Tuesday, March 16, 2010

The 2010 Census: Are we missing something?

Regular readers of this blog (both of you) know that I've spent a lot of my career doing market research. In the United States, the decennial (every ten years) Census is the single most important source of governmental and consumer research. Every household in the U.S. is required to participate in the Census, and it represents the "ground truth" dataset used to calibrate governmental, commercial and academic market research databases. Every ten years, the Census allows us to reset, recalibrate and build a strong base for forecasts.

I received the Census form yesterday and filled it out as instructed. In previous years, there were two forms: A short form, roughly equivalent to the 10-question form I filled out, and a long form which asked far more questions and got far more detailed responses. Most people received the short form, while the long form was sent to a statistically-selected sample of the population. In 2000, I received and filled out the long form.

For 2010, the long form is gone, and everyone fills out the short form. The 2010 form asks questions about age, race, ethnicity and housing. There are no questions about employment, income, or the more detailed ethnic, relational and housing questions that we've come to expect from the Census. Instead, a small sample of households, college residence halls, nursing homes, etc. will get the American Community Survey (ACS), which the Census Bureau sends every three years (2007 was the first year). Households who receive both the 2010 Census form and the ACS will have to complete and return both forms.

The information collected in the 2010 Census, while important, will no longer be as useful or insightful for governmental and commercial purposes as previous surveys. A lot of extremely useful information on employment, income, housing and more will be lost, and the ACS will only provide a statistical approximation. The whole reason for doing the Census is to completely measure the population, not a sample of the population. We may be losing too much for the little bit of money that we're saving.
Reblog this post [with Zemanta]

No comments: