duplications, and we just eliminate the dupes as we find them. It happens. It's a big country.

So, you don't know what frequency?

We don't do it like that. No one at the local level calls you to get your answer on the questions you didn't answer, or refused to answer.

So, why did they ask for my phone number and tell me that they would use it to call me in case they had any questions? And how come it seems that Census workers are unaware of what it says right there on the form?

It's okay for us to try to be sure we didn't miss anyone, isn't it?

That's totally unrelated to what I was asking, but nice attempt at being snarky.

If you had a clue how it works, you wouldn't even think of asking that. The actual methods we do use work well.

If you had a clue, you would understand that the system "working well" is irrelevant to whether it could work just as well without the information in question. I mean, the Census could ask for a naked picture of each family member and continue to do the bang-up job they currently do. Their "success" is no defense for asking to see UncaYimmy's Country Fried Love Steak.

I'll try to explain this so that you can understand. Full names, DOB, and phone numbers are not actually used as data points. Ostensibly they are gathered to enable to the Census to more accurately gather the data they actually have to report. We know, however, that the Census is not 100% accurate. Therefore, there is a judgment call about an acceptable level given resources allotted, blah blah blah. Same old story.

Collecting this data presents risks. Identity theft is a far larger problem than it was even in just 2000. It's like night and day compared to 1990 and earlier. Times have changed. Not only is there a risk of theft of data, but there's the risk of every single worker who sees that data using it improperly. Besides identity theft in the sense of banking and credit, there's the simple matter of passwords. Unfortunately, many people use the names of their children, DOBs, and even phone numbers and addresses as part of their password/PIN schemes. Workers lose laptops. And then you have the fake Census workers working the streets. The risk is real.

The idea is to balance that risk against the needs of the Census. Since the forms are address driven, full names and DOB for every member in the household are not required to detect if a duplicate form has been sent in from a given address. That problem can be resolved with far less information.

So, one class of "dupes" we're trying to identify are those individuals who are listed on more than one form with those forms originating from different addresses. It is not unreasonable to ask how often this happens and how they deal with it in order to judge it against the risks of collecting this information from 280,000,000 people and putting them all at potential risk.

Right now, I don't even know if the Census even attempts to do anything about those cases, much less how often it happens.

At the local level, almost all we check is the address, as it's pretty much all we need to check. I don't know what may be done at regional or central office levels, however.

Thanks again for taking the time to tell me you don't know the answer. Most people would just not answer at all, but it's much better if everyone who doesn't know steps in to make sure we all know they don't know.

 

[slingblade]

You're welcome!

 

[Cavemonster]

1) I never said they would work the same. That's your straw man.

And I never claimed you said they would work the same. That's your straw man.

Without knowing that, I can't really render an opinion on whether they should be collecting names and DOB like they do.

So you're admitting that the opinion you've rendered multiple times is ill informed?
[/QUOTE]

Yeh, I get that. Your situation could be resolved without requiring full name and DOB. Somebody at the Census should have a pretty good idea how many times each decade they encounter problems where full name and DOB are required for resolution.

Way to skip a full paragraph of very common situations that answered your incredulity that confusion between names might occur.

 

[ZirconBlue]

Another instance in which a person could be listed on more than one form: divorced parents who have shared custody. The child(ren) could easily be counted twice.

Really? Then you're in the minority. Go join one of the other census bashing threads where people are complaining about how many they got.

I got an annoying number of mailings from the Census Bureau, but only one of them was the actual form. I got a card telling me than the form was coming, the form itself, and a card reminding me to fill out the form, all in the span of less than a week.

But it is, b. Of course skin color varies. Of course facial features vary. Of course these characteristics are hereditary. That's not race. Race is the idea that the traits mean something beyond themselves. That notion of meaning is 100% socially constructed.

Which, of course, does not mean that the effects of this socially-constructed "race" are not real.

 

[Cavemonster]

I got an annoying number of mailings from the Census Bureau, but only one of them was the actual form. I got a card telling me than the form was coming, the form itself, and a card reminding me to fill out the form, all in the span of less than a week.

My reminder contained a duplicate form, and from people complaining about too many mailings in other census threads, this seems to be very common.

