by Calculated Risk on 2/03/2006 09:06:00 PM

Today's BLS employment report once again showed strong gains for construction employment. Dr. Kash noted this in his post: Job Growth by Industry

"Interestingly, construction employment still seems to be growing robustly, despite some concerns that the real estate market is cooling off; 46,000 of the new jobs created last month were in construction."

Here is a look back at the last consumer recession (1990-1991):


Click on graph for larger image.

NOTE: New Home Sales and Construction employment were normalized to 100 for each peak (see arrows). The '90/'91 recession is in gray according to the NBER dates.

New Home Sales peaked in July of 1989 and construction employment peaked six months later in January of 1990. The recession started in July of 1990.


The 2nd graph is looking at the current situation. It appears New Home Sales peaked in July 2005, although October 2005 was very close. It might take a few more months and a few revisions to determine the actual peak.

If the peak did occur in July 2005, then January 2006 would be six months later - a similar lag in employment just like the early '90s. Of course, no two slowdowns are exactly alike.

So far the housing market is following a predictable pattern: rising inventories followed by a decrease in transactions and finally falling prices (like the early '90s in California, Massachusetts and other bubble areas). Housing related employment lags the peak in housing sales.

There are other factors that make the current situation more dangerous: a larger area of "frothy" prices, more leverage and use of exotic loans, and far more equity withdrawal (that supported consumer spending). On the other hand, in the early '90s, there was a huge shift away from defense work (the "peace dividend") and that impacted both California and Massachusetts.

I'll be watching construction employment over the next few months, and I expect to see a reduction in housing related employment.