The Geographic Distribution of Households with Nonrelatives

A component of pent-up housing demand is the situation of collapsed households — individuals who reside with another household. From a data perspective, we can identify some of these households by estimating the number of housing units that contain individuals who are not related to each other. This is just a part of pent-up demand of course, as it excludes adult children who live with parents or other relatives.

NAHB has previously estimated that pent-up housing demand totals about 2.1 million potential households. The Census Bureau calculated a similar number, reaching 1.9 million households who are “doubled up.” And a recent Federal Reserve Bank of Cleveland study puts the number at 2.6 million potential households.

Data from the 2010 American Community Survey allow us to map the geographic distribution of the households with nonrelatives. The shares of such households are highest in states along the coasts and lowest in the South.  

The state with the highest share is Nevada, at 16.4%, followed by Hawaii( 15.5%) and California (15.3%).

The state with the lowest share is Alabama at 8.8%, with both Mississippi (9.8%) and Arkansas (9.8%) at less than 10%.

While these data are likely influenced by recent events in the housing markets, they also reflect cultural differences among the states. Households with nonrelatives includes homes occupied by the decidedly unromantic classification of “cohabitating partners.” This may be a more common practice along the West Coast and the Northeast than in parts of the South.

Excluding such couples, yields “shared households,” but unfortunately the ACS data do not allow an easy mapping of that population group, which has been growing as a direct result of the Great Recession.

The Geography of the Age of the Housing Stock

In January, Eye on Housing took a look at the age of the housing stock. In that analysis, we found that according to the 2009 American Housing Survey (AHS), the median age of owner-occupied homes in the United States is 34 years old, 11 years older than the median age found in the 1985 AHS. So it is clear that the U.S. housing stock is aging. And older homes are typically more expensive to maintain.

Using data from the 2010 American Community Survey, the geographic distribution of the median age of the entire housing stock (owned and rented) can be presented.  And clear regional clustering can be seen.

For the typical housing unit, the oldest homes are found in the Northeast. With the exception of the District of Columbia (median age of 63, but not a good geographic comparison with states consisting of both urban and rural areas), the state with the highest median age is New York, at 57 years.  Rhode Island is next at 56.

The youngest housing is present in the southern parts of the nation, where population growth has been the highest in recent decades.

An almost identical map portrays the share of the housing stock built before 1970. Again, New York tops the lists of states, with almost 70% of its housing stock having been built before 1970. Rhode Island (65%) and Massachusetts (63%) round out the top three. Again, D.C. ranks higher than any state, with 78% of its housing stock having been built before 1970.

The contrast in shares is strong between states. Among states with younger housing stocks, Nevada (11%) and Arizona (17%) have the smallest shares of their housing stocks having been built before 1970.

This information is clearly important for housing demand, as areas with aging housing stocks have higher demand for both remodeling and replacement housing construction. But on the other hand, the geographic distribution of the age of the housing stock reflects the movement of population within the United States in recent decades.

The following map shows net migration (state-to-state, as well as international) from 2000 to 2010, with the states in red losing population. It’s clear that there is a significant correlation between net migration patterns and the age of the housing stock.  And areas with higher population growth have greater demand for home building.

Immigrants Can Have Substantial Impact on Housing Demand

A new research paper from NAHB Economics investigates how immigrants affect US housing demand.  The study analyzes recent data from the American Community Survey (ACS) that has detailed information on the country of origin, age, family status and housing choices of newly arrived immigrants. The data show that new immigrants are a young and diverse group of people. More than two thirds of them are under age 35. Close to 42 percent of newly arrived immigrants come from Asia and another 40 percent come from Americas. European immigrants account for additional 10 percent of newly arrived immigrants, and the remaining 8 percent are accounted for by other regions.

The study finds that compared to the native born population, immigrants are more likely to live with parents, other relatives or friends rather than establish their own households. These tendencies are reflected in immigrant headship rates that are lower across all age groups. However, the longer immigrants stay in the United States the more likely they are to establish their own households. In case of European-born and other immigrants, their headship rates eventually exceed those of the native born population.

Similarly, the study finds that compared to the native population, immigrants are more likely to rent than own and move into multifamily units.  However, as duration of their stay in the US increases, income rises and socio-economic status improves they are more likely to buy homes and move into single family houses. Europe- and Asia-born households register the highest homeownership rates among all immigrants, reflecting their elevated socio-economic status in the US.

To predict future housing needs of immigrants, the study further builds a model that takes into account age of newly arriving immigrants, region of their origin and length of stay in the United States. For purposes of illustration, the model is applied to the Census Bureau’s low-end 2010 projection of 1.2 million net immigrants. If net immigration of 1.2 million persists for 10 years, new immigrants are projected to account for close to 3.4 million US households (see Figure below). They are estimated to occupy more than 2 million multifamily units and more than 1.2 million single family homes. More than 900 thousand of these new immigrant households are projected to become home owners.

55+ Households are Nearly Everywhere

NAHB analysis of data from the Census Bureau’s American Community Survey shows that, despite popular belief, the geographic distribution of households headed by someone age 55 or older is fairly even across most of the country.  In every state, these 55+ households account for over 30 percent of all households.

On a national level, 43.9 million households are headed by someone 55 years old or higher, accounting for nearly 38 percent of all U.S. households.   Among the 50 states and the District of Columbia, the 55+ household share ranges from 31 to 45 percent. West Virginia tops all states, with 45 percent of its households headed by someone 55 or older, followed by Florida at 44 percent, Hawaii and Maine (each at 43 percent) and Pennsylvania and Montana (at 42 percent). At the other end of the scale, Utah and Alaska are the only states where less than one-third of the households are 55+.

The 55+ household share is also over than 30 percent for 97 percent of the 3,143 county and county equivalents in the U.S.  At the high end, 44 counties have a 55+ household share of over 60 percent. Mineral County, Colo., and Sumter County, Fla., are the highest ranked counties in the U.S. with 77 percent of their households headed by someone 55 or older. Sierra County, N.M., follows closely behind at 74 percent, while both Esmeralda County, Nev., and Wheeler County, Ore., come in at 71 percent each.

Some of the extreme cases of “young” and “old” counties don’t contain many households, but there are five counties in Florida that have both a 55+ household share above 60% and more than 38,000 total households: Charlotte, Citrus, Highlands, Sarasota, and Sumter.  For developers who may be looking for “exceptional” 55+ markets, these 5 counties in Florida form a distinct category.

For tables showing the 55+ household share, as well as the number of 55+ owners, renters, and totals in each state and county, see the complete study.

Construction Self Employment Rates are on the Rise

Construction is known for employing a relatively high share of self employed workers. In fact, according to the 2010 American Community Survey (ACS), the construction sector registers the second highest share of self-employed among all industries, more than 26 percent of the employed labor force, i.e. more than one in four construction workers are self employed.  Only agriculture has a higher share of self-employed, close to 34 percent, while a national average for all industries stands at 10 percent.

It has always been common for some builders and remodelers to maintain relatively small payrolls and rely on subcontractors for a large share of the construction work.  Interestingly, self-employment rates in the construction industry started to rise during the housing downturn and increased from 24 percent in 2006 to 26 percent in 2010. At the same time a national self-employment rate fell from 11 to 10 percent, and self employment in agriculture declined from 41 to 34 percent. Moreover, states known to have been hit hardest by the housing downturn – Florida, California, Nevada, and Arizona – registered some of the highest jumps in the construction self-employment rates. According to the ACS, the share of self-employed construction workers rose in Arizona from 16 to 21 percent and in Florida from less than 24 to 29 percent. Similarly, the share of self-employed construction workers increased by more than 4 percent in Nevada and almost 4 percent in California. It is likely that during the downturn builders and remodelers who were no longer able to maintain a steady work flow may have tried to manage costs by eliminating payroll positions and joining the ranks of the self-employed.  It is also possible that some construction employees laid off during the downturn were able to stay in the industry by striking out on their own.

The 2010 ACS data also show that five New England states have the highest shares of self employed construction workers.  Maine, Vermont, New Hampshire register shares in excess of 40 percent – 43.1 percent, 41.1 percent, 40.3 percent, respectively – well above a national average. Connecticut and Rhode Island follow with 38.5 and 36.9 percent. Montana registers the sixth highest construction self employment rate in the nation, 34.9 percent, i.e. more than one in three construction workers in Montana are self-employed.  Interestingly, Maine, Vermont, New Hampshire and Montana also stand out for having relatively high shares of residential construction workers in their state employed labor force.

Residential construction employment and construction self-employment rates for all states can be found in NAHB: Residential Construction Employment across States and Congressional Districts (Table 1).