 

[Skeptical Greg]

...Collecting this data presents risks. Identity theft is a far larger problem than it was even in just 2000. It's like night and day compared to 1990 and earlier. Times have changed. Not only is there a risk of theft of data, but there's the risk of every single worker who sees that data using it improperly. Besides identity theft in the sense of banking and credit, there's the simple matter of passwords. Unfortunately, many people use the names of their children, DOBs, and even phone numbers and addresses as part of their password/PIN schemes. Workers lose laptops. And then you have the fake Census workers working the streets. The risk is real....

Good point UY ..


We keep being assured, how the Census Bureau does not use any of this personal information, but yeah, like we are supposed to believe they are using the Mind-Wipe tool from MIB, to make sure that none of the thousands of temporary census workers, decide to swipe an identity or two ...

 

[Cavemonster]

Collecting this data presents risks. Identity theft is a far larger problem than it was even in just 2000. It's like night and day compared to 1990 and earlier. Times have changed. Not only is there a risk of theft of data, but there's the risk of every single worker who sees that data using it improperly. Besides identity theft in the sense of banking and credit, there's the simple matter of passwords. Unfortunately, many people use the names of their children, DOBs, and even phone numbers and addresses as part of their password/PIN schemes. Workers lose laptops. And then you have the fake Census workers working the streets. The risk is real.

A single complete identity, went for around $15 in 2009. If the trend is continuing, it's likely even cheaper now, maybe even less than $10

http://www.associatedcontent.com/article/1405309/economy_identity_theft_and_the_black.html?cat=17

Considering that phone numbers and addresses are already publicly available, you're relying on sorting through a list of names to crack passwords, waiting for the very few stupid enough to have just a kid's name or birthdate as a password to something important (without knowing any usernames or what banks they use or exactly where you're trying out that password)

That's a payout, in the best case scenario, far worse than Mcdonalds for your time and with a large amount of risk if caught.

 

[ThunderChunky]

Anecdotally, in every Internet article, forum post, or blog I've read so far on this tedious subject, without exception, the people who say they've answered "human" or "American" identify themselves as white.

Now you have. Although, why should it matter?

I can only guess they believe that ignoring race and racism will make it magically go away.

Not institutionalizing it would be a nice start.

 

[slingblade]

Now you have. Although, why should it matter?

Now I have what?

Not institutionalizing it would be a nice start.

It would, wouldn't it. What do you plan to do about it?

 

[Uncayimmy]

So you're admitting that the opinion you've rendered multiple times is ill informed?

I have not rendered an opinion on whether the Census should or should not collect full names and DOB. The only thing I have said is that it's not needed for determining if two forms from the same address are duplicates, and I stand by that.

Way to skip a full paragraph of very common situations that answered your incredulity that confusion between names might occur.

At least Slingblade had the courtesy to say she didn't know. I wish you would do the same. You can come up with all the crazy scenarios you want - you still haven't answered my basic questions:

* How often does this happen?

* How, if at all, do they currently deal with these situations?

If you already know this answer, why are you holding out on me? If you're just Googling info and make up stuff as you go along, then just say so. You and others have posted thousands of words and still haven't answered the questions that started it all.

1) Why does the Census ask for my name and the names of my family anyway? I don't get that.

Somebody followed up with the vague claim of "duplicate checking" so I asked:

2) Can you describe how that works?

So far, nobody has described exactly how the Census uses this information to check for and resolve duplicate names at different addresses. Nobody has even presented evidence that they do anything at all about it. For all I know the real reason they ask for it is for genealogy buffs because despite their claims that they keep it private, they actually release the information after 70 years.

The smallest district by population has about 500,000 people, and that's Wyoming at-large. The districts average about 650,000 with the largest being around 900,000. How many duplicate names would it take to affect redistricting?

If you don't know, just say you don't know.

 

[Uncayimmy]

A single complete identity, went for around $15 in 2009. If the trend is continuing, it's likely even cheaper now, maybe even less than $10

http://www.associatedcontent.com/article/1405309/economy_identity_theft_and_the_black.html?cat=17

Is that supposed to mean something? Is the price somebody pays to be able to open a Sears card in my name and buy a washing machine somehow relevant? I guess you're arguing that since the price is so cheap, it means supply has increased tremendously. So, thanks for supporting my contention that it's a problem. I wonder if theft of Census data played a part in that?

