In June 2010, NAHB published an article on HousingEconomics.com which allows readers to compare local housing markets across the country. That article was based on American Community Survey, or ACS, data collected in 2008.
This article uses data from the 2010 ACS, released at the end of 2011, to gather information on a larger number of characteristics and organize them in a way that allows comparison across the entire country. Like the 2008 article, this one concentrates on owner-occupied homes. A companion piece on rental markets appeared in NAHB’s e-newsletter Multifamily Market Outlook.
The ACS is an ongoing statistical survey conducted by the U.S. Census Bureau. Previously the information collected by the ACS had only been collected in the long form of the decennial census. Although the ACS is not as comprehensive as the decennial census, the results are available every year rather than every ten years.
For each metropolitan area, this paper highlights the following statistics:
Number of owner-occupied housing units
Homeowner vacancy rate
Share of homes that are single-family detached
Value of homes owned
Income of homeowners
Change in number of single-family detached homes
Share of homes built recently
The definition of a metropolitan area used here is generally a Metropolitan Statistical Area (MSA), based on the list published and updated regularly by the U.S. Office of Management and Budget (OMB). Working with the Census Bureau, OMB defines MSAs by aggregating contiguous counties based on local commuting patterns. This commuting criterion is intended to capture local labor markets. In most places, the correspondence between housing and labor markets should be relatively close.
Several of the larger MSAs are broken into "Metropolitan Divisions," and where this occurs, this article will use the divisions rather than entire MSAs. This will allow us to show separate statistics for both San Francisco and Oakland, for example, or Dallas and Fort Worth. Wherever the term "metropolitan area" is used, it refers to this combination of MSAs and Metropolitan Divisions.
The following sections show the top metropolitan areas according to each of ten key housing market statistics. An alphabetic list of all metropolitan areas showing each measure, and where each metropolitan area ranks nationally according to the measure, is available in an attached Excel file that can be downloaded or opened online.
Number of Homes
Figure 1 show the ten largest metropolitan areas based on the number of owner-occupied housing units. If we expand the list to include the 11th largest metropolitan area, Riverside-San Bernardino-Ontario, CA, the list contains the same areas as the list for the top 11 largest metropolitan areas, based on population, although in a different order internally. For instance, although both New York-White Plains and Los Angeles are the two largest areas based on population, Chicago tops both of them when it comes to the number of owner-occupied units.
The top ten largest markets in 2010 are the same as in 2008, although again, in a slightly different internal order. Interestingly, only two of the top ten largest areas, based on number of owner-occupied housing units, increased in size. Both of these metropolitan areas are in Texas and both saw slight decreases in population between 2008 and 2010. Houston had 1,243,989 owner-occupied units in 2008, and 1,270,817 units in 2010. Dallas had 901,263 owner-occupied units in 2008, and 907,760 in 2010. The other eight areas all saw a decline in owner-occupied housing units between 2008 and 2010.
Since both Dallas and Houston saw a slight increase in number of owner-occupied housing units, it follows that they also both saw a slight increase in homeownership rates. Homeownership rates are calculated by dividing the number of owner-occupied units by the number of occupied units in a market area.
Figure 2 shows the top ten areas with the highest homeownership rates. In all ten of these areas, over 75 percent of occupied units are owned. Most of these areas are small metropolitan areas, with eight of them having a population of less than 100,000. Also, most of these areas tend to have lower home values, with eight of them having a median home value of less than $175,000. Nassau-Suffolk, NY stands out as the biggest outlier with a population of over 900,000 and a median home value over $400,000.
It is also interesting to look at the metro areas where homeownership rates are low. In some of these places, prospects for expanded homeownership may be limited by high housing costs or the transient nature of the population. Three of the areas, New York, LA, and San Francisco, are large metropolitan areas where the home values are among the highest in the country, pricing out potential homebuyers. Four of the areas, Manhattan, KS, Lawrence, KS, College Station, TX, and Ames, Iowa, are home to large universities (over 20,000 students), where the population is very transient. Jacksonville, NC has a large military presence, which also tends to be very transient.
The share of vacant homes in an area gives us insight into how tight a market is for owner-occupied housing. It would be harder for prospective buyers to find an available home from the existing stock in those areas with low vacancy rates, compared to areas with high vacancy rates. Figure 4 shows the ten metropolitan areas with the lowest homeowner vacancy rates. All of these areas have a less than 1 percent vacancy rate. Although all ten metropolitan areas with the lowest vacancy rates in 2008 continue to have vacancy rates under 1 percent, only Lebanon PA appears on the top ten lists in both years.
Keep in mind that homeowner vacancy rates may not be comparable across metropolitan areas with different sizes. Eight out of the 10 areas with the lowest vacancy rates have less than 100,000 owner-occupied housing units. For comparison, Figure 5 shows the ten metropolitan areas with the lowest vacancy rate, taking into account only those areas with at least 500,000 homeowners.
Twenty-seven metropolitan areas have at least 500,000 owner-occupied housing units, and were considered for Figure 5. Homeowner vacancy rates for these 27 areas range from 1.43 percent to 4.65 percent. The two areas with the lowest vacancy rates in 2010, Nassau-Suffolk, NY and Santa Ana-Anaheim-Irvine, CA, were also the two areas with the lowest vacancy rates in 2008 (28 metropolitan areas qualified in 2008).
Single Family Concentration
Homeowners choose from a variety of structures when buying a home. Condominiums tend to be more concentrated in parts of the county with higher housing (and land) costs, while single-family structures are more common in parts of the country with lower housing costs. Figure 6 lists the metropolitan areas where homeownership is more concentrated in single-family detached housing.
In all of these top ten areas, single-family homes account for more than 90 percent of the owner-occupied housing. Most of these areas are in low-cost areas in the central part of the country. In fact, in all but one of these areas, the median home value is less than $200,000. In Madera-Chowchill, CA, a state with traditionally higher home values, the median home value is $203,900.
Figure 7 shows the metropolitan areas where the median home values are the highest. Seven of the ten areas are in California, a state with consistently high housing costs, two areas are near New York City, and one is in Hawaii.
The share of single-family owner-occupied homes is less than 85 percent in each of the ten costliest metropolitan areas. Also, higher value homes tend to be found in larger metropolitan areas. All but two of the ten costliest metropolitan areas contain at least 150,000 owner-occupied homes.
In areas with high home values, fewer people own their homes. In eight of the top ten areas, less than 60 percent of the occupied units are owned. Oxnard has a 64 percent homeownership rate and Bridgeport has a 69 percent homeownership rate.
Interestingly, among metropolitan areas with the highest median home values, we see fewer newly built homes. Less than 5 percent of any of the owner-occupied homes in these ten areas were built after 2005. As you will see in later sections, this share is between 15 and 20 percent for areas with the highest share of homes built after 2005. Homeowners in these high cost areas on the whole are paying for the location, not a newer home.
The two metropolitan areas with the highest median home values—San Francisco-San Mateo-Redwood City, CA and San Jose-Sunnyvale-Santa Clara, CA, also held this rank in 2008. In light of the housing downturn it is not surprising that all ten areas saw a decrease in median home value between 2008 and 2010.
The ability to buy a home largely depends on two things—household wealth, often in the form of an existing home, and household income. Figure 8 shows the metropolitan areas where homeowners have the highest median incomes.
The high income areas also tend to have relatively large populations. With the exception of Anchorage, AK, all of the metros in Figure 8 have more than 250,000 owner-occupied housing units. Not surprisingly, there is a correlation between home value and owner’s income across metros. With the exception of Anchorage, AK, the areas in Figure 8 all have median home values greater than $350,000. The two areas with the highest median home value, San Francisco-San Mateo-Redwood City, CA and San Jose-Sunnyvale-Santa Clara, CA, rank second and fourth respectively, when it comes to median income.
The comparison of median income and median home value in a single metropolitan area may give us some insight into the cost of living in that area. If a homeowner could make the same income in two metropolitan areas, he/she could afford more home in the area with the lower median home value. For instance, although the median income in San-Jose-Sunnyvale-Santa Clara, CA ranks second among all metro areas at $108,643, the median home value is $631,400, also the second highest. The median income in Bethesda-Rockville-Frederick, MD, is $110,613, only slightly higher than the median income in San-Jose, but the median home value is $409,600, about a third less than the median home value in San-Jose. So although homeowners in San-Jose, and Bethesda have similar median incomes (and rank first and second, respectively, among all metropolitan areas), homeowners in Bethesda (DC area) are paying less for their home than homeowners in San-Jose.
Eight of the ten areas in Figure 8 were also on the top ten high-income list in 2008. Only four of the areas in Figure 8 actually show an increase in income since 2008—Bethesda-Rockville-Fredrick, Washington- Arlington-Alexandra, both in the Washington DC area, Cambridge-Newton-Framingham (Boston area), and Anchorage. The remaining six, mainly in the areas surrounding San Francisco and New York City, saw a decline in median income.
Fastest Growing Markets: Increase in Owner-Occupied Homes since 2008
It is also interesting to look at how the stock of owner-occupied housing in metro areas has changed since 2008. Just over half of the metros saw a decrease in owner-occupied housing units between 2008 and 2010, 11 percent saw no change, and 38 percent saw an increase.
Figure 9 shows the top ten metro areas with the largest increase in owner-occupied homes since 2008. With the exception of New Orleans, all of these areas are small markets with less than 80,000 owner-occupied units. So, although, the number of owner-occupied housing units in Hinesville-Fort Stewart, GA increased by a third between 2008 and 2010, the number of owner-occupied units increased by less than 4,000 units. The one exception is New Orleans, an area that is still in a re-building process after hurricane Katrina. The median home value in the ten fastest growing metros tends to be low, in all cases below $180,000.
Fastest Growing Markets: New Construction
The number of owner-occupied homes may increase because vacant units are absorbed, or because rental units are converted, or because new homes are built. Figure 10 shows the metropolitan areas where growth in the owner-occupied housing stock is due to new construction (based on the share built in 2005 or later, which is the most recent cut-off date available in the ACS data). ACS data are collected continually throughout the year, so that the 2008 ACS should capture roughly one half of the new units built in 2010.
Eight of the metro areas in Figure 10, like St. George, UT, are small markets, each adding less than 20,000 new owner-occupied housing units to the stock between 2005 and 2010 (half added less than 10,000). Austin and Las Vegas stand out as large metropolitan areas, with rapidly growing owner-occupied housing stock. Both belong in the top 50 largest metro areas, each adding about 60,000 new owner-occupied housing units to the stock between 2005 and 2010.
By combining Figure 10 percentages with the number of units shown in Figure 1 and the attached Excel file, it is possible to calculate the number of owner-occupied units that were built recently. For example, Figure 10 shows that 20.1 percent of the owner-occupied units in the St. George, UT metropolitan area were built between 2005 and 2010. Then, in the Excel file, you can see that there were 32,513 owner-occupied units in St. George in 2010, so that the 20.1 percent in Figure 6 translates roughly to a total of 6,500 owner-occupied units built between 2005 and 2008.
The cut-off at 2005 in the ACS for year the units were built is somewhat awkward, given the timing of the downturn that hit virtually all housing markets in the U.S. As a result, the units built between 2005 and 2010 tend to be heavily concentrated in the earlier years of that period. For example, total residential permits issued in St. George and Las Vegas each declined by 85 percent between 2005 and 2009 (although permit activity recovered somewhat in St. George in 2010, it declined even further in Las Vegas).
Nationally, although Las Vegas ranks seventh in terms of the share of units built in 2005 or later, but only 308th in terms of the change in owner-occupied units with a 3.5 percent decline. One of the results is a relatively high homeowner vacancy rate of 5.0 percent in Las Vegas.
To see where a metropolitan area ranks according to each of nine different measures, readers may open the attached Excel file and view the entire snapshot for a selected metro.
 “New Data Provide Snapshot of Local Rental Markets” June 2010.
 Defined as vacant housing units for sale divided by the sum of 1) owner-occupied units, 2) vacant units for sale, 3) vacant units that have been sold but are awaiting occupancy.
For more information about this item, please contact Heather Taylor at 800-368-5242 x8503 or via email at firstname.lastname@example.org.