Housing Starts Fall Through the Floor

Go to Original
By Dean Baker

The Census Bureau reported a sharp drop in housing starts in November from a downwardly-revised October rate. The 625,000 annual rate of construction reported for November is 24.8 percent below the September rate. It is 47.0 percent below the November 2007 rate and it is 70.9 percent below the rate in November of 2005 at the peak of the building boom. In fact, the start reported for November is lower than any rate reported for the last fifty years.

The November drop hit all regions of the country, but the Northeast saw the sharpest downturn. Starts fell by 34.6 percent in the region and are now down by 60.2 percent from year ago levels. This seems to be a case where the Northeast is catching up with the rest of the country, since starts had previously fallen somewhat less in the region. It is worth noting that the absolute levels of starts in the Northeast were far lower than in the other three regions, so this drop has much impact on the national economy.

This plunge in starts is a necessary part of the adjustment process in the housing market. There is no way to eliminate the vast inventory of unsold new and existing homes without a sharp slowing in construction. However, the drop in the last two months suggests that the housing market is in a qualitatively different state than it had been even three months ago.

Presumably, builders are cutting back in large part because credit conditions have tightened to the point that they have no choice. This would be consistent with the tightening of credit that took hold in September.

It is important to recognize that this is not necessarily a case of a “credit crunch” inhibiting economic activity. Given the vast oversupply of homes on the market, building new homes in many areas is not a good business proposition right now. Newly built homes can be expected to sit on the market for more than a year in many areas and may eventually sell for prices that are far below recent levels. Banks would be wise not to make loans to builders under such circumstances even if the banks were fully solvent and solidly capitalized.

It will be interesting to see how this plunge in starts is reflected in prices. There is a considerable lag in the data in the key price indices. The indices reflect contracted prices, but only get reported after sales are closed. The sales that are contracted in November will mostly be closed in January, which means that we will not get price data for the month until February.

However, if the plunge in starts reflects the inability of builders to continue to get credit from banks, then it likely indicates that they are also having difficulty obtaining the credit needed to sustain large inventories of unsold homes. This would imply a large sell-off of inventory, presumably at sharply lower prices. If this is the case, we should expect to see a big upturn in new home sales in the November report that will come out next week, with substantial price declines. (The new home sales data show contract prices, so they are more current than other indices.)

If builders are cutting prices to dump inventory, it does not appear that homeowners have yet followed the same path. The Mortgage Bankers Association's purchase applications index continues to show very low levels, even as mortgage interest rates are approaching 5.0 percent. The low measure on this index is the most glaring refutation of the claim that people are unable to get credit. If creditworthy applicants were being denied loans by banks unable or unwilling to lend, then the ratio of mortgage applications to home sales should be soaring. Since there is no notable increase in this ratio, access to credit is obviously not an issue.

The drop in housing starts indicates that housing will again be a sharp negative in GDP this quarter. With sharp downturns in consumption and a weakening trade picture, 4th quarter GDP is certain to show an extraordinary decline.