Montana’s At-Large Congressional District Has More Residential Construction Workers Than Any Other District

A new research paper from NAHB Economics presents the 2010 estimates of residential construction employment by state and Congressional district. Despite significant employment losses that took place in home building during the housing downturn the industry continues to employ a substantial number of workers in most parts of the country. NAHB estimates show that, nationally, close to 3.4 million people (including self-employed) work in residential construction (RC) in 2010, accounting for 2.4 percent of the US employed civilian labor force.  California has more residential construction workers than any other state, almost 475 thousand, accounting for 2.9 percent of the state employed labor force. Montana tops the state list with the highest share of RC workers, 3.7 percent of the employed labor force. The average congressional district has around 7,700 residential construction workers but that number is often significantly higher.

The map below helps visualize the distribution of RC workers across the Congressional districts. Perhaps somewhat surprisingly, many areas that were once booming and consequently hardest hit by the housing downturn still show higher than average numbers and shares of RC workers.

Montana’s At-Large Congressional district (Rep. Rehberg, Dennis – R) registers the record number of residential construction workers among all districts – 17,190. The 44^th District of California (Rep. Calvert, Ken – R), that includes the city of Riverside, and Texas’s 29^th District (Rep. Green, Gene – D) that serves the eastern part of the Greater Houston area, come second and third respectively with more than 14,000 workers each.  The top ten list also includes three districts in the state of Florida. The 18^th (Rep. Rooney, Tom – R) and the 25^th districts (Rep. Rivera, David – R) cover South Florida and each has nearly 14,000 residential construction workers. The 8^th district (Rep. Webster, Daniel – R) that includes most of Orlando concludes the top ten list with 13,290 residents working in the home building industry. The remaining districts on the top ten list are California’s 49^th (Rep. Issa, Darrell – R), Idaho’s 1^st (Rep. Labrador, Raul R. – R), Arizona’s 6^th (Rep. Flake, Jeff – R) and Colorado’s 7^th (Rep. Perlmutter, Ed – D), each registering between 13,400 and 14,000 RC workers.

By design, Congressional districts are drawn to represent roughly the same number of people (even though in 2010, Congressional districts were still based on the 2000 Census population counts). Consequently, large numbers of RC workers generally translate into high shares of RC workers in their district employed labor forces.  The 29 District of Texas has the highest share of RC workers in its employed labor force, 5.3 percent. California’s 49^th District is a distant second with 4.4 percent.

Affordability Pyramid Shows Most Americans Only Qualify for Lower-Priced Homes

The Census Bureau’s American Community Survey (ACS) provides detailed data on the income distribution of US households. The NAHB Priced Out Model translates the income data into the distribution of homes that US households can afford and allows generating the Housing Affordability Pyramid. The pyramid shows how many households in the United States can afford homes in various price ranges. At the base of the market for housing is a large number of households with relatively modest incomes. The homes that these households can afford are also relatively modest. As the price of a home goes up, there are fewer and fewer households in each tier who are able to afford it. 

Based on conventional assumptions and underwriting standards, it takes an income of about $26,430 to purchase a $100,000 home. In 2012, about 28.9 million households in the U.S. are estimated to have incomes lower than that threshold and, therefore, can only afford to buy homes priced under $100,000. These 28.9 million households form the bottom step of the pyramid. Of the remaining 87.5 million who can afford a home priced at $100,000, 23.3 million can only afford to pay a top price of somewhere between $100,000 and $175,000 (the second step on the pyramid).

This trend continues up the pyramid of house prices. Each step represents a maximum affordable price range for fewer and fewer households. The peak of the pyramid shows a very small share of households able to afford homes priced above $1.25 million. It’s possible to have more million dollar homes than this in the U.S., because many households would have initially purchased homes at lower prices which subsequently appreciated.

The pyramid is based on an income threshold and a 10 percent downpayment assumption. Households at the high end of the market may be more likely to have equity in a previously owned home or other accumulated wealth for a larger downpayment. However, it is less likely to be the case at the low end where affordability is a major concern. Increased development costs can easily price these households out of the market for a new home.

Tightest Housing Markets in the U.S.

A simple measure of tightness in a market for owner-occupied housing is the homeowner vacancy rate (number of homes for sale divided by the number either for sale or owner-occupied). Builders are often interested in markets that are tight by this measure, because it indicates prospective buyers will have difficulty finding a suitable home among the available existing units. 

