The Washington PostDemocracy Dies in Darkness

‘Great Resignation’ Wasn’t That Great, Historically Speaking

The rate at which US workers are quitting their jobs, which has become a much-watched economic indicator during the current economic expansion, fell in January and is now well below the peaks that sparked talk of a “Great Resignation” in 2021 and 2022. But at 2.5 quits per 100 jobs, it’s still higher than at any point before the Covid-19 pandemic.

Then again, these numbers go back only to December 2000, which is when the US Bureau of Labor Statistics launched its Job Openings and Labor Turnover Survey. That was at the tail end of the long 1990s boom, with the unemployment rate beginning to rise in January 2001 and a recession starting that April, and the quits rate in early 2001 was only slightly below where it is now. Quits are seen as a measure of labor-market strength — people don’t leave jobs voluntarily unless they think they can find new ones — and the past two decades weren’t exactly known for strong labor markets. Maybe it’s the sub-2% quit rates of the 2000s and 2010s that were anomalous, not the 2.5% we’re seeing now.

If only we knew what quit rates were before December 2000, right? Well, we do, sort of. Not for the 1990s or most of the 1980s, and not for all industries. But for decades the BLS conducted a Manufacturing Labor Turnover Survey, only to discontinue it in 1981 because of budget cuts. Last spring, Ira Regmi of the Roosevelt Institute and Bart Hobijn of the Federal Reserve Bank of San Francisco (he has since moved to the Chicago Fed) both used post-World War II MLTS data to show that that the recent Great Resignation was no record-breaker, at least not in manufacturing. My main contribution here is to run the numbers (harvested from the Census Bureau’s Historical Statistics of the United States: Colonial Times to 1970 and back issues of the BLS’s Monthly Labor Review) all the way back to 1919.(1)

The Metropolitan Life Insurance Co. started the turnover survey in 1926, subsequently collecting data back to 1919 from the companies it surveyed. In 1929, it handed the project over to the BLS. The early surveys were pretty small, with MetLife sending questionnaires to 160 large firms in 1926 and 350 in 1929, but by the mid-1960s the BLS was surveying 40,000 establishments (for JOLTS it now surveys 21,000). So while some of the early numbers might not be perfectly comparable to today’s, the overall historical picture provided here is probably pretty accurate — and it shows that the manufacturing quits rates of the past two decades have, in fact, been anomalously low, with 2009’s 0.7% the lowest on record.How well do these manufacturing quits rates reflect overall labor market health? Well, from the early 2000s until three years ago, quits were much lower in manufacturing than in other sectors. This makes sense: US manufacturers hemorrhaged jobs from 2000 to 2010 amid the “China shock,” and even now they employ 25% fewer people than they did in 2000 while overall nonfarm employment is up 18%.

So while the manufacturing labor market of 2009 and the early 2010s was possibly even weaker than that of the Great Depression, the overall labor market may not have been. Since the pandemic, though, manufacturing and non-manufacturing quits have come much closer together, meaning that current manufacturing quits may well provide a good measure of overall demand for labor relative to past eras.

What they show is, again, that the “Great Resignation” of the past couple of years was nothing special. This is what the US labor market is supposed to look like when the economy is growing.

A return to such conditions is obviously good news for workers, especially the younger, lower-paid ones who drove most of the recent quits increase, according to Hobijn’s research. It’s more complicated for employers and investors, as the wage increases needed to keep and attract workers could cut into profits and put upward pressure on inflation — although they could also bolster productivity. In any case, while a recession would drive quits down temporarily, demographic trends point to this being the new normal. Get used it.

(1) The BLS also collected some labor turnover data from 1910 to 1918. The numbers aren’t entirely compatible with those from 1919 onward, but DePaul University economist Laura Owen converted them and put them in a chart in 2004, in case you really want to see (the quits rate appears to have been quite high then) .

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

Justin Fox is a Bloomberg Opinion columnist covering business. A former editorial director of Harvard Business Review, he has written for Time, Fortune and American Banker. He is author of “The Myth of the Rational Market.”

More stories like this are available on

©2023 Bloomberg L.P.