Friday, June 5, 2020

Jobs Report Not Really All That Surprising

I am a bit taken aback at how shocked so many are about the new jobs report showing that net hiring in May was positive.  For regular readers here I have made several posts here noting that the US economy was almost certainly growing, probably for at least a month. The most recent was my one a few days ago on Rising Oil Demand, and an earlier one, where I was vaguer about the US economy, was the one on Rising Carbon Emissions.  It has been clear to me that the US economy hit bottom in terms of output about a month ago, which put it about a month behind the world economy as a whole and two months behind China.  All of this correlates with how the relative patterns of the pandemic have gone, with China a month ahead of most of the world and about two months ahead of the US.  I think it has been pretty clear that US GDP has been growing, so nobody should be all that surprised that the labor market has turned around and net hiring is now positive.

How did all this confusion come about?  I think the issue is that we get weekly reports on fresh layoffs as measured by new applications for unemployment insurance while we only get monthly reports on net hiring, with our monthly BLS reports such as the one that came out today and surprised the heck out of so many observers who should have known better.  I note that I did not forecast an increase in net hiring, but I had avoided making any forecasts on employment beyond the comment that it is a lagging indicator behind output, which would allow for net hiring to have still been negative. But I was somewhat mystified by what seemed to be such an disjuncture, clear evidence GDP was rising while there were these ongoing weekly reports of many more getting laid off.

The answer is now fairly clear.  Indeed lots of layoffs are happening and probably will for some time to come.  But, not only have those numbers been falling, but the ongoing layoffs are increasingly concentrated in certain sectors, such as education where net hiring was negative in May.  Indeed, we are likely to see a surge of layoffs in the local and state government sectors as those have not seen revenues rise and are not getting federal aid and also face balanced budget constraints.

However, hiring has been going on in other sectors, not publicly reported until today. Among those are hospitality and tourism, construction, and manufacturing (hence rising oil demand and carbon emissions).  Some numbers I have seen on some blogs, so not sure they are accurate, include a claim that a full 54% of the hiring occurred in the restaurant sector.  Yes, there have been a lot of reopenings there, if still somewhat limited.  Another odd figure Tyler Cowen reports on MR is that supposedly 10% of the new jobs are in the dentistry sector. Really? Who am I to say.

I do note that by some alternative procedures than reported, the unemployment rate should be 3% higher than reported, and thus would have increased.  Nevertheless, it does seem that there was a net increase in jobs during May.

What does this portend for the future?  I remain doubtful of a "V" shaped recovery, despite a lot of crowing about such today. This growth is so far fairly slow, and the decline in the unemployment rate not all that dramatic.  The pandemic is still expanding in some states, with Florida hitting a new daily high for new cases on Wednesday, even if new cases are gradually declining nationally.  Even if there is no second wave, fear of such and the continuing presence of the virus will hinder rapid growth for some time, probably at least through most of the third quarter. But I would say that the probability of a "U" shaped recovery has probably risen, with a possibly much more rapid growth rate in the crucial fourth quarter if indeed the virus continues not only to decline but to really stay down, thus reducing peoples' fears and allowing them to really get out there and spend and "go back to (almost) normal."

Barkley Rosser


Anonymous said...

Really nice analysis.

Anonymous said...

Devi Sridhar @devisridhar

Reflecting on early modelling: perhaps the obsession with deaths missed the larger issue: post-covid health complications. As an ICU doc told me, she’s less worried about # of beds & ventilators & more concerned about healthy people having long-term problems like lung damage.

1:10 PM · Jun 5, 2020

[ I am shocked at what I find is the extent to which people are paying no attention to social distancing now. I fear this is a serious mistake. ]

Kaleberg said...

How much of the employment pop is workers being rehired as required by the terms of Federal bailout grants/loans?

2slugbaits said...

I think we need to be a little bit leery of the May numbers. Roughly three-fourths of the job gains were in the services sector, and of that the Leisure and Hospitality sector accounted for well over half. The Retail and the Education & Health sectors accounted for the lion's share of the remaining growth in the service sector. But let's take a closer look at the Leisure and Hospitality sector. Average Hours Worked plunged in March from 25.8 to 24.1. Then in May Average Hours Worked jumped to 25.8. But Average Hourly Earnings jumped from $16.86 in March (what was a new high) all the way to $17.95 in April even though Average Hours Worked plunged. Then in May the Average Hourly Earnings fell back to $17.39. In short, we're seeing a helluva lot of instability in this sector. One has to wonder about some possible data quality issues. And then there's the suspicious Average Hours Worked across the entire economy. Average Hours worked fell sharply in April, but then jumped to a record high (going back to March 2006!!!) in May.

Finally, we should keep in mind that state & local government employment fell by almost 600K. said...


There is a lot of noise in May's data and also some of this is due to federal aid to small businesses that rehire their workers, with that aid disappearing in July. Another element is workers merely furloughed who count as still being employed.

Anonymous said...

Adam Tooze @adam_tooze

At least to judge by test results, the US, unlike the Europeans, appears to have given up on sustained COVID-19 containment. Consistent with the numbers for mortality too.

[ ]

5:30 PM · Jun 5, 2020

Fred C. Dobbs said...

If the recent employment jump is not just a blip,
and the stock market continues it's recovery,
maybe Trump will be swept back in, without
a need for last-ditch chicanery. As if!

After all, election day is still 5 months away.

But, do consider Biden's VP choices...

Joe Biden is under pressure to choose a woman of color as running mate. Who’s most likely? via @BostonGlobe

... So really the choices are Stacy Abrams, Val Demings, and Kamala Harris.

