Vacant existing homes and potential additions remain a leading hurdle to new-home production. An overabundance of empty houses at desperation prices are tough competitors when the costs of building have not seen the same devaluation. A look at underlying vacancy measures shows promise for the future.
The excess is entirely within the existing stock. The number of newly built for-sale homes is at a four-decade low but 14 million year-round existing homes or nearly 11 percent of the total housing inventory are vacant. On the face, that would appear to be a huge glut, but a deeper dive reveals a more reasonable conclusion. The excess over what the actual level should be is more on the order of 1 to 2 million and has been declining for a year. The large difference in the actual number and the expected number is due to the growing need for a larger inventory of homes to exchange. Between 1969 and 1999 the share of the housing inventory left vacant increased steadily from about 7 percent to 9.5 percent. A continuation of that trend would suggest about a vacancy rate of 9.7 percent or a current excess of 1.3 percent of the year-round housing stock.
Vacancy levels have fallen the least in the places that experienced out-of-bounds house price and production increases or underlying economic change. About one-third of the states and 40 percent of the largest metropolitan areas still have homeowner vacancy rates that are more than one percentage point greater than they were in 2005. Areas with relatively high vacancy rates are concentrated in the states experiencing the greatest housing distortions (California, Nevada, Arizona, and Florida); states dependent on manufacturing (Ohio, Michigan, Illinois); East Coast states (Virginia, Maryland, New Jersey, Connecticut, and Maine); and the Northwest (Oregon and Washington). The remainder have returned to normal or nearing normal.
Locations with a larger share of vacant for-sale inventory were faster-growing areas containing more recently built homes. The homeowner vacancy rate for homes built before 2000 has remained steadily below 2.5 percent while the rate for homes built after 2000 rose to nearly 7 percent in 2008 but has since declined to around 4 percent. More recently built homes have higher loan-to-value ratios, larger shares of the household income devoted to mortgage payments, and are near the competition of newly built homes. The share of mortgages underwater for homes built since 2000 is three times the share for homes built before 1990.
The stream of foreclosures that add to the vacant inventory and depress prices will continue but the flow does appear to be diminishing. The share of mortgages seriously delinquent has started to fall in the U.S. and half the states. The shares 30 days and 60 days late have fallen in nearly every state since 2009. Unemployment has been the single largest cause for foreclosure most recently, and while still a long way from normal, employment has begun to pick up and this also will relieve payment burden pressures and allow more homeowners to continue paying their mortgage.
And, finally, the number and share of homeowners with an outstanding mortgage greater than the value of their home, may be overstated by the automatic valuation models. Estimates of 20 percent to 25 percent of homes underwater are based on modeling values. Homeowners’ own valuations according to the 2009 “American Housing Survey” result in much lower estimates of distress. Since a homeowner’s evaluation is the first step in a homeowner’s decision to default, this approach may be more telling than model results. Using homeowner values, 7.4 percent of homes had mortgage balances more than their value in 2009 and another 6.5 percent had values near the mortgage balance. More than a third have no mortgage and another third have at least 40 percent equity.
Excess vacancies will continue to compete with newly built homes in the more distressed markets, but the worst appears to be behind us and improvements will spread as the economy heals.