by Calculated Risk on 2/03/2009 02:59:00 PM

This morning the Census Bureau released the Housing Vacancies and Homeownership for Q4 2008. This is a very useful report, and earlier I posted graphs of the decline in the homeownership rate, and changes in the homeowner and rental vacancy rates.

Nerd alert: I've mentioned this before, but as a reminder readers should use caution when using the Estimates of the Housing Inventory. The homeownership and vacancy rates come from a survey of a sample of households, but the inventory data is based on two year old housing unit controls. See the discussion at the bottom of Table 4. Estimates of the Total Housing Inventory

The totals shown above have a two-year time lag (4Q2007 uses 2005 housing unit controls from Population Division, which are projected forward and 4Q2008 uses 2006 housing unit controls from Population Division which are projected forward).

We can clearly see the inventory increases are too high for 2007 and 2008. First, the inventory each year increases by the number of housing units completed, minus scrappage and net manufactured homes added (a few scrapped housing units may be rehab'd, but that is minor).

The Census data shows inventory increased by 1.998 million in 2007, and 2.191 million in 2008. These numbers are based on two year old housing unit controls and are clearly way too high. Total completions in 2007 were 1.502 million (plus 95 thousand manufactured homes) and completions were 1.116 million in 2008. Add in some scrappage, and the housing inventory probably increased by less than 1 million in 2008 (less than half the amount the Census Bureau reported this morning).

This is just a reminder that users should use caution when using the inventory numbers.