Three weeks ago we received our 2010 Census form in the mail. The very next day I sat down with Benjamin and we filled it out (after I explained to him what it was, why we had to fill it out, and why we should want to fill it out — the need to do that has some interesting implications, more on that later). The day after we filled it out, it went out in the mail.
Imagine my surprise and frustration, when I received a second form in the mail yesterday saying, essentially “we haven’t received your census form, please fill it out”. What!? It includes this helpful note that “if you’ve already filled it out, no need to do so again”. It isn’t as though the two crossed in the mail, I mailed the first one three weeks ago! If they hadn’t received it yet, they probably weren’t going to get it (thanks US Postal Service)1. And because I really want us to be counted, I filled it out again last night and it’ll go out in the mail today.2
Back to the discussion with Benjamin about the census form. Given his lack of knowledge about the census, including what it was, why we do it, and what the data is used for, it’s no surprise that the Census Bureau is having a hard time getting some people to respond. I think part of the challenge was that he’s never filled out a form before. 10 years ago he was still living at home. I on the other hand filled out my first census form during the 2000 census while I was at A&M. I suspect that the 18-30 year olds who weren’t paying close attention in civics class and aren’t active news readers will have a low mail-in response rate.
And what’s the deal with questions 8 and 9? Just for reference, they are:
8. Is Person X of Hispanic, Latino, or Spanish origin?
9. What is Person X’s race?
Question 8 I get. The responses for question 9, however, seems mutually exclusive for people who mark any “Yes” response for question 8. It took Benjamin and I, two college educated individuals, several minutes of head scratching to figure out what the heck we were suppose to put down on question 9 for him. In the end we threw up our hands, marked “Some other race”, filled in “Latino”, and moved on. If question 9’s instructions hadn’t been “Mark one or more boxes” we would have left it blank for him altogether. I bet that the data gathered from question 9 for people who responded “Yes” to question 8 is going to be complete garbage since the question was complete garbage in that context. I can only assume that a bunch of white people formulated the question’s wording.
1 Interestingly, I received the second form yesterday (March 31st) but the document was dated April 1st. While the US Post Office is unable to deliver my first form from me to the Census Bureau within three weeks, they were able to deliver the second one to me back in time.
2 I’m not worried about being double counted as I’m more than certain the census database’s primary key is our address and they’ll presumably kick out one or the other. There are only two downsides to me sending in the second form, one actual and one potential. The actual downside is that I just cost the government another $0.42, and even that “actual” downside is only theoretical as if they didn’t receive the first form the postal service may not have charged them for it to begin with. The potential downside is that if, for whatever reason, I didn’t fill out the second form exactly the same as the first form (specifically Question 9 about your race confused the hell out of Benjamin and I for him) and they end up getting both forms, are they likely to toss out both results and thus we’ll not be counted at all? After weighing that potential downside I decided that if I were in their situation and that were to occur I’d use some deterministic algorithm to accept one or the other forms (eg: keep the first and toss the second), hence me filling and sending in the second one.