Thursday, April 1, 2010

A Case for the 2010 Census, from a GIS perspective

Yesterday, I mailed in my first census.

For myself (and some of my housemates), I filled out 10 questions, regarding our ages and ethnicities, that the government is required by law (see US Constitution: Art. 1, Sec. 2, Clause 3) to ask me every ten years or so (notably, I am also "legally required"-- as a US resident--to answer these questions). Overall, I'm happy to report is was a positive experience. Beyond taking exactly 20 minutes to fill out (longer then average--but shorter then that Facebook quiz I took to find out whether whether my musical choices made me an 'Emo' or a 'Hick'), I learned how to spell my Filipino roommate's last name, and I feel like I significantly contributed to the future of my community (...no really, I feel like I did just that).

This marks one of the few cases of my adult life that I was actually happy to oblige the government with what it was asking of me (I doubt I'll feel the same towards the IRS on the 15th, but that's another story). In fact, I was so eager to help my government, I scrounged 1.75 from my broke self (thanks again IRS) to encapsulate and mail the thing (First Class) to the National Processing Center in Jeffersonville, Indiana (yes, I lost the paid-postage envelope that came with the form). I do not--in ANYWAY--have second thoughts about supporting the 2010 US Census.

...however, there are some among us whom apparently do.

In February, a poll by the Pew Research Center found that nearly 1 in 5 people were not willing to fill out the Census (notably: 1/4th of those 'no' responses said their refusal was due to "distrust of government or concerns about privacy".) In web surfing, I uncovered a lot of evidence that corresponded with the poll's findings. Just by goggling the words "scared" and "Census" and "2010", I wandered across a whole lot of forum posts and news sources (note: if you consider Fox.com a real "news" source) dedicated to voicing doubts, fears-- and even anger--at the 2010 Census and the Government's (alleged) intents with collecting the data. Among some of the most note-worthy comments were:

I will not fill in names or ages of my children. The last thing i want is some pedophile working for the census bureau having my address and the choice of ages that he likes. Yes, I have a tinfoil hat, no, I don't care that people think I'm crazy.


...Why does the Census care about race? How about questions like education, health, etc.. Why so much on race? I thought we were supposed to be blind to race. I smell a big pile of politics.


...and my personal favorite (mostly because it is soooo over-the-top and devoid of any logical, factual, and reasonable thought):

...Above all, ACORN is pulling the strings! They're secretly behind everything, and they're going to use the census to bring down the country! DON'T TREAD ON ME, ACORN! I may not KNOW very much, but based on what little I know, or think I know, let me tell you, I am angry, angry, angry! Very angry! I want my country back, Acorn! AARRGHH!


Okay people....where to start?

Well, for starters, let me just say that I have a healthy amount of respect for those who take the time to question authority (be it government or otherwise). However, there is a difference between 'questioning' authority and bringing to the table a fact-based inquiry of the system....versus indulging in some paranoid 'over-the-top' rant of things that you (clearly) don't understand and don't appear to want to understand.

Overall, it seems to me, that the Census Bureau is facing the same problem that NASA is: bad PR.

The question is simple:
Why do so many people distrust the Census?
Answer: ...because very few people know what the Census is used for.


Allow me to enlighten you a bit.

The Census is a multi-year effort to create a cumulative data source that accounts for the number and demographics (i.e. age, race, and on a more general scale income) of our entire population. As anyone who has worked with large-scale and multi-year data collecting will tell you-- an effort this big is bound to have some complexities and flaws.

For instance, the "race" categorization.

Over the years, our definition and understanding of race has changed and evolved to levels that cannot easily be defined within check-boxes 1 or 2. Who in this country-- at this point--doesn't epitomize the American melting pot? I, myself, can name 4 "racial" definitions that could apply to me (i.e. Caucasian, Pakistani, Middle Eastern, Asian-Pacific). Race is not a simple question anymore, and perhaps, it never was.

However, we are talking about standard data entry here. We are also talking about data entry over a large time scale, where we hope to compare previous data sets (i.e. previous census results) to current data sets. Sure, I find the Census term "Negro" offensive and wildly out-dated. But I have no doubt that it is being used-- not because the government is interested in Civil Right's abuses--but because data entry is one particularly tedious and laborious task. How many Census years have African-American's defined themselves (and have been defined by others) as "Negroes". Keep in mind that the first census was taken in 1790. I'd bet the fill-in-the-blank "applicable race" section of the census saw that term many times over many different time periods**.

It is a common question we run into in GIS: How do we 'deal' with our data? How do we go about collecting our data, organizing it, and distributing it? How do we--from start to finish--create a framework of research and data collection from which we can create the most unbiased, most accurate--but still incredibly generalized-- analysis possible?

It's not an easy question to answer, and I have faith that the Government struggled just as long as I have in creating a data-collection process that synthesized the wide-variety of answers into a few clear (but generationally comparative) check boxes.

As for the (paranoid) insistence that the Government is "out to get us" with the Census. All I can say, is I doubt that is true. Census workers, just like all government employees, have to go through a rigorous (and red-taped to the tenth degree) background check. If you've ever applied for a government position--you know what I'm talking about.

Furthermore, as a GIS tech, I am privy to one of the many positives that the US Census data provides us: the Tiger/Census files. Recently, these (free) files helped out a non-profit group I interact with (and volunteer services for) occasionally.

This particular non-profit group works with Bay Area teenagers, specifically, teenagers who are suffering from a wide range of disabilities: including severe asthma and chronic lung diseases. The teenagers are almost all African-American and Mexican-American--there is not a single Caucasian student among them. Recently, authorities have reported that diesel emissions are a leading cause of childhood asthma and cancer.

The non-profit asked me if I could map the dominant Under-18/race types (i.e. White, African-American, and Mexican-American) from the 2000 Census and overlay it with their refineries and port cargo facilities data (which they knew to be--based on topography and emissions data--the dominant contributors of diesel exhaust in the Bay Area). I did that and drew a 5-mile buffer around the facilities to visualize the density of race types in the facilities neighborhood. The results were sobering.

Here is the UNDER-18/Caucasian map:





...versus the UNDER-18/African-American:



...and the UNDER-18/Mexican-American:



Without the 2000 Census, there would've been no way for us to (reasonably) assume and/or argue that the causes of these kids illnesses could be due to their close proximity to the factories. Furthermore, there wouldn't have been much reason to even suspect that these illnesses were not (just) due to genetic, racial, or even age susceptibilities.

Just think: if every parent who provided that 2000 Census information (used here) left their kids age and race off the Census form...we'd be looking at a much different series of maps. It's possible we'd also be running the risk that these sick kids would never know the true cause of their illnesses. Nor would the government know how to help them, and identify the other 'at-risk' communities in their neighborhood.

The Census is a big deal, people. Not just to the Government (and your Congressman, who wants to keep his/her district), but it is a big deal to you, and for me.

Do yourself a favor: drop the fear, and fill it out.

**True, with database technology these days, it would be feasible to do a (so-called) "Find-and-Replace" query to re-classify the race "terminology" of decades of census data. However...in doing so, you run into numerous possibilities of errors (for ex: how to re-classify, or even recognize, misspellings). With database entry, one must always carefully consider the risks of introducing statistical errors into raw data versus the advantages of changing 'offensive' terminology.

1 comment:

  1. I admire your point of view on how the census data is relevant in issues of health studies. That isn't something I considered. Personally I found it offensive that the only thing that was on this census was age and race. To me, i don't like the idea of the government having access to racial data in the event that they engage in gerrymandering or just adjust policy and run candidates based on who they think would be popular in certain states, etc. This information, like anything is just plain data and nothing has to be implied by it to begin with. But why is that the only thing on there really? If in the future if there is a census that collects blood samples from everyone in order to map and understand our genetics vs health statistics etc. I'd be all for it. (I don't think anyone else would be though). I was so miffed by the question of race that i put down other: "Human Race". To me what you identify yourself as isn't as important as real genetic or health information. But that information would be more vulnerable to abuses as a whole. Thanks for pointing out a real life application for race as a data point. I would be less suspicious of answering the census if they had more diversity in their questions so that there were many uses to everyone other than maybe understanding where racial groups are living. To me that can give politicians a chance to manipulate or take advantage of voting districts.

    ReplyDelete