A very belated jobs day thread, after a morning spent scrambling to adjust to a very surprising report. (Suspect I'm not the only one who had to throw out a lot of prewritten copy this a.m.!)
The first version of my story is here: nytimes.com/2020/06/05/bus…
The first version of my story is here: nytimes.com/2020/06/05/bus…
First, let's be crystal clear: This morning's report was good news. The rebound in the job market started sooner than we expected. The unemployment rate didn't get as high as we feared. There are important caveats and cautions, but we shouldn't lose sight of that core takeaway.
A quick word on the data: This has been an unprecedented period, and it has strained the ability of our economic statistics to track it. Simply put, our models weren't built for this. So it's reasonable to treat all these figures as having more uncertainty than usual.
But there is NO evidence that anyone is mucking with the numbers. In fact, @BLS_gov has been admirably transparent about the assumptions and adjustments they're making, including publishing monthly FAQs: bls.gov/cps/employment…
I did a thread last year on why I still trust the monthly jobs figures, which I stand by. Former @BLS_gov Commissioner @EricaGroshen (an Obama appointee) has a good post today that discusses this as well:
ilr.cornell.edu/work-and-coron…
ilr.cornell.edu/work-and-coron…
OK, now onto the data! I'm mostly going to skip my standard charts, since it's already nearly noon and, besides, they're all out of whack.
So let's start with some context: Even after May's unexpected gain, we are still in a deep, deep hole. We've lost 18 million jobs since February (including 1.4 million in March, which is 2x as bad as we originally thought). 
The unemployment rate was 13.3 percent, which is still higher than any month since the Great Depression *other* than April.
And that understates the true level of pain, as @aliciaparlap dissects here:
nytimes.com/interactive/20…
And that understates the true level of pain, as @aliciaparlap dissects here:
nytimes.com/interactive/20…
But enough bad news! (Well, for now.) Where did job growth come from in May? A lot of it came from leisure & hospitality, especially restaurants. That's a sign PPP was working -- small businesses were bringing back workers.
But it wasn't just leisure & hospitality. In fact, most sectors were adding jobs, as the diffusion index shows:
That +2.5 million figure is a net number. It's interesting to break it down into job losses and job creation: BOTH were elevated in May, although losses weren't nearly as bad as in April. (Note this is household survey data, so doesn't exactly line up with the payroll figures.)
The number of workers on temporary layoff fell in May (but remains very high). That's probably mostly due to rehiring, but some may also be due to temporary layoffs turning into permanent losses. As @JedKolko notes, permanent unemployment is still rising:
As @BLS_gov noted in their release, they're still struggling with the issue of furloughed workers being mischaracterized as "absent from work." The unemployment rate would have been 3 points higher had they been properly counted.
Given the rebound in employment last month, it's disappointing not to see more improvement in the number of people working part-time for economic reasons. Still a lot of people out there working reduced hours.
The employment rate rebounded off its record low (I had feared it would fall below 50%). But full-time employment experienced a more modest bounceback -- job gains were disproportionately part-time.
Important to note that the drop in unemployment was not universal. The unemployment rate actually *rose* for both Black and Asian workers.
