Tuesday, October 12, 2010

A closer look at Employment: Construction

- by New Deal democrat

One of the points I've made several times recently is that the expiration of several stimulus programs - the $8000 home buyer's credit, and aid to the states - served to "reset the clock" for jobs in those areas, as layoffs that would have happened one year ago, are instead happening now. Last Friday's payrolls report confirmed that the great majority of renewed job losses are happening in these two areas.

Today let's look at jobs in residential and nonresidential construction. Both of these sectors are very seasonal. Here's residential construction, nonseasonally adjusted:


and here is nonresidential construction, non seasonally adjusted:


The BLS does seasonally adjust both series, however. Let's start with residential construction. Here is the seasonally adjusted graph:


Notice that the number of job losses seems to have slowed dramatically. To better visualize this, it is best to look at the trends in year over year growth or losses.

In the case of residential construction, the year over year maximum decline in jobs took place in April 2009 (coinciding with the all time lows in both housing permits and starts), totaling -196,200. Presently, however, year over year losses have abated greatly and only total -37,200 as of this month. Here's the graph:


This statistic continues to trend towards the better, even though it hasn't turned positive yet. And it is encouraging that the post-housing credit expiration decline in new house sales hasn't led to that much of renewed decline in jobs in this area.

The situation is even less pessimistic in the nonresidential construction sector. The absolute bottom in these jobs took place earlier this year, and jobs have stabilized since:


Here the year over year maximum job loss took place in November 2009, totaling -113,400. As of last Friday's report for September 2010, year over year losses have abated to a mere -17,200. Here's the YoY graph for nonresidential employment:


A few weeks ago I noted that it might be time to start watching for the bottom in nonresidential construction. The fact that we are trending close to zero net job loss year over year is another encouraging sign.

In summary, while construction job losses may not have hit bottom in their absolute scale, the trend is encouraging, and suggests a turning point where year over year there is actual job growth in both residential and nonresidential some time next spring, i.e., in about the March to May window. After 5 years of relentless losses, that will be a welcome relief.