BTW, Associated Content is a few steps below Wiki when it comes to being a reliable source. It's not hard to get published there, and there's no fact checking or any real repercussions for being wrong.

Considering that phone numbers and addresses are already publicly available, you're relying on sorting through a list of names to crack passwords, waiting for the very few stupid enough to have just a kid's name or birthdate as a password to something important (without knowing any usernames or what banks they use or exactly where you're trying out that password)

1) People can choose not to have their name and phone number published.

2) Family members including minor children are typically not listed.

3) DOBs are not listed either.

I don't know what point you're trying to make, but the Census has in the past paid for people to have identity theft protection for when their security was breached, such as when laptops have been lost or stolen. If you don't think it's an issue, you're in the minority.

As for ordinary things like passwords, I know the real names of lots of people I've encountered on-line. There are people at the JREF who know my login, e-mail and real name. If they get access to the names of all my family members including DOB, they can sit around and try to hack into the accounts UncaYimmy has scattered around the globe. In my case it won't help them, but for some people, whether you call them stupid or not, it will.

From there it's not too hard to figure out what e-mail server someone uses. Once you have one password cracked, there's a chance they can get into that person's e-mail. If the mail is stored on the server, which is standard practice with systems like GMail, HotMail and Yahoo, what might they find? Bank account numbers? Password resets?

That's a payout, in the best case scenario, far worse than Mcdonalds for your time and with a large amount of risk if caught.

You have GOT to be kidding. Let's pretend it's $15 for a name. Do you think they are selling just one name? Really? Somebody has a laptop full of confidential information, and they are going to sell just one of the 1,000 names they have?

Gimme a break.

 

[Cavemonster]

You have GOT to be kidding. Let's pretend it's $15 for a name.

You misunderstand, it's not $15 for just a name, that's for a full identity including all the stuff on the census form and social security number. Without the SS, it's pretty useless. No one's going to buy a list that they can get almost all of from public sources on the off chance that they can find a password to "something".

And if you don't believe my source, here's Slate magazine who sets $15 as the top value(and that's 2 years old)

http://www.slate.com/id/2189902

This one references a number as low as $2, again with SS number
http://www.theinquirer.net/inquirer/news/1013264/recession-hits-spammers-id

Actual credit card numbers, ready to use, go for 40 cents a pop. You say identity theft is a huge problem, that's exactly why there isn't a large risk of this information being sold, because the black market is glutted with much better information already. If I can get credit card #'s ready to use for .40 a pop, how much am I going to pay for the chance to try to mine... (what exactly? bank accounts? email?) with some brute force algorythm out of DOB and kid's names? Personally I doubt there'd be any bites with SSN's and CC#'s so cheap as it is, maybe few cents each? None of the temp census workers have direct access to enough data to make pizza money, let alone enough to be worth the risk of getting caught.

 

[Uncayimmy]

You misunderstand, it's not $15 for just a name,

Lighten up. I used "name" as shorthand.

And if you don't believe my source, here's Slate magazine who sets $15 as the top value(and that's 2 years old)

It has nothing to do with "believing" your source. This is a skeptics board. If you are going to make factual claims, cite a reliable source. You didn't do that. It seems to me you are a classic case of GAG - Googling as you go. You have a thought, find the first link that supports it, then post it as "evidence" that you are right.

Problem is, your whole argument is ridiculous. Why do you think the Census keeps the information private in the first place? Huh? Are you really demanding that I prove to you that having this information getting into the wrong hands could present a problem? Seriously? Really, I find your argument to be little more than pissing on my leg.

There are numerous ways this information could get into the wild, from somebody selling it to carelessness to sophisticated theft. If people are selling information for a buck, then, guess what? That means people are selling it, which is the only burden I have to establish risk.

