Thread by @bencasselman

OK, after dealing with the nightmare that was the @BLS_gov site this morning, my usual jobs day chart thread to follow.

First, the big picture: Job growth is strong! More importantly, it is remarkably consistent. 103 straight months of growth, with quick rebounds from the occasional weak month.

Average hourly earnings were solid: up 3.2%, significantly better than earlier in the recovery. But while wage growth is better, it's not gaining steam. Basically just holding at this level. Suggests we've still got room to run.

The unemployment rate hit a 50-year low. And broader measures of un- and underemployment have likewise showed strong improvement (though the broadest, U-6, was flat this month).

But the unemployment rate fell for the "wrong reasons" in April. Unemployment fell, but so did employment (as measured in the household survey). Labor force participation fell.

My advice: Don't get too excited about the household survey data in either direction. It was weak this month, but the trends are good. At most, seems like there's been a bit of softening, but too early to say even that with confidence.

This much is clear: Labor supply has been *much* bigger than the unemployment rate alone would suggest. The strong labor market is pulling people in off the sidelines, which is part of why wages aren't rising faster.

More than 70% of the people getting jobs (moving into employment) were out of the labor force the previous month -- meaning they weren't actively looking for work, but got jobs anyway.

Adjusting for the aging population, the employment rate is now well above its prerecession level. Still well below the level of the late 1990s/early 2000s though. And with inflation still low, there's no obvious reason we can't get back there.

Turning to the establishment survey, that weak February number looks more and more like an anomaly -- and also less weak, since it's now been revised up to +56k (was originally reported at +20k).

Still, there were some areas of weakness. The slowdown in manufacturing hiring is now very clear (although March was revised up from a tiny loss to no gain).
nytimes.com/2019/04/04/bus…

In fact, the manufacturing diffusion index dipped below 50 in April, meaning a more manufacturing subsectors are now cutting jobs than adding them.

And the grim times continue in retail. The sector lost another 12k jobs in April, the third loss in a row. Retailers have cut nearly 50k jobs in the past year.

On the other hand, construction hiring was very strong, the latest sign of life in the housing market.

There was some talk that early hiring for the 2020 Census might boost April's numbers, but any effect was small. Government added 27k jobs, 11k of them federal.

Note that growth in *total* earnings (hourly earnings times total hours worked) has fallen back a bit lately. Consistent with evidence of slowing consumption growth since middle of last year.

(That chart is new this month. Curious if folks find it useful or hard to interpret. Or, I suppose, easy to interpret but still not useful.)

Good idea from @VirginiaJim: Here's contributions to total earnings growth. Makes last year's strength (and the modest drift downward since then) clear.

The long-term unemployment rate has finally fallen below its pre-Great Recession average. But still *way* above where it usually is in a supposedly booming job market.
cc @marthagimbel

@marthagimbel There are still more than 750,000 Americans who have been out of work for a year or longer, and 1.2 million who have been looking for more than six months.

Stories about structural changes in the labor market haven't held up well in this recovery. But the rise of long-term unemployment sure looks like an exception.