Poe Does Employment
JACKSON HOLE, Wyo. -- First you have to admit you have a problem.
The slowdown crowd has been promising softer employment numbers for nearly a year. One shop -- initials
-- predicted 170,000 new nonfarm jobs for October 1997 and promised that the employment picture was moderating. Payroll growth came in at 288,000 that month. It has averaged 262,000 since.
Four months ago -- seizing on a temporary downtick in moving averages -- another shop promised moderation during the months ahead. Two months ago -- floating the asinine notion that service employment was lagging what was "really" going on -- the same shop promised the same thing.
And every single month the poorest economic forecasters dismiss the employment report as backward-looking. Why? Because the job numbers keep refusing to go their way.
What happened during July-August was particularly disgusting even for this lot. Said seers swore up and down that the July employment report reflected more than just a strike -- that it was a mistake to ignore the obvious signs of serious slowing underneath. Worse still, they were loath to mention the ex-strike employment gain (207,000) that the
reported. But, lo and behold, today they are quick to point out that August payrolls grew "only" 215,000 excluding strike effects, and -- surprise -- they are just certain that apparent signs of strength are aaaaaaaaall strike. Merely this and nothing more.
There are two ways to analyze the employment picture. You can listen to
and a whole host of other market cheerleaders tell you what the markets want to hear. Or you can look at the numbers. And here they are.
237,000 during the first eight months of the year (including an average 254,000 gain over the last five months). Compare that to an average 282,000 increase for all of 1997 (when
boomed 3.9%) and an average 233,000 gain in 1996 (when GDP jumped 3.4%). Note that the 12-month moving average for job increases stands at 274,000 -- this marks the fourth highest such average in more than three years. Service-sector payroll growth, meanwhile, averaged 233,000 during the first eight months of the year (including a July-August average of 255,000). Compare that to an average 240,000 increase for all of 1997 and an average 202,000 gain in 1996.
Overall employment is now rising at a huge pace instead of a really huge pace, and service-sector employment hasn't slowed -- period. Is that the kind of slowdown the Fed is looking for? That is not it at all. And keep two things in mind:
- It takes monthly employment increases as little as 150,000 to keep the unemployment rate from rising.
The labor force grew at an average 176,000 per month last year. It grew at only an average 31,000 this year.
Thus, employment is likely to show some genuine slowing at some point for the simple fact that firms are running out of people to hire. But do note that this kind of slowing is not the kind that will relieve wage pressure.
Average hourly earnings are
at a 4.3% rate. That marks the fastest pace of wage acceleration of the cycle. Goods-producing wages are rising at a 3.3% rate (roughly twice the pace of the biggest inflation estimate out there). And on the other side of the economy -- the one that accounts for 82% of total output -- they are soaring. