Supply of Vacant Houses Remains Swollen
By Marshall J. Vest
EBR and Forecasting Project Director
A key indicator of mobility is the number of vacant housing units. Realtor data on the number of listing (or months supply) offer little insight into mobility. It’s the absorption of vacant inventory that’s important. Unfortunately, most housing studies cover only a submarket (such as apartments or single family new construction). The 2010 Census found that over 16% of Arizona’s housing stock was vacant. After adjusting for second home ownership the rate still approached 10%, a very high rate. (Roughly 6.5% of Arizona’s housing stock – over 184,000 units -- is used seasonally, recreationally or occasionally.) The main question is how have vacancies changed since then?
The U.S. Census Bureau produces a measure of vacancies each quarter that is based on a very small survey that unfortunately is too small to provide meaningful results. About 72,000 nationwide addresses are surveyed each month as part of the Current Population Survey/Housing Vacancy Survey (CPS/HVS). Included are renter and homeowner vacancy rates for states and 75 largest metropolitan areas. According to this source, rental vacancies in Arizona have declined from 13.7% at the time of the 2010 Census enumeration to a still-high 11.2% in the second quarter 2012. Over the same period, homeowner vacancies have moved upward, from 2.6% to 2.8%.
A second measure is produced by the U.S. Department of Housing and Urban Development (HUD) using data from the U.S. Postal Service (USPS). Included is the entire universe of addresses (both residential and commercial) that USPS has in its database. Addresses are recorded as vacant if delivery staff on urban routes observes no mail being collected for 90 days or longer. Data are aggregated and made available to researchers at the Census Tract level (using year 2000 definitions).
At the time of the recent decennial Census in the spring of 2010, the HUD/USPS data recorded a statewide vacancy rate of 4.9% compared to the Census reading of 16.3% (463,536 units). Corresponding readings for the metro Phoenix area were 4.93% vs. 14.53% (227,000). For metro Tucson the comparisons were 5.92% and 11.85% (52,000). Although the total count of addresses compares quite closely, the HUD/USPS vacancy rates are much lower than corresponding readings from the 2010 Census. Nevertheless, changes in the HUD/USPS measure are indicative of household formation and absorption of vacant units. Current readings (for June 2012) show small declines over the past two years for metro Phoenix – from 4.93% to 4.7% -- and an increase from 5.92% to 6.6% in metro Tucson (Exhibit 1).
|Exhibit 1: Residential Vacancy Rates, 2012Q2, HUD and USPS|
All the evidence suggests that little progress has been made so far in absorbing the large numbers of vacant units and that migration flows remain depressed.