Your lack of imagination is no reason to assume that there no other ways to abuse the information. What about the estranged husband who gets a job at the Census and fishes around to find out where his wife his hiding? What about the pedophile who uses little Timmy's personal information to coax him into a car? What about the school that uses a "connection" at the Census to find out the race of applicants so they can exclude blacks while telling the government, "I never even saw the person. How could I possibly have discriminated based on race?"

Tell you what, Cavemonster. How about respond to this thread posting everything you filled out on the Census form. You won't because you're not stupid. There's a reason nearly everybody on this and most other boards uses an alias. Many don't even reveal their city/state or even country, and wouldn't do it if asked. There's a reason people have unlisted numbers.

So, really, cut the ********. This information is protected for a reason, but that protection is not perfect. There needs to be a compelling reason for collecting it. We should look at alternatives. And without knowing how the information is actually used and how often, nobody can say if it's a good idea or not.

Did you not hear about workers with access to passport information accessing the files of celebrities? **** happens.

 

[Cavemonster]

Lighten up. I used "name" as shorthand.

Shorthand for what? For Name, ID info and SSN? Because that's what the figure was for, but it doesn't correspond to what you wrote.

Somebody has a laptop full of confidential information, and they are going to sell just one of the 1,000 names they have?

Nobody has a laptop of names with SSNs.

It has nothing to do with "believing" your source. This is a skeptics board. If you are going to make factual claims, cite a reliable source. You didn't do that. It seems to me you are a classic case of GAG - Googling as you go. You have a thought, find the first link that supports it, then post it as "evidence" that you are right.

No. I have information, but I don't keep a rollodex of where I found it. My actual source for the cost of identities was a radio piece on NPR, when I need to prove something, I'll Google.

Problem is, your whole argument is ridiculous. Why do you think the Census keeps the information private in the first place? Huh? Are you really demanding that I prove to you that having this information getting into the wrong hands could present a problem? Seriously? Really, I find your argument to be little more than pissing on my leg.

There are numerous ways this information could get into the wild, from somebody selling it to carelessness to sophisticated theft. If people are selling information for a buck, then, guess what? That means people are selling it, which is the only burden I have to establish risk.

Your lack of imagination is no reason to assume that there no other ways to abuse the information. What about the estranged husband who gets a job at the Census and fishes around to find out where his wife his hiding? What about the pedophile who uses little Timmy's personal information to coax him into a car? What about the school that uses a "connection" at the Census to find out the race of applicants so they can exclude blacks while telling the government, "I never even saw the person. How could I possibly have discriminated based on race?"

Tell you what, Cavemonster. How about respond to this thread posting everything you filled out on the Census form. You won't because you're not stupid. There's a reason nearly everybody on this and most other boards uses an alias. Many don't even reveal their city/state or even country, and wouldn't do it if asked. There's a reason people have unlisted numbers.

So, really, cut the ********. This information is protected for a reason, but that protection is not perfect. There needs to be a compelling reason for collecting it. We should look at alternatives. And without knowing how the information is actually used and how often, nobody can say if it's a good idea or not.

Did you not hear about workers with access to passport information accessing the files of celebrities? **** happens.

As a note, you criticize my sources and then cite a Fox News story? Are you for real?

But on the broader point, yes, some risk exists, selling bannanas creates a non zero risk that someone will drop a peel and someone else will slip on it.

The examples you cite are as likely and as dangerous.

 

[Uncayimmy]

Nobody has a laptop of names with SSNs.

I hope your "nobody" is confined to Census workers.

As a note, you criticize my sources and then cite a Fox News story? Are you for real?

Huh? Are you suggesting that Fox is news is lying when they cite government reports and AP articles? Are they inventing quotes by senators? Please stop being silly.

But on the broader point, yes, some risk exists, selling bannanas creates a non zero risk that someone will drop a peel and someone else will slip on it.

The examples you cite are as likely and as dangerous.

This is getting tiresome.

The first issue is the potential damage, which can be loosely called risk. There are numerous very serious "risks" if that data were to get into the wild. This is why the Census does not release that information and has policies in place to protect the information. If the entire Census database were published on the Internet, there would be huge problems. You seem to be downplaying this risk, and that's just silly.

The other type of "risk" involves the odds of some or all of the information getting into the wild. First off, there is cyber security. Systems get hacked. It's a fact of life.

