Driving home from the office the other day, I heard a story on public radio about the accuracy of US Census data. The story provided an overview of several metropolitan cities that appeared to be shrinking according to US Census data. The story then went on to discuss demographic data providers who utiltize mutliple means of data collection to provide a much more accurate demographic picture of these areas.
There were several accounts of local government agencies who challenged the US Census conclusions using reports from third-party demographic data providers and won. Since many federal funds are tied to population statistics, these local agencies were set to receive less funding than they should have (according to the third-party demographics), so they challenged the Census figures to ensure that they received their fair share of federal monies.
The whole story got me thinking about how accurate the Census really is. I worked for a County government GIS department when the 2000 Census was being planned, and remember all the work that everyone in the department did to provide the Census Bureau with accurate and up-to-date address locations to help them plan their censusing efforts. We combined the results of a number of different data sources into a comprehensive list that was scrubbed using some AMLs, and in some cases manually checked for accuracy.
Since I was a Geographer by education and a GIS person by profession, I was thrilled about helping the US Census Bureau prepare for the Census. As it turned out, I never received a Census form or a visit from a Census taker. Was I missed again? I’ll never know. But I was disappointed that I never had the opportunity to participate in the actual Censusing either by completing the form or talking to an interviewer.
What made it most frustrating was that in 2000, I owned a home. So I thought “this time I’ll be sure to be included”. When the previous 1990 Census was taken, I was renting a condominium while attending college studying Geography. So I was excited about the Census and ready to participate. Collection came and went, no form and no interview. I asked my landlord if he received a form for the Condo I was renting, thinking that perhaps he completed the form on my behalf. He didn’t. Was I missed?.
So here were are in mid-2009, with the 2010 census being planned. I know local Census takers are running around town, though I’ve not seen any. No one has come to my door yet. Will I get a form this time? Hopefully. I would really be disappointed to be missed a third time.
My experience leaves me to wonder what exactly happens to the data us GISers provide to the Census Bureau? How is it used by the Bureau to determine who should receive forms and who should get a visit from a Census taker? The Census is not a survey, it’s supposed to be a complete Census of all individuals in the US. So ideally no one is skipped. So, how is it that I’ve been missed in the last two censuses?
Has the Bureau found another way to obtain demographic information on me such that they don’t need to send me a form or direct a Census taker to my house? Surely they can learn some information from County assessor records, but this doesn’t include the other sociological and economic information they collect. Are they getting this from other data sources? What is the statistical possibility that I was missed in the last two censuses? What does that say about how accurate the data really is?
Any fixed-odds on me getting missed again this time around?