New-home construction is expected to get back to a “normal” level of 1.6 million starts per year by 2016. If the industry reaches that plateau again, it is also expected to create an estimated 412,000 new residential construction jobs between 2012 and 2016. But that increase would still leave the industry’s construction employment nearly 1 million workers short of the 3.4 million it hit at the peak of the last housing boom in 2006.

These are the conclusions of a May 6 analysis of supply and demand in housing and employment by Fannie Mae’s Economic and Research Group. The analysis adds to the growing body of research that raises questions about the housing industry’s ability to sustain its recovery.

The Fannie Mae paper can be seen as a response to an analysis of residential construction jobs released last February by Kris Dawsey and Hui Shan of Goldman Sachs. In their paper titled “Housing Sector Jobs Poised for a Comeback,” the researchers estimate that the housing industry’s construction employment fell by 1.5 million, or 42%, between 2006 and 2011. They also introduced the controversial notion that builders “hoarded” workers during the recession—that is, companies didn’t lay off as many workers as they could have, given the collapse in construction activity.

This hoarding premise is based on a calculation of the economic value added per worker, which Goldman estimates fell to $60,000 in the fourth quarter of 2012 from $80,000 in 2006. But productivity among construction workers has been gaining, leading the researchers to infer that hiring would again account for a larger share of future increases in residential investment output.

Consequently, the Goldman Sachs researchers believe that as home building expands, housing-related employment this year and next could grow at a rate of 25,000 to 30,000 jobs per month.

Other studies anticipate a similar job-creation spurt. Last December, Michelle Meyer of Merrill Lynch wrote that while employment did not keep pace with housing starts in 2012, “looking back at prior cycles, it appears that it is normal for construction jobs to lag output by about a year. We think we are on the verge of construction hiring. As demand for housing continues to improve, construction companies will likely become more comfortable expanding their workforce.”

Fannie Mae throws some cold water on these predictions.

Fannie basically agrees with Goldman’s estimates for job losses during the recession, stating that jobs related to the construction of residential buildings and specialty trade contractors declined by 1.4 million, or 41%, between 2006 and 2011. But where it differs from other analyses is in its observations about the connection between construction employment and housing starts, which Fannie contends was “highly correlated” in the pre-bust era. It expects this correlation to reassert itself over the next three or four years.

So if the industry eventually gets back to 2006-level construction, Fannie argues that employment should rise to that level, too. Fannie projects a healthy 20% increase in residential construction jobs between 2012 and 2016. But adding 412,000 new jobs over this period would only bring the industry’s employment ranks to about 2.41 million, or 1 million less than where employment stood in 2006.

The paradox in Fannie’s analysis lies in its estimate that at the peak of the housing boom, there may have been 1.6 million more workers than fundamental housing demand justified. That imbalance corrected itself only last year when, says Fannie, construction jobs were essentially in line with fundamental demand.

The good news is that this realignment “sets the stage for sustained growth in home building jobs over the next several years,” states Fannie’s analysis.

John Caulfield is senior editor for Builder magazine.