American Census Bureau

May 27, 2005 3:00 PM

I received a huge questionnaire in the mail from the US Census Bureau. It says there is a legal penalty if I do not fill it out. I do not like ANYONE...even our trusted government asking all this personal info. What will happen if I do not fill it out, will the Feds show up?

It is very detailed and asks for my name, income level, and a whole lot of other very personal information that I am supposed to mail back to them (24 pages that they say takes 38 minutes to complete). I have no reason to think that the form did NOT come from the government. The packet came with extensive documentation including multiple colored pamphlets that I know would be very expensive for the average phisher to bother putting together...especially when there are more efficient and cheaper ways for them to operate. However, I just do not like the idea of this. Digging through the details of the info, it says I have been selected as a sample of my community, which implies that not everyone got one of these.

I simply do not believe it is anyone's business to know things like, how much money I have in retirement or if I have trouble bathing and dressing??? I mean, if it's a census...COUNT ME...don't audit me.

posted by SparkyPine to law & government (43 comments total)

According to this, you are required to respond. Also take a look at Are You In a Survey?, or I would just contact your regional census office.
posted by istewart at 3:09 PM on May 27, 2005

From the US Census site...
"Every question in Census 2000 was required by law to manage or evaluate federal programs or was needed to meet legal requirements stemming from U.S. court decisions such as the Voting Rights Act. In addition, the data collected by them is as much a part of our nation's infrastructure as highways and telephone lines. Federal dollars supporting schools, employment services, housing assistance, highway construction, hospital services, programs for the elderly, and more are distributed based on census data."

Not sure I like it but at least that is the explanation.
posted by sexymofo at 3:10 PM on May 27, 2005

if it's a census...COUNT ME...don't audit me.

Well it's not just a head count: the government needs that demographic information too, so that when congress decides to redistrict again they can place you in a pool of voters who reflect their party's best chance of gaining seats a common set of needs and interests.
posted by PinkStainlessTail at 3:14 PM on May 27, 2005

This is unbelievable. I was going to say "lie," but apparantly there's a $500 fine. I would def. call your regional census office and tell them you are uncomfortable filling out such a detailed survey, and politely requesting not to fill it out. After that, I don't know.
posted by muddgirl at 3:18 PM on May 27, 2005

I received one of these, too. Didn't you also receive, about a week ago, a letter from the Dept of Commerce saying the Community Services Survey was on the way?

I'm usually pretty paranoid about stuff like this, too, but the feds already know a whole lot more about me just from my tax return than they're learning from this survey. Moreover, I'd rather they think I'm a cooperative good citizen (heh) than give them any reason to scrutinize me more closely. Maybe I'm deluded.

Anyway, I filled it in and sent it back. The most time-consuming part of it was finding the utility bills from last month.
posted by Alylex at 3:28 PM on May 27, 2005

If you really feel strongly about this and don't get a satisfactory answer from the Census Bureau, you might try to contact your Rep or Senator and see if they can intervene on your behalf. Acting as an intermediary between their constituents and the federal agencies is one of the things they're often good at.
posted by HiddenInput at 3:44 PM on May 27, 2005

Well, do you have trouble bathing and dressing?
posted by xmutex at 3:49 PM on May 27, 2005

I don't think the census will actually arrest you; you’re just a number they are collecting like a pushy collection agency.
If they have a problem list, I bet I’m on it. When asked my color: I told them the choices did not match me. Then their race question is racist. Think I had just joined Mefi too.
posted by thomcatspike at 3:51 PM on May 27, 2005

The below information should be placed above my previous comment for better reading.

They came by my home several times after it was not turned in. Then called me until I would talk to them, which was the only way to end their annoying calls.

iirc, I was picked to fill out the longer book version. I was surprised by their aggressiveness so I only answered the questions about me that they had previously recorded, like my ss# or equal.
posted by thomcatspike at 3:58 PM on May 27, 2005

You know, answering indepth census questions helps researchers who are trying to help everyone else, like finding out how many people have disabilities, etc.
posted by jb at 4:11 PM on May 27, 2005

was it sent certified mail?