Picking Abrams could be good politics. Georgia is on the verge of being a swing state and having her on the ballot could really excite Democrats in her home state and force Republicans to defend what has traditionally been GOP turf. And even if Democrats don’t win Georgia in the presidential race, the turnout among Democrats could help their chances in not one, but the two US Senate seats being decided there this fall. But Biden has a Capitol Hill bias, and Abrams has never run anything larger than her state legislative office and her current nonprofit organization on voting rights.

The final two

If Biden wants to pick a woman of color to be his running mate, his choice comes down to Harris and Demings. Both choices may raise some red flags given that both have backgrounds that activists have criticized as being too pro-police. Harris was criticized during the presidential primary last year for, among other things, refusing to back a bill that would require her, as California attorney general, to investigate all deadly police shootings. Demings has been criticized for promoting excessive police tactics and for a lack of transparency when she was chief, even as violent crime went down during her tenure.

That said, Harris and Demings are probably Biden’s best choices if he is looking for a woman of color to be his running mate.

Then again, two months ago there was attention on which possible Biden running mate had the best background to handle the coronavirus. Two months from now, when Biden’s deadline comes up, the focus could be something else entirely.

ken melvin said...

If we've learned anything it should be that, first and foremost, the office requires competency; not someone who agrees with whatever one's pet reform is, mean as a junkyard dog, ... but first and foremost they must be able to competently run the country. Things like preparation for emergencies, response, listening to advisers, ...

Given Biden's age, this applies to his VP choice

Anonymous said...

Paul Krugman @paulkrugman

Aha. I was pretty sure that claims about health effects of quarantine were bunk, but here it is from the people who really know

Trump's pet theory about the fatal dangers of quarantine is very wrong
There’s little to no evidence that ‘deaths of despair’ track unemployment rates.
By Anne Case and Angus Deaton 

10:18 AM · Jun 6, 2020

Anonymous said...

Paul Krugman @paulkrugman

Putting yesterday's job report in perspective: many people have put up some version of this chart 1/

[ ]

Here's another way to put it in perspective: Bureau of Labor Statistics says that if measured properly unemployment would be 16.3%. That's far above Great Recession; how does it compare with Great Depression? Data from Millennial Historical Statistics 2/

[ ]

The reason it doesn't feel like the Great Depression is emergency relief, which has compensated many workers for lost wages. But almost all the relief will expire over the next 2 months 3/

This will cause great hardship and could cause a relapse in the economy too. Unfortunately, GOP opposition to further relief has been reinforced by 1 good month 4/

Anonymous said...

Barkley Rosser:

I do note that by some alternative procedures than reported, the unemployment rate should be 3% higher than reported, and thus would have increased.

[ No the same factor was there in April and would have increased the unemployment rate then by 5%. Paul Krugman and Dean Baker noted this. ]

Fred C. Dobbs said...

On V-, or W-, or wwwww-shaped recoveries...

One of my indexes consists of ten stocks,
the top five being Tesla, Netflix, Apple,
Microsoft & Amazon. In terms of performance,
Tesla & Netflix shine. But the other three are
now solid Trillion Dollar market valuation stocks.
Investors seem to be moving in the direction of
such megacorps. Is this what Wall Street
is trying to tell us? said...


You are correct that if one uses the alternative measure of UR for both April and May, the rate will still have declined from April to May.

On Econbrowser, Jim Hamilton has an excellent and fairly lengthy discussion of a variety of issues related to the numbers, with a lot of things off in various ways. The bottom line is that probably the increase in jobs and decline in UR were not as great as the official estimates suggest, but that indeed there still was some net gain in jobs and net decline in the unemployment rate, just probably less than reported.

Fred C. Dobbs said...

US recession began in February, ending longest expansion on record

BostonGlobe - AP - June 8

WASHINGTON — The U.S. economy entered a recession in February as the coronavirus struck the nation, a group of economists declared Monday, ending the longest expansion on record.

The economists said that employment, income and spending peaked in February and then fell sharply afterward as the viral outbreak shut down businesses across the country, marking the start of the downturn after nearly 11 full years of economic growth.

A committee within the National Bureau of Economic Research, a trade group, determines when recessions begin and end. It broadly defines a recession as “a decline in economic activity that lasts more than a few months.”

For that reason, the NBER typically waits longer before making a determination that the economy is in a downturn. In the previous recession, the committee did not declare that the economy was in recession until December 2008, a year after it had actually begun. But in this case, the NBER said the collapse in employment and incomes was so steep that it could much more quickly determine that a recession had begun.

“The unprecedented magnitude of the decline in employment and production, and its broad reach across the entire economy, warrants the designation of this episode as a recession, even if it turns out to be briefer than earlier contractions,” the NBER panel said.

The unemployment rate is officially 13.3%, down from 14.7% in April. Both figures are higher than in any other downturn since World War II. A broader measure of underemployment that includes those who have given up looking and those who have been reduced to part-time status is 21.2%.

On Friday, the government said that employers added 2.5 million jobs in May, an unexpected gain that suggested job losses may have bottomed out. A recession ends when employment and output start to pick up again, not when they reach their pre-recession levels. So it's possible that the recession could technically end soon.

Even if so, most economists expect a full recovery could take two years or more, with the unemployment rate likely still 10% or higher at the end of this year. With few Americans traveling, eating out or shopping at anywhere near their previous levels, consumer spending — the primary driver of the U.S. economy — could keep economic activity weak for many more months. ...

Roger Fox said...

This is a very professional analysis. I completely agree that China a month ahead of most of the world and about two months ahead of the US.