The 2010 Census: Annuit Coeptis has hit a bit of a snag…

It’s 2010, the year of the decennial census. We all get letters in the mail asking us several probative questions, all with the intention of taking down information for statistical analysis. The federal government uses data collected every ten years for various purposes, including congressional seat allotment, education funding, etc. Censuses have been used since before the time of Christ to gauge the population changes in a certain region, and gather demographical information. At first glance, this may seem as a positive thing, something useful to have; however, is it really effective? Is a door-to-door headcount really the best way to gather information about the population of our country?

This year’s census will cost $11 Billion, which includes the marketing, mailing, and finally the door-to-door interviewing of the entire population of the US. This is more than half the annual budget of NASA, and 20% of our annual federal funding for education. It costs, on average, $57 for a census worker to visit a house. The process has been described as tedious, costly, and inaccurate in op-ed pieces in the Washington Post, NYT, Christian Science Monitor, and about a dozen other major metropolitan newspapers. A senior fellow at the Brookings Institute says that it would be a mistake for the census to be carried out without sampling. So why does the census not use statistical sampling?

The constitution, at the time it was written, calls for a decennial enumeration of the general population for the apportionment of representatives.

Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons. (Article 1, Section 2)

At the time it was written, a door-to-door headcount was the only known way of counting the population. In today’s day and age, there is a general consensus that statistical sampling would greatly improve not only the accuracy of the census, but also carry only a fraction of the $11 billion price-tag that the current headcount has. The language of the constitution itself in this particular case is arcane. Adding the whole number of free persons and three fifths of all other persons. 3/5 of all other persons?? Clearly this passage was written in a time and matter that reflected the best thinking of it’s time. I think it’s fair to say that it’s time to update this with a more effective method. Well, how come we haven’t done so yet?

As it turns out, such an update of the constitution is extremely controversial due to the political nature of the change. The census generally overcounts kids that go off to college (whose parents may have counted them, in addition to their own census filing at school) and military servicemen serving away from home. In the same respect, the census has undercounted hard-to-reach minorities and homeless people in traditionally urban and other densely populated areas. These folks, were they to vote, would probably vote Democrat (a generalization). This would pose a long floor-fight to even bring such a measure to vote. Nevertheless, this is definitely a worthy cause to consider, as it would alleviate a great tax burden upon the population, as well as increasing the accuracy of the population statistics.

Ok…didn’t mean to put you folks through this whole long thing…but it’s one of those things that really annoy me. I appreciate that we should be getting statistical data about the population to decide how many seats a particular state gets or how much federal education funding each gets, but there are better ways of getting to that data. If only we could push aside the partisanship and get to a better place.