If not, throw it away, and if they hound you tell them you never got it.

(this only works if you haven't contacted them already)
posted by CCK at 4:27 PM on May 27, 2005

I don't think there is anything especially nefarious about the census and its data is extremely useful. But, lots of people never fill it out and nothing happens because it would be a gigantic pain to track down everyone who hasn't filled out a census and prove that they didn't. So you can always take that chance. But, seriously, private companies probably have way more important and personal data on you already.
posted by Falconetti at 4:36 PM on May 27, 2005

I've seen how useful and essential historians, geneologists, anthropologists and the like have found old birth, death, marriage, and census records to be. Think how much more we'd know about the past if they'd made careful and detailed records of demographic information.

I remember during the 2000 Census, a lot of emphasis was put on collecting information about serially under-counted groups. Towards this end, the CB spent a lot of time assuring people who are traditionally suspicious of the goverment about the privacy protections that surround the Census (e.g. the 72 year delay before individual answers are made available). I think the census folks understand that the only way they're going to get the cooperation from the public that they need to do their jobs right is by being completely scrupulous and circumspect with the data they're collecting.

EPIC has a good summary of the privacy risks involved with the Census, and I think they're pretty neglible, as far as individual privacy is concerned. They do mention that "The Washington Post quoted Former Senate Majority Leader Trent Lott [as] encouraging citizens not to answer invasive questions." If Trent Lott is against it, myself, I'm all for it.
posted by tew at 4:49 PM on May 27, 2005

My wife was an enumerator supervisor during the 2000 Census. Folks who didn't respond get a visit; those who still refuse get a visit with an accompaniment from the Highway Patrol or State Police; those who still refused got bench warrants. I don't think I heard of more than one or two going that far though.
posted by luriete at 4:55 PM on May 27, 2005

As thomcatspike notes, there's a long form and a short form. Everybody (well, every household) gets the short form, and a randomly-selected sample gets the long form. And, as should be clear, the Census Bureau is interested in a lot more information than just the head count.

I shred my mail, pay for nearly everything with cash and intentionally maintain a telephone number with somebody else's name on it, and I don't have any beef with the census. They're not particularly interested in the personally-identifiable portion of your data, and their standards of confidentiality are excellent. Is it just that you would prefer to keep a low profile? If so, the best advice is probably to fill out the form, rather than calling your Congressperson or whatnot.

In answer to your question, though, it's probably fairly unlikely that you'll be prosecuted, but it's extremely likely that, if you don't fill it out, census-takers (as opposed to, y'know, FBI agents) will be calling and stopping by and so forth, in an effort to persuade you to complete the thing.

On preview: What tew said. And what falconetti says about tracking people down doesn't square with my experience. I distinctly remember census employees knocking on doors and leaving flyers for the other residents of the apartment building I was living in during the 1995 and 2000 censuses.
posted by box at 4:57 PM on May 27, 2005

Oops, I missed luriete's comment. It sounds like it's not so much that the Census Bureau is unwilling to prosecute, but that they will exhaust every alternative before doing so (it also sounds as if a visit from a state trooper is enough to make many people reconsider their objections).
posted by box at 5:01 PM on May 27, 2005

Oh for crying out loud, answer the questions. It's the law and it's for the public good. The data is anonymized and if they really wanted to investigate you (are you a terrorist?) they could find out all this and more rather easily, but at much greater expense. Do yourself and the rest of us citizens a favor and just answer the questions please.
posted by caddis at 5:12 PM on May 27, 2005

HiddenInput: This is not something a congressional office is going to involve itself in. If you Soc Sec check is lost, sure, but not when you can't be bothered to fill out some forms. It's a little like jury duty. When you're number's drawn, you gotta do it.

I got one of these last year and it kinda sucked and was a bit personal at times. I didn't dig out any old bills, just guestimated, and no one called me on it.
posted by jaysus chris at 5:12 PM on May 27, 2005

Not that I'm advocating this, but: there is the option of filling it in, using made-up information. "I am 96 years old and I live with six unrelated women. Beautiful women, by the way." Have some fun, as George Carlin says.

