Posted by: Josh Lehner | February 5, 2013

Construction: Employment and Starts

One particular topic that has been gaining attention on the housing recovery is the relative lack of new construction jobs even with housing starts increasing substantially. After the recent revisions to the employment data, the data now show that construction employment is up nationwide nearly 300,000 since bottoming out in early 2011. However, housing starts in December were up nearly 40% from a year earlier while construction employment was only up 1.6%. In Oregon our construction employment has been a bit stronger, however we believe this to be related, in part, to the Intel expansion(s). It is an interesting topic to discuss as there is clearly a long-run relationship between housing starts and construction employment. Tim Duy wrote recently about the the ratio of construction employment to housing starts at the US level and notes that payrolls still appear to be larger than the underlying level of housing starts would suggest. The graph below illustrates the same ratio for Oregon. It should be noted that this is not the actual number of individuals working on each housing start but, rather, the ratio between the two series.


This ratio remains higher than the long-run average (for both Oregon and the US) and is likely due to a number of reasons, including the fact that housing starts plummeted to a larger degree than construction employment did, and also the level of non-residential construction is a larger portion of the overall industry today given increases in items such as public works and also the lack of single family activity relative to historical norms. These factors are related and would keep this ratio elevated. In order to gauge the level of non-residential activity in recent years, I went and grabbed the data on state government finances available from the Census Bureau. The graph below shows Oregon’s expenditures on capital outlays and highways, as examples of public works spending. Note this is the data compiled by the Census and is available for all states, and is not directly from, say, ODOT who may report slightly different numbers. There was a clear increase in 2009 and 2010, at the same time that new home construction was at the lowest levels seen in a generation (or two or three). Even though construction employment overall collapsed about 35% in Oregon, and the ratio to housing starts shot up to 10, it likely would have been larger without these public increases, fueled in part by both the national and state stimulus programs.


Going back to the ratio graphed above, our office does not model the forecast explicitly in this way (and maybe we should) but the outlook does include a return of this relationship back into its long-run historical range in the coming years. Our overall housing outlook remains relatively unchanged and even as housing starts are increasing strongly today, coupled with rising home prices, expectations remain for supply/demand responses in the existing home market. Stable to increasing demand, coupled with declining inventories is helping to increase prices which will bring more and more homeowners back above water on their mortgages, allowing them to sell, if they choose/need to. This pattern is expected to play out in the next couple of years and to me, going back to the near term outlook for construction employment is as Tim Duy wrote:

This suggests to me that housing starts need to go higher until construction firms begin hiring more aggressively.  Until then, expect firms to increase the hours of existing workers.

Finally, it should also be noted that Trulia chief economist Jed Kolko has written recently that if you look the number of units under construction (different than starts) and just the residential construction employment then the relationship is tighter and we are seeing gains in proportion. However this does not address these larger construction-related comparisons which still hold. Another factor that will influence the industry moving forward is worker productivity and the technological changes that may impact the number of workers needed per housing start. Just another item to keep your eye on.


  1. Josh, you should be able to get hours worked for the industry, compare it with QCEW, and compute a quarterly FTE employment for construction. That would tell you whether hours worked was increasing (if FTE employment is rising faster than QCEW). Construction, as you indicated, is a lot more than new housing–there’s home remodeling, commercial, industrial, infrastructure. I would guess that a new home is on the order of 1 job on an annualized basis, but industry people could probably give a better estimate.

  2. […] and at the national level, the following takes a look at employment and hours worked. After the recent post on construction employment and housing starts, our good friends over at the Employment Department […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: