A question was posted to our twitter feed.

Labor growth does not match housing starts growth. Is the housing starts data missing any portions of the workforce?

Ed Zarenski, @EdZarenski


It’s always possible that the employment surveys do not fully capture what is going on in the market. Sampling error for a sector is larger for a specific industry compared to the economy as a whole. The payroll survey (where the 30,000 number comes from) depends on the accuracy of the firms reporting their payroll numbers and the model the Bureau of Labor Statistics (BLS) uses to do an estimate for the whole industry based on the survey responses. The model can drift from reality. This is why the BLS does a benchmark revision of its payroll employment data each year based on more complete data from several sources. The benchmark revision is meant to keep its models on track and adjust for any drift from reality.

Four other factors may be at work here, which are as follows:

  1. To the extent they could, construction firms “stockpiled” labor, i.e., kept their best workers on payroll even when there was limited work for them. As single-family construction ramped up, they could put these workers to better use, limiting the need for additional hires. Larger, national and regional builders would have been in the best position to keep under-utilized workers on their payroll.
  3. Smaller, family-run construction companies typically employ various family members. They may keep these family members on the payroll with a reduced salary. These individuals may take work outside of the industry to make ends meet even as they continue on the family firm’s construction payroll.
  5. Some firms engage in both residential and nonresidential construction. With residential construction ramping up and nonresidential building construction slowing of late, companies may have moved workers from their nonresidential projects to their residential projects.
  7. Many homebuilders kept their companies alive and their workers employed by moving into residential remodeling. With increasing housing construction activity, they may be moving their workers back to homebuilding. Both activities are considered construction employment.
Bernard Markstein
US Chief Economist, Reed Construction Data

Ask Your Own Question

Contact Information

First name:
Last name:

Your Question