The Most Awesomest Recovery

An economist named Karl Smith, who writes at a blog called Modeled Behavior, seems to have invented a statistic that finally can demonstrate what an awesome recovery we are in. He calls it the "Labor Soakage Rate". I don't know if he actually did invent it - he doesn't make a point of claiming he did. But if you Google "Labor Soakage" you'll find references to his blog post and a bunch of stuff on soaking up water leaks and the work needed to do so. So I'd say he invented it.

At any rate, he's using it because it appears to show a very steep rate of job growth. He calculates this "soakage" rate as the percentage change in employment minus the percentage change in the civilian population from one month to the next. So, even if jobs are being lost, and the percentage change in jobs is negative, as long it's better than the prior month, the graph will trend upwards. So it in no way measures the magnitude of job growth, or even lets us know anything about job growth itself - it only tells us about the rate of change in the change - the 2nd derivative, basically - and that this change in the change is very high!

Now maybe this is a useful statistic - if people working in banks had looked at the 2nd derivatives of housing prices, a lot of pain might have been avoided. And of course generally it's useful to find peaks and troughs. But all you've needed to be doing over the last few months is pay attention to the weekly unemployment claims numbers - which have generally dipped below 400k of late - to know that there is some turnaround in the labor market. But this statistic gives you no feel for the what kind of recovery we're in. For that, I'd prefer to compare the actual magnitude of job gains.

One other dubious approach he takes is to compare this recovery to the last one - but early-2000's recovery was routinely derided as the Jobless Recovery. So it's not much of a comparison. Plus, that recession was relatively shallow, as was the 1991 recession. The last recession to rival this one for depth was the 1982 recession, where the jobless rate actually exceeded 11% at one point. But the economy began to recover sharply in 1983. The current recovery is now in its 29th month. The chart below compares the accumulated increase in employment net of civilian population growth over the first 29 months during the recoveries from the 1982 and 2008 recessions (or the Reagan vs. Obama recoveries, if you will). You can see what a dud this recovery is.

From this view, one can't even discern any improvement in the current labor situation whatsoever.

Yet it's hard to pin the blame on Obama. This dismal job situation has been pretty much the "new normal" since 1999. Note this chart, which shows total accumulated job creation net of population growth since the 1982 recession.

From 1983 thru 1999, the economy managed to create at least one new job for each additional adult. But since job-creation peaked at 6 million accumulated net new jobs in December 1999, there has been a steady erosion in the number of working people with respect to the adult population, with only a slight flattening of the decline during the housing boom. Another way to look at it is that in this century, the civilian adult population has added 30 million more people than jobs have been created. The most obvious trend behind this is the baby-boom retirement explosion, but high levels of immigration is no doubt involved as well. Not that immigrants are jobless, but clearly ten-or-so million jobs have gone to immigrants that have not gone to the native-born.

Is there any way out? Over the next decade, the baby-boom retirements should peter out, and baby-boomers should start dying in earnest during the 2020's. So if you're young enough, I suppose there's some reason to hope.