Labor

Where Did All the Restaurant Workers Go?

The latest jobs report shows that one sector accounted for all of September's losses. Not even two hurricanes can explain that.

Tough business.

Photographer: Charles Mostoller/Bloomberg

The 83-month-long streak of payroll employment gains appears to have been broken in September. The 33,000-job decline announced Friday by the Bureau of Labor Statistics could be revised away in coming months, and in any case it's more likely due to two big hurricanes than any economic slowdown. Not much to see here, in other words.

But there is one aspect to the payroll report that I couldn't take my eyes off of. A single sector -- food services and drinking places -- accounted for all the job losses and then some, with payrolls declining by an estimated 104,700 (0.9 percent) in just one month. Which means that all the other sectors put together added 71,700 jobs.

Why would this one sector be so dramatically affected? Are the Houston area (where Hurricane Harvey caused massive flooding at the end of August) and the state of Florida (which Hurricane Irma rolled through on Sept. 10 and 11) especially restaurant-and-bar-heavy economies? 1 No, not really: Food services and drinking places make up only a slightly higher percentage of payroll employment in metropolitan Houston and the state of Florida than in the nation as a whole.

One plausible explanation has to do with how employment is counted. The household survey, from which the unemployment rate is derived, showed gains in overall employment in September. It was only in the establishment survey, source of the nonfarm payroll numbers, that employment declined. One reason for the difference, noted the BLS, was that:

In the establishment survey, employees who are not paid for the pay period that includes the 12th of the month are not counted as employed. In the household survey, persons with a job are counted as employed even if they miss work for the entire survey reference week (the week including the 12th of the month), regardless of whether or not they are paid.

If food services workers are more likely than those in other sectors to miss out on a paycheck when their workplace is closed for the week -- which seems reasonable to assume -- then their payroll employment numbers would be more likely to take a big hit from a hurricane. With Florida and metropolitan Houston together accounting for almost 9 percent of national food services and drinking places employment in August, that, plus some statistical noise, may explain most of the September food services jobs collapse.

But after big gains in July prompted me to observe that food services and drinking places employment was on track to pass manufacturing employment in three years, Derek Thompson at the Atlantic to figure out that the sector had added more jobs than any other in 2017, and Joshua Wright at the regional-economic-data firm Emsi to calculate that it had accounted for more than 40 percent of job growth in some metro areas in 2016, 2 it does seem worth at least contemplating whether more might be afoot.

Up, Up, Up and Then ...

Employment at food services and drinking places, seasonally adjusted

Source: U.S. Bureau of Labor Statistics

Note: Y-axis does not go to zero.

OK, apart from the September drop, there's not any real sign of a slowdown there. There are modest signs of one since the beginning of the year, however, in the Census Bureau's monthly count of spending at food services and drinking places:

Slowing Down in 2017

Sales at food services and drinking places, seasonally adjusted

Source: U.S. Census Bureau

Note: Y-axis does not go to zero.

In a much-discussed piece published on Eater this week, Elizabeth G. Dunn attempted to explain the current downward spiral at several big casual dining chains thusly:

TGI Fridays and Applebee’s and their ilk are struggling as the American middle class and its enormous purchasing power withers away in real time, with the country’s population dividing into a vast class of low-wage earners who cannot afford the indulgence of sit-down meal of Chili’s Mix & Match Fajitas and a Coke, and a smaller cluster of high-income households for whom a Jack Daniel’s sampler platter at Fridays is no longer good enough.

As an explanation for why these restaurants have, as Dunn puts it, "subsisted in a dismal and persistent state of decline for about a decade," this makes a lot of sense. The real estate bust and recession pushed tens of millions of Americans off a financial precipice. But the casual restaurant chains' decline has accelerated over the past two years, even as the middle class recovered somewhat and overall restaurant spending rose. So maybe people are just tired of TGI Fridays.

Still, as is apparent in the previous chart, spending on food services and drinking places does seem to have plateaued this year. Also, the Russell 3000 Restaurants Index (which is dominated by McDonald's Corp. and Starbucks Corp.) is down 5.5 percent since hitting an all-time high in early June.

Down From the Peak

Russell 3000 Restaurants Index

Source: Bloomberg

It may just be that, after years of gains in restaurant spending and employment, a bit of a slowdown was inevitable. There's certainly not much sign that, the best efforts of Blue Apron notwithstanding, we're about to see a big shift back to cooking at home instead of dining out (or ordering takeout). At this point, in fact, the signs are mostly just muddled. But they're worth paying attention to over the coming months, in case a clearer signal emerges.

(Corrects percentage decline in payrolls in second paragraph.)

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

  1. Puerto Rico suffered a much worse hurricane hit than either Texas or Florida, of course, but (1) Maria hit after the reference period of the employment surveys and (2) Puerto Rican jobs data isn't counted in the national totals.

  2. Baton Rouge, Hartford, Wichita, Birmingham and Cleveland.

To contact the author of this story:
Justin Fox at justinfox@bloomberg.net

To contact the editor responsible for this story:
Brooke Sample at bsample1@bloomberg.net

Before it's here, it's on the Bloomberg Terminal.
LEARN MORE
Comments