Despite the economic recession of 2001, this decade is off to a great start for the housing sector. Indeed, the strength of production in recent years has provoked charges of "unsustainability" and forecasts of lower production in the years ahead. But the NAHB's current long-term forecast shows housing production numbers that compare quite favorably with recent performance. In fact, the housing production pie promises to expand significantly over the next 10 years.
Long-term forecasts of housing production are based upon three major factors that determine the fundamental demand, or need, for new housing units. The most important is net household formations, and the key drivers of household formations are changes in the size and age structure of the population along with factors, such as divorce rates, that affect the number of households that are formed out of age groups of the population.
Forecasts of population and household growth are anything but easy to make, and net immigration has turned out to be the hardest component to predict. Using the Census Bureau's "middle" series, we come up with a projection of 1.28 million household formations (annual average) for the coming decade, a number that exceeds both the 1980s and 1990s; furthermore, use of the Census "high" immigration series generates an estimate of nearly 1.6 million household formations per year! While it's hard to put probabilities on these two scenarios, it's likely that reality will turn out to be about in the middle, based on recent immigration numbers, with average household formations around 1.4 million for the next 10 years.
Fundamental housing demand also stems from the need to replace units that are removed from the housing stock through demolition or conversion to non-housing uses. We estimate that net removals will average 350,000 units per year during the next decade, up by nearly 100,000 units per year from the 1990s.
The third major factor in the long-term housing equation is changes in the number of vacant units, a number that includes second homes as well as vacant units for ownership or rental. Although the rental vacancy rate now is quite high and may decline in coming years, we expect the total number of vacant units to show some growth during the next 10 years, primarily driven by the second-home component.
The housing numbers
Our forecasts for household formations, net removals, and housing vacancies place our 10-year forecasts for new housing units (annual averages) in a range from 1.82 million (middle immigration series) to 2.10 million (high immigration), and an expectation around 1.9 million seems quite reasonable. If achieved, this performance will exceed both the 1980s and 1990s and approach the robust performance of the 1970s.
The types of housing units to be produced depend on a variety of factors that range from the projected age structure of households to the structure of housing policies that support homeownership or affordable rental housing. Everything considered, we expect that conventionally built single-family homes will account for more than 70 percent of the total housing units produced, with multifamily production a little less than 20 percent, and manufactured homes at 10 percent (the proportions are similar for both the middle- and high-immigration scenarios). These production patterns definitely will generate further increases in the nation's homeownership rate from the current record (68.2 percent seasonally adjusted).