Housing false starts

Even if you have a pessimistic bent, the latest housing starts figures are dispiriting. They fell 10.2% in September, to 1.19 million units, well below the consensus forecast of 1.28 million units. This translates to a full 31% drop from year-ago levels.

Are we near the bottom? Not quite. If the past is any guide, they could go as low as 800,000 units. The question now is how fast we'll get there.

A couple of additional points. First, it's hard to exaggerate just how far forecasters have been behind the curve. According to the Philly Fed's third quarter survey of professional forecasters, in the fourth quarter of this year housing starts would average 1.4 million units (annual terms). Needless to say, reality has rendered this outlook worthless and we should keep this in mind when assessing 2008 growth forecasts.

Of course, this also feeds into my current pet obsession: how can this collapse in housing construction be seen everywhere except in residential construction employment?