Friday, June 15, 2012

Census Bureau released 2011 housing units estimates

The Census Bureau released 2011 housing units estimates for the nation, states and counties.

It is important to note that the methodology for this series is different than before. From the methodology description:

"Our method of producing housing unit estimates is different this year than in years past. The Vintage 2011 housing unit estimates do not rely on the usual components of housing change (building permits, non-permitted builds, mobile home shipments, and housing loss), which we used last decade to produce the housing unit estimates. Instead, we created the Vintage 2011 estimates by extrapolating the average monthly change in housing units at the county level, then summing these estimates to create estimates for the states and nation. To produce subcounty housing unit estimates, we distributed the extrapolated county estimates down to each subcounty area within a county based on 2010 Census proportions.
We chose not to use the components of change model for 2011 since we are presently evaluating the 2010 housing estimates relative to the 2010 Census results, and considering improvements to the existing housing unit method for the new decade. We expect to return to the components of change model for the 2012 estimates."

This means that a version of the usual methodology will be used next year. Every series start from the last Census and the numbers in the just released estimates will be replaced with numbers using the 'old' methodology next year.

1 comment:

Anonymous said...

If I'm reading the methodology correctly, the Census is just extrapolating the average monthly change between 7/1/2009 and 4/1/2010 into 2011, right? This seems really simplistic to me, especially given they take the trouble to rank the 'top 100 fastest growing counties' using the methodology. Or am I missing something?

To be fair, this question stems from puzzling over why Queens County lost 2,995 housing units between 4/1/10 and 7/1/11...