There’s a question I’ve been asked several times since last Friday’s employment report: If demand for construction labor is so high, why did employment decline by 20,000 jobs in May? Given how hard it is to find workers, wouldn’t construction companies think twice before letting people go?
Here’s some relevant background: construction is coming back to life for economic, post-pandemic, and seasonal reasons, and industry demand for workers is accordingly high. Unemployment has been elevated for 14 months, and many governors are truncating the duration of enhanced federal unemployment insurance benefits. This serves as a perfect recipe for construction job growth.
Usually, when a perfect recipe goes awry, my spouse is involved, but I can't hang this one on her. What I can do is give you some plausible answers. Bon appetit!
Let's get cooking with some March 2021 Bureau of Labor Statistics data. March was a perfectly fine month for job creation. In February 2021, there were 143.27 million Americans on payrolls. One month later, the number was 144.06 million Americans on payrolls. That translates into 785,000 net new jobs.
The key word in this instance is net, and what that net number hides is the massive churn in the labor market. Every month, millions of Americans are hired, and millions of Americans are separated from their employment (be it voluntarily or involuntarily).
Now stay with me here. We call people who are separated from their work for whatever reason separations. We call those who are hired hires. My apologies for all the jargon. Subtract separations from hires and you have net change in jobs.
Imagine a fictional economy with two companies, Allmart and Wamazon (if you are a techy, you can use Gapple and Oogle). Say a worldwide pandemic causes Allmart to lay off 10,000 workers but Wamazon to hire 15,000 workers. The economy would have added 5,000 jobs that month.
Back to reality. In March, U.S. employers hired 6 million people! But 1.4 million were laid off, 3.5 million quit, and 334,000 were separated for other reasons. So, hiring 6 million people translated into the net creation of only a few hundred thousand jobs.
This is nothing new. Each month from 2001 to 2019, about 5 million people were hired on average, a shade under 2 million people were laid off or fired, and 2.6 million people quit their jobs. Another 350,000 people lost or left their job for other reasons (which BLS classifies as “other separations”), which brings average monthly separations to just under 5 million. That's churn, baby!
More precisely, an average of 4,968,232 were hired each month and an average of 4884,048 people were separated from their job each month. That gives us about 84,000 net new jobs each month.
Now, back to our question: If demand for construction labor is so high, why did employment decline by 20,000 jobs in May? Given how hard it is to find workers, wouldn’t construction companies think twice before letting people go?
Construction has a higher rate of churn than other industries, accounting for 4.9% of payroll employment but about 7.9% of hires and separations. Intuitively, this makes sense; contractors hire and lay off workers when jobs end, or activity ebbs in one region and picks up in another, or even when a project is significantly delayed.
Construction companies are most certainly clinging on to workers. In March, the most recent month for which we have Job Openings and Labor Turnover Survey data, only 150,000 construction workers were laid off or fired, about 36% fewer than in an average month. But workers can still quit, and contractors can’t afford to keep employees on their payrolls if those employees aren’t actively working.
Here's the big thing. The leisure and hospitality sector is bouncing back, adding more jobs than any segment. But those are low-wage jobs you say, especially compared to construction. For the most part, I agree. But there are low-paying construction jobs. And those restaurant and hotel workers are in such high demand that their average wage is now approaching $16/hour, rising nearly 5% over a recent 2-month period. One suspects that some construction workers are being lured away to leisure/hospitality, Amazon, and other segments that are presently hiring in bulk.
BLS data on hires and separations comes from a separate survey than the payroll employment report, so you actually get a slightly different number.
You are so uncannily accurate it rocks my soul