While waiting for new information, it’s always good to go over the old stuff and see how we’ve done so far. Barataria used to use the weekly unemployment initial claims as an instant guide to how the Depression has been running, but stopped a year ago. The rationale at the time was that we were close to equilibrium at about 360k jobs lost every week. There wasn’t much reason to expect a change.
How did that assumption fare? And was this really proof, along with job growth leading GDP growth, that we are in an unusual economic event? Let’s go back over some old ground and see how the old predictions went. Some of this went well and some not so well.
Let’s start with what didn’t go so well. With initial claims at about 360k per week the economy was losing jobs at a pace of 1.55M per month. Comparing that to the 200k net increase in jobs per month gave us a figure of 1.75M created every month. Whoops! It turns out that the rest of 2012 saw a much slower net job creation of about 160k jobs per month. With all the jobs created and destroyed in any given month that’s not much of a difference, really, but it shows how hard it is to take two really big numbers and make sense of a tiny difference.
But what stands out in the graph below is how initial claims has not fallen significantly in the last year, pretty much as we predicted. The first stage of the restructuring, wholesale destruction, did indeed end last year:
Initial claims have fallen to around 350k per week, or about 1.50M per month, and may be dropping more. It’s not a big difference, but it is important. Remember that we have two large numbers that we subtract to get a very small difference, so all that noise in the data makes everything dodgier.
Confirming that job destruction hit a plateau in 2012 gives us a bit more room to pay attention to the things we have been looking for in 2013, which are secondary signs that the economy is growing. Now that the net job growth per month is indeed 200k per month and we can expect that to be repeated when the ADP jobs report comes out (and it may have by the time you read this!). But the fall in initial claims is enough to have accelerated the net gain in jobs from 160k to 210k or more. In other words, job creation has not accelerated yet – we have simply stopped losing as many.
We have not really turned the corner yet, but we do appear to have printed a nice bottom. In addition to the things we’ve already talked about looking for this year we should see an actual increase in the net job creation as the year goes on to well over 200k per month (and I expect March data might well surprise to the high end, more like 220k).
So how much is job growth leading economic growth? Not as much as we have hoped, but it is still ahead of GDP growth. That hasn’t been north of 2.8% in real terms since 2006, a level that in every other postwar year would generate a net gain of zero jobs per month. At the very least, this is not your typical recovery.
Does it make sense to look back on initial claims data? It hasn’t moved much in the last year, but then again job creation apparently hasn’t budged much either. Looking back over the indicators we’ve abandoned shows us that it was a good move to drop them, since they didn’t change much at all, and that 2012 was indeed the year that the corner was turned.
Back to wishing for a good 2013!