The jobs report came in better than expectations today with gains of 1.8 million, but it's important not to lose the headline: the recovery has slowed.

Unemployment is down to 10.2%, and importantly permanent job loss has not gone up. Nor has it meaningfully declined however, and remains at 2.8 million near the peak of the tech bubble recession. A large enough number to constitute a meaningfully large recession on its own.

One caveat is that govt jobs were up a seasonally adjusted 301,000 but this is probably largely noise. Normally govt employment declines in July, necessitating a seasonal upward adjustment to keep it level. It declined earlier than normal this year, making the seasonal "wrong"

This month suggests that the slowdown didn't turn negative, but it did seriously slow down. At this pace of job growth, by the end of 2020 we will still be 4 million jobs short. Around 1.5 million will be permanent, which are much slower to replace.

This jobs report certainly could have been worse given a variety of indicators, but we should not let the expectations game move our goal posts: the recovery has slowed down, and this is very bad news. This is not V-shaped job growth.

The details of the report aren't great news. Yes, permanent job losses held steady. But the labor force participation rate declined. Greater levels of detachment from the labor market will take more time to recover from, this is bad news.

In addition, as @nick_bunker shows, the absolute rate of job loss remains seriously elevated. Above the worst month of the Great Recession, and in fact picked up this month. Businesses are failing, and new job losses are *still happening*.

We continue to see the accumulation of the kinds of economic scarring that does not bounce back at a pace of millions of jobs a month, but takes years and years to dig out of. It's important to not lose sight of where we are in an absolute sense, and get caught up in expectations