Then there are people on the inside. They hired 750,000 temporary workers. They put them through background checks for a reason. Bad people are going to slip through the cracks. Good people are going to be tempted - it's not like the people snooping celebrity passport information were "bad" people that could have been stopped with a background check.

Then there's human error. People make mistakes. According to The Washington Post, 1,100 laptops were lost or stolen between 2001 and 2006. About 250 were from the Census, and that's not even during busy time for the Census.

In effect you've answered my question by using several thousand words to say that you don't know if the Census actually uses the data to detect duplicates from different households. If they do, you don't know the process, and you certainly don't know the frequency. When I suggest that it's worth looking at because if security issues, you tell me that the "risk" is like a banana peel.

Next time just say that up front.

 

[Cavemonster]

In effect you've answered my question by using several thousand words to say that you don't know if the Census actually uses the data to detect duplicates from different households.

Yimmy, this is a lie. I linked you to where they explicitly said they use the data for that purpose.

Are you suggesting that the Census is lying about it's process?

If they do, you don't know the process, and you certainly don't know the frequency.

I'm not sure how the exact process is relevant. It doesn't take much imagination to assume you can run a program that picks out multiple entries with the same name and birthdate. It's likely more complex than that, but it really shouldn't be hard to imagine how it may work. It should go without stating. As for the frequency, that's another silly question, I listed a number of very common situations that can easily result in double counts, but if you absolutely must have numbers,
http://www.jos.nu/Articles/abstract.asp?article=224655

Estimated about 5.8million duplicates out of ~281million respondents

 

[Cavemonster]

Nobody has a laptop of names with SSNs.

I hope your "nobody" is confined to Census workers.

What an odd non sequitur. You do realize that the census doesn't ask for Social Security numbers, don't you? Nobody has a laptop full of names and SSNs because they don't record SSN's.

 

[Uncayimmy]

Yimmy, this is a lie. I linked you to where they explicitly said they use the data for that purpose.

Are you suggesting that the Census is lying about it's process?

Seriously, unclench your cheeks and open your eyes. The Census simply says that it uses names "to help ensure people are not counted twice in the census." I specifically stated that we have no evidence that they use the information to check for duplicates that come from different addresses. We have a first-hand account that they use it to check for duplicates from the same address. Nobody has presented any information that they use it to check for duplicates from different addresses, and I have never once claimed that they don't. I just don't know one way or another, so I asked.

ETA: And I asked this question of a Census worker as an aside. I never asked you to get all Google crazy on me.

 

[Uncayimmy]

What an odd non sequitur. You do realize that the census doesn't ask for Social Security numbers, don't you? Nobody has a laptop full of names and SSNs because they don't record SSN's.

If I didn't realize that the Census does not collect SSNs as part of the Census, I wouldn't have phrased the question like I did. You made the unqualified statement of "nobody" so I was just clarifying that you meant "Census workers." The Census Bureau does (or at least did until recently) deal with SSNs as part of their function outside of the head count.

Furthermore, we were talking about identity theft in general, so you could have meant that no government agency ever carries SSNs on a laptop. No need to get all worked simply because you were unclear, and I was just trying to make things clear.

 

[Uncayimmy]

I'm not sure how the exact process is relevant. It doesn't take much imagination to assume you can run a program that picks out multiple entries with the same name and birthdate. It's likely more complex than that, but it really shouldn't be hard to imagine how it may work. It should go without stating. As for the frequency, that's another silly question, I listed a number of very common situations that can easily result in double counts, but if you absolutely must have numbers,
http://www.jos.nu/Articles/abstract.asp?article=224655

Estimated about 5.8million duplicates out of ~281million respondents

Did you even read your citation? They didn't actually do anything with the information because they have never had the computing horsepower until 2000 to even attempt it. Doesn't that make you wonder why they ever needed it before 2010?

There's also no indication that they are going to do anything with the information this time around because, like I said, it takes manpower to resolve at least some of them. They also didn't address what kind of duplicates they had. You cannot assume that each duplicate pair was from the same address. They could have been of the type already discussed.

So far, it seems that the Census doesn't really need my name.