Note, by the way (as no one else has) that this is 2005, and that the decennial census is five years away.
posted by yclipse at 5:21 PM on May 27, 2005

Last census-year (Good ole Y2K), I declined to participate. It helped that I was coming off a transient spell wherein I arguably resided in at least seven states. They seemed not to notice. Were I in your situation, I'd ignore the survey until it became clear that someone with authority noticed. Then I'd fill it out.

As an aside, a professor of mine worked in public libraries in Alaska where one of her duties was to run around the woods demanding that the cabin-folk fill out their census surveys. She was shot at on more than one occaision and, she told the class, was not above bringing the sheriff out to get the survey completed.
posted by stet at 6:19 PM on May 27, 2005

You're talking about the American Community Survey, I'm guessing. It looks pretty much the same as the "long form" of the 2000 Census, but they're calling it something different. Here are pdf's of the 2005 ACS and the 2000 long form for reference. For what it's worth, there were a spate of articles in 2000 about (mostly Republican) opposition to the long form, which (mostly Democrat) suporters saw as political opposition to simple accuracy. I'm not so sure. Here's one from March 2000 that gives a lot of info, noting the form goes to one in six households and quoting then-candidate Bush:

"We want as accurate a count as possible, but I can understand why people don't want to give over that information to the government," Bush said Thursday during a campaign stop in Milwaukee. "And if I have the long form, I'm not sure I would, either."

The Texas governor stopped short of advising people not to complete the form. But he said, "If they're worried about the government intruding into their personal lives, they ought to think about it."

The non-partisan question is this: Does an accurate count for purposes of distribution of federal funds *really* require information about the "emotional condition" of folks in the house who might "have trouble learning, remembering or concentrating"?

The data is anonymized

What? The 2005 form's mailing instructions clearly state, "make sure the barcode above your address shows in the window of the return envelope." How on earth is that anonymized?
posted by mediareport at 7:15 PM on May 27, 2005

Oops that first link to the ACS form should be this.
posted by mediareport at 7:32 PM on May 27, 2005

The Census Bureau has changed its approach. The short form will continue to be used for everyone every ten years (so, next up, 2010). But the long form is being sent out every year to a much smaller number of people (around or under 1 percent of households).

So yes, with the 2000 census, there often wasn't a lot of followup for those who didn't fill out the long form. But since the long forms being sent out now are for a much smaller percentage of people, and the Census Bureau essentially has ongoing staff working on this (as opposed to ramping up and ramping down every ten years), it's not surprising that they do a lot of followup for those who don't respond. It IS important to try to get a high response rate when the sample is much smaller.
posted by WestCoaster at 7:58 PM on May 27, 2005

The census bureau does more than the census, they collect a lot of the data that we (researchers, social scientists) use to understand what's going on in the country. No one wants to know *your* personal stuff, but it is important to understand what the overall patterns are and the only way to get the patterns is to aggregate from the individuals.

Information that is very identifying (e.g. addresses, names etc.) is removed before the data are analyzed. Of course if someone knew you and every detail of your life, they could probably say "Makes 45,687, pays $48 for gas in May 2005, has 3 kids who he talks to on average 4 times a week, and commites 12 minutes to work??? That's Bob!" Well guess what? They worry about that, too. That's why (by computer algorithm) individual variables are altered (so for example they would change it to say you commute 7 minutes to work). Because of the large N's changing a variable here or there, especially when you change only a couple per respondent and change different variables for each respondent, won't affect correlations much.

Generally for any individual-level data that still has cencus tracts or blocks attached (so there would be some vague idea where you live) there are very strong precautions. For example, when I've seen data like this used it's had to be on a computer that is password protected, and located in a locked room, which no one has access to except the person authorized to use the data and the computer cannot be connected to the internet or any network.

Getting a high response rate is absolutely crucial to getting accurate results. That's why there's a penalty. If you don't participate it results in misinformation about what's going on in the country.

Yes, it is important to know about the emotional state of people who have trouble learning and concentrating. Every question in a survey like that adds at least 10K to the cost of adminstering the survey, which means (believe me!) every question is debated and has gone through a long culling process.

What is that question useful for? Not my area, but I would guess that there is concern that disorders involving learning and concentration can correlate with high levels of frustration and depression *under some conditions*. Understanding what these conditions are can show where money should best be spent to alleviate problems associated with those disorders. Without the research, you'd just be throwing money into the wind.

So I don't know what the penalties are, but *please* just fill out the survey. If you really want to know what the data is used for, go on google scholar and enter the name of the survey, I'm sure you'll find some papers you can look through. If you're worried about how the data are anonymized, call and ask them to explain their procedures.

Believe me, just because they're asking about your stuff doesn't mean there's a single person out there who actually cares about you or your personal information. It's all going to be used for aggregating and nothing more.
posted by duck at 8:23 PM on May 27, 2005

That is indeed the American Community Survey, and I just wanted to tell you that you're not alone if feeling like its awfully detailed and intrusive. I work for a citizen's action non-profit and have gotten easily a dozen calls in the past two days from Maine residents who got the survey and are completely freaked out about it. Note for other posters that this is very different from the "long form" that you might have seen in years past.

We're currently recommending that you contact your Senator and US Representative. The Census Bureau claims that the answers you give are covered under U.S. Code Title 13, the same as "regular" Census answers are, which I currently have no reason to doubt. The Census Bureau does have an excellent privacy track record, but I would encourage you to make your concerns know. I've never actually heard of anyone being fined for failing to respond to a Census survey, but the law is on the books, so anything is possible.

Sorry I can't be more help.
posted by anastasiav at 8:26 PM on May 27, 2005

You know, I just looked at the survey and umm...maybe I'm a jaded sociologist, but it doesn't look so bad. It seems pretty obvious to me why they would want to know these things. These are all questions that essentially speak to what services a person is likely to need and what their economic situation is.

*Of course* the government needs to understand the economic conditions across the country and the services that people in different areas need. How else can they possibly be expected to provide services or to evaluate the effects of their policies?
posted by duck at 8:49 PM on May 27, 2005

I know some people who work for the Census Bureau. Their privacy standards are quite stringent. I am pretty strict about who gets my personal information, but I wouldn't have a problem filling out the long form.

That said, if you are really against it, I assume you can contact someone at the agency to get out of doing it.
posted by bedhead at 9:56 PM on May 27, 2005

While I love the once-a-decade Federal Census for all the wonderful information my grubby genealogically-inclined hands can wring out of it, I do have to take issue with the history of privacy protections of the accumulated data.

Specifically, you may not want to answer the ethnic background or ancestral nationality question. During WWII, Federal (and possibly local?) census data from 1940 and 1930 was improperly used by the US government--long before the 72-year moratorium was up, of course--to help identify people who were Japanese and Japanese-American. Those people were sent to internment camps. Fom EPIC's privacy page, linked above:

"It has been recorded that even before the Japanese attack on Pearl Harbor, President Franklin Delano Roosevelt ordered the Census Bureau to collect information on "American-born and foreign-born Japanese" from the Census data lists. Information was gathered from the 1930 and 1940 censuses on all Japanese-Americans and then given to the FBI and top military officials. These sources point directly to the census information as one of the reasons that led to the internment of almost 110,000 Japanese-Americans on the West Coast, two-thirds of whom were U.S. citizens."

EPIC doesn't list it, but it is also speculated that the 1910 Federal Census data was used in the smaller-scale WWI internment of Germans and German-Americans. And I really wish I had a link to give here, but I have read news stories, which have circulated among the genealogical community, that the current government has been using data from the 2000 census--in the aggregate only, or so they say, and thus legal--to identify neighborhoods with high percentages of people with Pakistani, Saudi, Syrian, and other predominately Arab or Muslim backgrounds to increase "outreach" in those neighborhoods. Your fear is as good as mine as to what that could mean for the future.

