by Calculated Risk on 10/29/2014 05:28:00 PM

From housing economist Tom Lawler:

Yesterday the Census Bureau released the “Residential Vacancies and Homeownership” Report (commonly referred to as the Housing Vacancy Survey, or HVS) for the third quarter of 2014. The HVS is based on a relatively small sample of housing units, still uses an outdated “sampling frame,” and does not perform adequate follow-ups to surveyed units, and for well over a decade the HVS has overstated the housing vacancy rate and overstated the US homeownership rates. The HVS is a supplement to the monthly Current Population Survey, which also suffers from sampling “issues.”

Here are a few summary statistics from yesterday’s report.

Housing Vacancy Survey "Results"
	Q3/14	Q2/14	Q3/13
Gross Vacancy Rate	13.5%	13.6%	13.6%
Rental Vacancy Rate	7.4%	7.5%	8.3%
Homeowner Vacancy Rate	1.8%	1.9%	1.9%
Homeownership Rate NSA	64.4%	64.7%	65.3%
Homeownership Rate SA	64.3%	64.7%	65.2%

The report also shows “estimates” of the US housing inventory, which are “controlled” to independent estimates of the housing stock by the Population Division using a questionable methodology. Here are some summary stats on the housing stock from the report.

HVS "Estimates" of US Housing Inventory (000's)
	Q3/14	Q3/13	Change
All Housing Units	133,331	132,843	488
Vacant	18,021	18,075	-54
Year-Round	13,447	13,603	-156
Seasonal	4,575	4,475	100
Occupied	115,310	114,769	541
Owner	74,240	74,897	-657
Renter	41,070	39,872	1,198

If one were to believe these estimates, US households grew by just 541,000 over the last year, and just 1.012 million over the last two years. Given Census’ estimate of housing completions and manufactured housing placements (which strangely are not used by the Population Division in estimating the housing stock), the change in the housing stock shown in the HVS over the last year looks too low, possibly by around 160,000 or so. Even if one assumed the housing stock grew by 160,000 more, however, the HVS-based household estimate would have increased by only 679,000.

With respect to the homeownership rate, below is a table comparing homeownership rates from the Decennial Census for April 1 to the homeownership rates from the HVS for the first-half of the year for 1990, 2000, and 2010.

Different Homeownership Rate Estimates
	1990	2000	2010
Decennial Census (April 1)	64.2%	66.2%	65.1%
HVS (first half average)	63.9%	67.2%	67.0%

The HVS homeownership rate for the third quarter of 2014 was 2.6 percentage points lower than the HVS homeownership rate in the first half of 2010. If the “true” homeownership rate in the US fell by a similar amount over this period, then the “true” homeownership rate would have been about 62.5% last quarter, the lowest homeownership rates since the 1960’s.