Several federal government surveys provide homeowner vacancy rates, but the one with the greatest geographic detail by far is the Census Bureau’s American Community Survey (ACS).  In a recent study, NAHB tabulated the most recent (2010) ACS this data for all metropolitan areas in the country.

Overall, the tightest markets tend to be relatively small: Corvallis, Oregon (with a homeowner vacancy rate of 0.23%), Lebanon, Pennsylvania (0.49%), Billings, Montana (0.54%), San Angelo, Texas (0.61%), and Eau Claire, Wisconsin (also 0.61%). 

Because it may seem difficult to compare these areas to larger markets, NAHB also looked separately at the 27 metropolitan areas that have at least 500,000 owner-occupied homes.  Homeowner vacancy rates for these 27 areas range from 1.43 percent to 4.65 percent.  The ten tightest large markets are shown below.

The two tightest large markets in 2010—Nassau-Suffolk, NY and Santa Ana-Anaheim-Irvine, CA—were also the two tightest large markets the last time NAHB looked at the ACS data in 2008.

The NAHB study provides a rundown of the top-10 metros according to nine key measures, including: owner-occupied housing units; homeownership rate; home owner vacancy rate; share of single-family detached homes; value of homes owned; home owner incomes; growth in stock of single-family detached homes; and share of homes built recently. It also has a spreadsheet that shows how more than 350 other metro areas stack up in each category.

House Prices: the “Priced Out” Effect

NAHB Economics regularly receives requests to evaluate the effects of pending new regulations on housing affordability in local markets where regulatory actions are expected to raise home prices.  The NAHB Priced Out Model provides straightforward answers on the issue. The model estimates how many households can qualify for a new home mortgage before and after a house price increase. The resulting difference is the number of priced out households.    A new research paper from NAHB Economics discusses the priced out methodology in detail and presents the new 2012 estimates for the United States and 325 metropolitan areas.

The 2012 estimates show that nationally a $1,000 increase in the home price leads to pricing out about 232,447 households.  The size of the impacts across metropolitan areas ranges from more than 6,000 households in Chicago-Naperville-Joliet, IL-IN-WI to only 14 households in Napa, CA. These large differences mainly depend on metro population, new home prices and income distribution. The Chicago-Naperville-Joliet, IL-IN-WI metro area registers by far the largest priced out effect in the nation, in part because it is a relatively affordable metro area where 43 percent of households can afford a new home, and in part because it is a populous area with almost 3.5 million households residing there. On the other hand, in Napa, CA, where half of all new homes sell for more than $700,000, only 13 percent of households can afford new homes to begin with. So adding another thousand to the price disqualifies only 14 households from buying a new home.

Looking at the affordable metro areas, where more than fifty percent of households can afford new homes, the priced out effects are large and can often disqualify thousands of new home buyers. In Houston-Sugar Land-Baytown, TX, almost 4,700 households are priced out of the new home market as a result of prices rising by $1,000, in Atlanta-Sandy Springs-Marietta, GA – 3, 771 households. 

Even though the NAHB Priced Out Model does not estimate effects of new regulation on new home sales or housing starts, it highlights often overlooked effects of regulation on affordability of new homes.

The research paper also notes that every time a local or regional government raises construction costs by, for example, increasing building permit or impact fees, the final price of the home to the buyers usually goes up by more than the increase in the government fee. This is because other costs such as commissions and financing charges automatically rise as well. As shown in Table 1, these add-on charges range from 0 percent if a fee is imposed directly on buyers to 39 percent if cost is incurred when applying for site development approval. So that for every $1 increase in fees incurred, for example, when acquiring a building permit, the final price of a new home to its final customer rises by $1.20.

Top 2011 Posts: Where are the Nation’s Second Homes?

With the end of 2011 approaching, the contributors of NAHB’s Eye on Housing thought it would be useful to take a look at the updates that attracted the most readers over the last year.

In August, we mapped the locations of the nation’s second homes. We highlighted the fact that the tax definition of a “second home” for the purposes of the mortgage interest deduction includes a significant number of homes that many people would not think of as a second home. This includes: (1) a home that used to be a primary residence due to a move or a period of simultaneous ownership of two homes due to a move; (2) a home under construction for which the eventual homeowner acts as the builder and obtains a construction loan (Treasury regulations permit up to 24 months of interest deductibility for such construction loans); or (3) a non-rental seasonal or vacation residence.

On the other hand, many homes that people think of a second homes, such as expensive beach homes, are in fact rented and rental units do not qualify under the rules for the mortgage interest deduction.