Also, I'm not sure about the 2005 ACS, but your responses to the once-a-decade census will eventually be publically available and thus visible to your descendants someday, who might be looking up your data as they do research on their families. So feel free to write "hi, grandchild!" in the margins, if you want.
posted by Asparagirl at 10:25 PM on May 27, 2005

You know, I've always understood the census and felt okay about it. But then, last time it was an issue for me , the government still resembled an American government, still seemed to respect the law more than not. Times have changed, we have single-party rule, and that party doesn't seem to care about pesky things like laws and rights.
posted by Goofyy at 10:32 PM on May 27, 2005

Information that is very identifying (e.g. addresses, names etc.) is removed before the data are analyzed.

Oh, please. The facts are simple: At least one government office is able to connect your name and address with the data you provide on the ACS. A citizen can *decide* to trust that this particular office will treat his/her privacy as an inviolate trust, but that hardly means it's true in all cases. Join the real world, you know? But don't assert that the data is anonymous (at best it's confidential) and expect everyone else to follow along.
posted by mediareport at 11:24 PM on May 27, 2005

Of course if someone knew you and every detail of your life, they could probably say "Makes 45,687, pays $48 for gas in May 2005, has 3 kids who he talks to on average 4 times a week, and commites 12 minutes to work??? That's Bob!" Well guess what? They worry about that, too.

This is actually a rather interesting mathematical problem, which I wish I could remember the name of so as to provide links. (IANAstatistician) One version of it doesn't have you change the 12 minutes to 7 minutes, but rather limits how many categories you're allowed to intersect at one time.

A bad case: say you are allowed to ask not only "How many Kuwaitis" and "How many kickboxers", but "How many Kuwaiti kickboxers", and find out there's only one in Scranton, PA. If you're then allowed to ask "How many Kuwaitis that eat kangaroo", and get zero, you've learned that the kickboxer doesn't eat kangaroo, and can probably find out much more about her with other such questions. There are, apparently, well-defined statistical methods to prevent such identifications.

This is probably orthogonal to your privacy fears, which I completely respect. I just find it interesting.
posted by Aknaton at 11:32 PM on May 27, 2005

What Goofyy said. In the last 3.5 years I've seen so many governmental abuses I previously never thought possible in the land of the free and the home of the brave that it really gives me pause about blindly trusting the government on anything anymore. Which is very, very sad.
posted by grouse at 5:09 AM on May 28, 2005

But don't assert that the data is anonymous (at best it's confidential) and expect everyone else to follow along.

I didn't assert (nor did anyone) that the data is anonymous, I asserted that the data is anonymized. It is put through a process that turns it into anonymous data. Your survey that you mail in is not completely anonymous, it is (as the census website says), confidential.

Though I wouldn't be surprised if there were procedures in place so that no single person ever knows both your identification (name address) and your survey answers. Using barcodes (instead of names/addresses) as someone implied this survey does, is one way of setting up those procedures.
posted by duck at 7:07 AM on May 28, 2005

...I'm pretty sure kangaroo would not be halal and that most Kuwaitis are Muslim...
posted by alumshubby at 7:31 AM on May 28, 2005

You know, I usually fall on the side of more privacy = better. But in the case of the Census, I think that the public good outweighs other considerations. My reasons for feeling this way have been amply outlined in the above posts.

Fill it out, and thanks.
posted by sic at 8:57 AM on May 28, 2005

...according to Google, Qatar Airways serves kangaroo in their halal meals. See also Halal Helpline.
posted by Aknaton at 9:45 AM on May 28, 2005

Here in the UK we take the whole Census thing a lot less seriously
posted by Lanark at 12:07 PM on May 28, 2005

See also this article on use of census data in human rights abuses worldwide, including in the US on multiple occasions.
posted by IshmaelGraves at 4:09 PM on May 28, 2005

I'm just surprised (as a Canadian) that you can be fined or pursued legally for not filling out this form.

I do agree with the others that you likely have very little to fear from filling it out, and that it does help your country in various ways. Indeed I've received the long form census survey here in Canada and had no problem filling it out. But there was no legal threat involved. Indeed, just seeing a threat in fine print would make me more inclihned to not fill it out. So I understand your fear.

But really, filling it out is the easiest solution.
posted by aclevername at 7:15 PM on May 28, 2005

Have you thought about taking the Fifth on any questions you find invasive? I hear it can be done on income tax returns.
posted by glibhamdreck at 7:33 PM on May 28, 2005

In Canada the legal requirement stands, but the privacy situation is entirely opposite the US American one.

First, the Edmonton Public Library tells us that Canadians are legally required to take the census:

The taking of census did not become a legal requirement until 1867. The first Dominion census was taken in 1871. A full census has been taken every ten years since then.

The first census in Canada was taken in 1666 but many of the early censuses were merely head-counts and contained no names. Returns prior to 1851 are rarely complete for any geographical area and list only heads of households. Returns after 1851 list every person in the household with age, sex, country or province of birth, marital status, education, religion, and occupation.

Privacy-wise, it is currently illegal to share the private (non-anonymous) information of any post-1901 Census. Statistics Canada refuses to release the records to the public archive. Several court decisions indicate that Statistics Canada is correct in its interpretation of the law.

There is an a movement to open these records, mainly by the genealogical/family-tree crowd. I didn't spot any recent news on their progress.

Personally, I'd be happier were the records to remain closed. I prefer we keep our privacy rights rock-solid, instead of waiving them where it's inconvenient.
posted by five fresh fish at 7:41 PM on May 28, 2005

Trust me, no one can get at your information, except the government. And they can find you much, much more easily than through the long form census.

I know Canadian researchers who worked with longitudinal health data. They had to sign tons of forms to have the security clearance to even use the databases, and had strict guidelines on what they could publish (nothing with very low numbers, like less than 5).

Everything duck has to say is very true - I've pulled those articles based on census data, and used to work at a research unit that used both American and Canadian census data extensively. No privacy was ever threatened - there was no possible way for any of us to identify anyone in the survey, and still there were all sorts of strict rules about privacy.

As for ever being release, in the UK censuses are not released for at least a century (1901 just recently opened), and it may be that they decide they will never release any non-anonymized health information from the twentieth century censuses.
posted by jb at 4:59 PM on May 29, 2005

I just went through the process of submitting the ACS form.

I agree with the above posters who point out that (a) census data, including the stuff they ask on this form, is vital for understanding what's going on so that social programs etc. can actually reflect the needs of the populace, and (b) this is probably not the most important way -- even close to the most important way -- that our privacy rights are being trimmed away. And that to the extent that it is an intrusion, it may well be justified by (a).

But I did hate feeling threatened. When I sat down to fill it in the first time, it was clear that to follow the instructions and fill it in accurately for my family was going to take a large amount of time and lots of number-crunching -- they ask for things like each family member's income over the last twelve months, a different number than what we filed, for example, with IRS. I don't have the foggiest idea what my self-employed partner made during that time.

And there was no indication that I could skip or even estimate on any question. So I put it aside, grumblingly, and figured I would do it when I got to it.

A very ominous you-will-be-fined letter showed up in my mailbox about three weeks later. We were outraged and a little flustered, and I got my back up and considered writing Congresspeople and writing "under protest" on every page.

A week after that -- I was still dithering about what to do -- I got a phone call. Why haven't you filled it out? Can you do it now? I sputtered my various discontents to the sweetly insistent woman on the other end of the phone, who then said that she could take my responses over the phone and it would be just fine. We scheduled a call back, and at 8 PM that night I did this in 15 minutes by phone.

It might as well have been a different survey -- we sped past stuff that didn't apply, she assured me that all they needed were approximations of most of the quantitative entries, and that of course I didn't have to answer any question I didn't feel comfortable about!

So, in the end, I was rewarded for dragging my feet and being a Reluctant Citizen, and I don't feel particularly grand about that. But I am perplexed that the survey doesn't make it clearer that one can "opt out" of certain answers. And that they don't get that if they hadn't threatened me with a whopper of a fine I would have felt better about doing it. I'm the sort of person who thinks that weaseling out of jury duty is inexplicable withdrawl from social obligation. But something about the way this whole survey was framed set me at odds with it...
posted by BT at 5:32 AM on May 30, 2005