Scott Sumner recently argued
that the fact that housing starts fell significantly between early 2006 and the spring of 2008 without any significant increase in unemployment, something which according to Sumner proved that it wasn't the housing bust that caused the increase in unemployment, but a drop in nominal GDP (NGDP).
Bob Murphy replied here
. His argument was in short that Sumner was focusing on the wrong data, overlooking the fact that construction employment developed similar (but greater in terms of percentage) to overall employment, something he argues might be due to work on previously built homes. He also mentions how housing prices supports this story, but fails to fully spell out all the implications of this, something which I will elaborate upon below.
Scott Sumner replied to Murphy here
arguing that the smaller decline in construction employment until 2008 was the result of a continued boom in commercial real estate or non-residential construction.
Murphy responded to Sumner's response here
by arguing that Sumner had confirmed his arguments of construction workers being needed to maintain old homes. Actually, he hadn't mentioned that point, and had instead discussed non-residential construction which follows a somewhat different cycle than residential construction.
What's to make of this? Well, data on the real value of different type of construction spending
confirms partially both Murphy's and Sumner's points. Between its peak in Q4 2005 and Q2 2008, real residential construction fell 41%. That is less than the 50% decline in new homes that Sumner tried to point to, suggesting that work on existing houses helped reduce the decline in residential construction activity.
But the main reason why construction employment fell so little was a boom in non-residential construction (called "structures" in GDP statistics) , which expanded by 33% during that period.
Does this mean that Sumner is right and that it was the decline in NGDP rather than the end of the housing bubble that caused the recession?
No, because first of all Sumner wrongly assumes that the end of the housing bubble had nothing to do with the NGDP decline and that the NGDP decline were completely predictable and preventable. I will however not elaborate upon this point for now, but I will probably soon return to this issue.
Secondly, to understand this situation we must first differentiate between the late 2005 to late 2007 period and the late 2007 to mid 2008 period.
The first period featured a continued, albeit slow, overall expansion, the second period featured the beginning of the 2007-09 recession. A recession is usually defined in terms of real output and not employment.
Why didn't the contraction in residential construction cause an overall contraction in the economy if housing was the key factor? Well, basically for the same reason that the big 2001 contraction in non-residential investments didn't cause more than a very shallow recession: namely that there were a bubble in another sector. In 2001 that was residential construction, in 2006-07 it was non-residential construction.
Furthermore, house prices peaked much later (in 2007) than housing construction, something which helped fuel unsustainable level of borrowing for consumption as people used their houses as ATMs.
Then there is the issue of why unemployment rose so little during the first 6 months of the recession, and then rise dramatically during the following 6 months.
Well, this is clearly related to the fact that real hourly wages was reduced
by 1.5% during the first 6 months (after having previously dropped by 1% during the 6 months before that), only to increase as much as 5% during the following 6 months.
Real weekly wages rose somewhat less, 4%, because of a decline in the average work week.
The big drop in employment was thus mainly the result not of a contraction in real output, but from a big real wage shock. The big decline in NGDP thus didn't primarily lower real output, it instead because of a sudden change from high inflation to deflation and nominal wage rigidity redistributed from people who lost their jobs to people who held on to their jobs.