The Condo Market: Coming of Age
In this country, “buying a home” is overwhelmingly likely to mean buying a single-family detached house. But with changing demographics, and the shrinking supply of ever-more-expensive close-in land, the demand for condominiums is beginning to grow.

Baby boomers and empty nesters who are retired or anticipating retirement are fueling an increased interest in homes that demand less maintenance. These tend to be people who've owned homes, and many  would still rather buy than rent. So it’s no surprise that the condo absorption rate, weak in recent years, is beginning to rebound.

That information, and much more, comes from the 2000 Census – recently released, and analyzed by NAHB’s Housing Policy economists.

But first, some background

Before jumping into the numbers, let’s be clear about the words. A condominium isn’t a building – it’s a legal form of ownership. Condo owners hold fee-simple title to their housing units, and they share ownership of common areas. Cooperatives are a similar legal arrangement, but owners have shares in a corporation that owns both housing units and common areas; individual owners have exclusive rights to occupy a unit.  

The American Housing Survey (AHS) indicates that there are about eight times as many condos as co-ops. Because they are similar (even though there’s a legal distinction), we’ll refer to condos and the condo market with the understanding that the reference often will include co-ops as well.

Some structures are more suitable for condominium ownership than others. According to the AHS, more than a third of condos and co-ops are single family (including attached single family town homes), and nearly half are in structures with fewer than five apartments (Figure 1).

Source: 2001 American Housing Survey, U.S. Census Bureau and the Department of Housing and Urban Development.

That said, the focus of this article is on owner-occupied units in the larger, five-plus structures – that is, condos in buildings with at least five units – which tend to be of greatest interest to builders specializing in multifamily construction.

At the national level, some information is available on an ongoing and relatively timely basis.  For example, the Survey of Construction reports on the share of multifamily units started that are intended for sale. That share generally has hovered around 20% since 1990 (Figure 2).

There was a spike around 1993, caused by exceptional weakness in the production of new rental units rather than a surge in condo construction.

In addition, HUD funds a quarterly follow-on to the Survey of Construction, called the Survey of Market Absorption (SOMA), to measure the rate at which new housing units in structures of 5 or more units are rented or sold. The SOMA shows that, although the absorption rate (the share rented or sold within three months of completion) for condos weakened over the past couple of years, it never fell to historical lows like absorption rate for rental apartments (Figure 3).  Moreover, the condo absorption rate rebounded sharply in the most recent SOMA data (based on units completed in the third quarter of 2002).

Absorption rates are for structures with five or more housing units that are privately financed, nonsubsidized, and unfurnished. The results are presented for units completed during the indicated quarter. The quarterly rates are not seasonally adjusted. Source: U.S. Census Bureau, Housing Reports Series H-130, Market Absorption of Apartments.

The 2000 Census

The major shortcoming of most of these condo statistics is that they don’t provide much geographic detail. For even a state-by-state breakdown, the only federal statistics available are those associated with owner-occupied five-plus housing units in the decennial Census. The limitation of the Census data (beyond the obvious ten-year gap between observations) is that we have to base judgments on owner-occupied units.

In five-plus structures, these should include only condos and co-ops.  The 2000 Census therefore can tell us where condos tend to be concentrated, and it can provide some information about them. Thanks to suggestions made by NAHB and adopted by the Census Bureau, the 2000 Census Summary Files contain more information about condos than the 1990 files did. In particular, the 2000 files permit us to calculate the average market value of occupied condos.  

Although we can observe the number of owner-occupied five-plus units in both 1990 and 2000 in a given area, it’s difficult to say whether a large increase occurred for a particular reason.  New condo structures could have been built, but vacant units also could have become occupied, or rental units could have been converted to condos (attempts to back into the numbers suggest that, over a 4-year period, about 150,000 condos were created from net conversions)¹.   However, any of these indicate that the condo market is growing, and for some purposes that knowledge may be as useful as understanding precisely how the growth is being accommodated.

Before looking at smaller market areas, let’s consider Census numbers for the 50 states, the District of Columbia, and the U.S. as a whole in Table 1.

The table shows that in 2000, the number of owner-occupied 5-plus units in the U.S. was just under 2 million, representing a roughly 30% increase from 1990.  

Among the states, the number of occupied five-plus condos ranges from a low of less than 1,000 in Wyoming to more than 400,000 in Florida. Nevada showed the greatest growth rate of any state, a 154% increase between 1990 and 2000. Six other states showed an increase of at least 75% over the decade. Four are clustered together in the Rocky Mountain-northern plains area (Colorado, Montana, South Dakota, and Utah). The others are Mississippi and Vermont. In general, these six states developed modestly sized condo markets during the 1990s where almost none existed before.

Compared to renters, or even owners of other types of housing units, five-plus condo owners tend to be somewhat older. Across the U.S., their median age is 56. A conventional explanation is that as householders approach that age, they are attracted by a living arrangement that frees them from the need to do yard work and exterior maintenance, but are not necessarily ready to become renters. Given that the large baby-boom cohort is now moving into what appears to be the prime condo-owning age bracket, it’s perhaps surprising that production of new condos hasn’t been even stronger than shown in Table 2.

