Wednesday, September 2, 2009
This chart shows the monthly changes in nonfarm payroll employment as estimated by ADP (Automatic Data Processing Inc.) and the BLS (Bureau of Labor Statistics). I would note two things: One, you don't need to be a statistician to look at this chart and realize that these two lines have a pretty tight fit; they don't exactly match up every month, but they do a remarkable job of tracking each over time. Two, there's been a rather significant improvement in the data this year; indeed, this is probably the clearest evidence that we are in a V-shaped recovery.
Skeptics will immediately object that the "recovery" in jobs I'm referring to this year is not really a recovery in the sense that we are adding jobs. But I would counter that while jobs are still being lost, they are being lost at a much slower rate. This is unavoidable whenever the economy moves from recession to recovery; you don't go from losing hundreds of thousands of jobs one month to adding hundreds of thousands the next month. This economy is a like a supertanker—it takes a long time to turn, especially after a sharp recession such as we've had, and particularly given the headwinds that we are facing in the form of misguided fiscal stimulus.
The market has been selling off of late as people apparently worry that the jobs situation isn't improving fast enough to support a recovery in consumer demand. (As a supply-sider this doesn't concern me, because demand is not the driver of the economy. I look instead at signs that production and investment are expanding.) The market's concerns were fueled this morning when the ADP estimate for August job losses came in higher than expected, leading many to project that Friday's employment report will also be disappointing. I have no particular insight into what the BLS number will be, but I'm pretty confident, judging from the trends of the past few months, as well as a variety of other indicators suggesting improvement in the economy (e.g., the Challenger layoff announcements, the tightening of credit spreads, the general decline in implied volatility, rising commodity prices, rising global trade, the steep yield curve, increased capital spending) that whatever the number is it will not be so different from the ADP number as to suggest that all of a sudden the economy is losing momentum or turning down.
I would expect to see both lines on this chart trending higher over the rest of this year, and perhaps reaching positive territory by year-end. That would be perfectly consistent with an economy that is recovering. Nevertheless, the employment situation is always a lagging indicator, and I don't expect employment gains to be strong enough over the next year to result in a significant decline in the unemployment rate. The recovery is V-shaped, it's just that the right side of the V isn't likely to be as sharply upward sloping as it could have been had we had better fiscal policy.
I could point to the fact that, following the 2001 recession, employment gains did not turn decisively positive until late 2003, almost two years after the recession ended, and hold that out as a model for what we might expect over the next year or two. But the improvement actually should come a bit faster this time, given that the economy has so much unused capacity. It will be a long time before businesses have to build out new capacity in order to hire people, which means that the marginal cost of a new worker will be relatively low. Plus, the sharp drop in activity last year resulted in a significant accumulation of cash reserves (and a reduction in money velocity), and the spending of those reserves could provide fuel for rising consumption and a further depletion of inventories, which would then encourage businesses to ramp up hiring and production.
Mike Hammill of the Atlanta Fed has a good analysis which shows that the sharper the recession, the sharper the recovery, and that consensus forecasts for future growth (and even mine) are therefore quite pessimistic given historical trends. Milton Friedman's Plucking Model for GDP, which I discussed here, also supports the notion that a sharp recession should be followed by an equally sharp recovery.
Still, it should not be surprising to see a few months here and there of disappointing jobs numbers. If Friday's number is disappointing and the market trades lower, that would present a buying opportunity in my view. All of the important indicators continue to point to a recovery.
Posted by Scott Grannis at 2:08 PM