An MIT economist is now forecasting a rapid recovery in the housing construction segment (click here to see article). While I agree that construction will indeed return, growth is not imminent. In the long run, a growing population will require extensive construction. But in the short run, the massive shadow inventory of housing and almost zero job growth portends continued low levels of new construction. The U.S. Census Bureau forecasts that the U.S. population will grow from an estimated (estimated since we still do not have a count estimate from the 2010 Census) 307.9 million people to 398.5 million by 2050. That’s an increase of 90.6 million in 40 years (2.265 million additional people per year and a gain of 29.4 percent in the 40 year period.)
In 2009, the U.S. Census Bureau estimated an average household size of 2.57 people. Assuming that number remains constant, then an estimated 881,300 additional households will form per year for the next 40 years for a total 35.2 million new households (and absorption of 35.2 million dwelling units by 2050). To forecast the total demand for new housing you must also consider demolitions. A study published in Environmental Research (Volume 9, Issue 2, October 2005) estimated 1.8 million total residential demolitions for the decade. Again assuming a constant level, then in 2010 to 2050 an estimated 7.2 million residential demolitions would occur, combining for a needed 44.4 million new dwelling units—or slightly more than 1 million dwellings per year. And to put this demolition issue in perspective, 8.705 million of the 128.2 million dwelling units (U.S. Census 2000) were manufactured housing (6.8 percent of the total) which have, unfortunately, a shorter life span. So the upside shows demand for significant increases in construction from the current level of building activity proxied by new building permits as reported by the US Census and easily accessed via the Real Estate Center at Texas A&M University. The last 12 months found a total of less than 600,000 permits issued—40 percent less than the demand being created by population and growth (and this is both owner-occupied and rental housing). What clouds this is the massive shadow inventory of housing not listed for sale or even for rent (much less the huge inventory of homes that are in the process of foreclosure or that would be listed for sale when prices finally recover a bit). The table below shows the total inventory and status of US Housing (an estimate from the U.S. Census Bureau) as of the second quarter of 2010. With a total 131 million dwelling units in the country, 19 million are vacant of which 12.6 million are neither for sale nor for rent. And that’s a bunch when you consider we sold 5.2 million existing homes in 2009 and just 400,000 total new homes.
Estimates of the Total Housing Inventory for the United States: Second Quarter 2009 and 2010- census.gov Then throw in the estimated 1 million-plus foreclosures in 2010, of which many are not even listed for sale on the market, and you have a massive shadow inventory that has yet to be counted on the market. So the prognosis? In the short run (read that as the next 18 to 36 months) we will see restricted new housing construction—and double down on that bet if job growth continues to be close to zero. In the long run, however, we need to add 1 million dwelling units per year to inventory.