In some states that contain relatively few five-plus condos, the ones that do exist seem primarily targeted to older adults. In Nebraska and South Dakota, for instance, the median age of five-plus condo owners is over 70.

Finding the Top Counties

There are several ways to identify the top counties for five-plus condos in the Census data. The simplest is to look at the counties with the most condos in 2000, and Table 2 lists the top ten on that scale. There is little mystery about these counties and why they made the list.  All are either in areas with extremely high housing costs, or in Florida.  In high-cost areas, relatively affordable multifamily condos are one of the few ways many families can become homeowners.  In Florida, the relatively large proportion of the population over age 50 becomes a factor. In addition, Florida is a traditional winter destination, where some households buy condos as a second homes that they may rent in the off-season.

Table 2 shows counties that had a large stock of multifamily condos in 1990, and added to that stock at close to or under the national average rate during the subsequent decade.

Another way to look at the market is in terms of a growth rate—that is, determine which counties showed the largest percentage increase in occupied five-plus condos between 1990 and 2000.  Table 3 is based on that approach.

In order not to be misled by the inclusion of sparsely populated areas where a single garden-style condo building could indicate a very large growth rate, the table shows only counties that had at least 1,000 five-plus condos in 2000. Results clearly are sensitive to the choice of a cut-off. More than half of the counties in Table 3 have between 1,000 and 2,000 five-plus condos. Two of the counties show approximately 40 owner-occupied units in buildings with 50 or more units. This can occur if a single building has, in addition to occupied condos, a number of vacant or rental units.

Table 3 identifies counties with between 150% and 300% growth in five-plus condos during the 1990s. Most of these are counties that belong to metropolitan areas but do not contain the metro area’s central city. As such, they generally are adjacent to counties where a five-plus condo market existed before 1990 and now appears to be expanding into counties that are becoming more closely integrated into the metro area. 

Exceptions to this rule include Clark County (Las Vegas) Nevada and Eagle County, Colorado. Clark, with the most condos of any county in Table 3, is part of an area that is experiencing rapid growth in general. Eagle, with the highest-priced condos of any county in Table 3, is part of a resort area just west of the mountains. Utah County (the Provo-Orem metro) area, at the top of the fast-growth list, also is slightly west of that same mountain range and also could be receiving some overflow activity from Salt Lake City, a substantially larger metro area that lies just to its north. The five-plus condos in Utah County are spread across many of the county’s 85 Census tracts.

Another way to create a top-10 list is to look for upscale markets and counties where condos have the highest average value. Table 4 takes this approach. Again, we want to avoid counties with an extremely small number of condos. We assume, however, that where price rather than the growth rate is the issue, a threshold lower than the one used in Table 3 would be reasonable.  Table 4 therefore excludes only counties that had fewer than 100 occupied five-plus condos.

Unlike the previous two tables, Table 4 includes a wide variety of condo markets with drastically different sizes and growth rates. The number of occupied five-plus condos ranges from 180 in El Dorado, California to well over 100,000 in New York County (Manhattan), NY.  These counties also were at opposite extremes in terms of growth between 1990 and 2000 (17.5% for New York, 300% for El Dorado).

Despite the high average price of the condos in the counties in Table 4, prices are only 2.5 to 3 times as large as the average income of their owners in New York County, Nassau County, Florida and Suffolk County, Massachusetts (Boston). New York and Suffolk are the centers of large Northeastern metro areas that are well known for containing affluent households in significant numbers. Nassau County is to the north of Jacksonville, on the border between Florida and Georgia, and part of the Jacksonville metro area. Neither the county nor the larger metro area has especially high incomes or house values. However, there is a concentrated area in Nassau County, consisting of two Census tracts, where incomes and house values are quite high. All of that county’s five-plus condos lie in that area.

For the seven other counties in the table, the values of the units and incomes of the owners seem out of balance. Most of these are in California, where the high cost of housing relative to household income is well-documented, and the political environment has been, perhaps, less friendly to multifamily construction – and home building in general – than market forces suggest it should have been.

This article has shown only a portion of the information about condos NAHB has compiled from the 2000 Census – a full treatment would require volumes. But a complete set of tables is available with detailed Census information for five-plus condos in every county in the United States. The information includes when they were built, their average value, when occupants moved in, and their characteristics such as age, household size, and average income by age of the household. Keep in mind that not all counties contain five-plus condos. Of the 3,141 counties and county equivalents in the U.S., nearly 700 have none at all, and another 1,700 have fewer than 50.  In addition to condos, the tables show similar information for five-plus rental units in the county. In all cases, state and national numbers also are provided  for comparison.  This information, customized for a particular county or group of counties, can be obtained for a reasonable fee. Contact multifamily@nahb.com for more information.

1Michael Carliner, “Despite Strong Demand, Supply of New Condos is Limited,” NAHB Multifamily Market Outlook, February 2001.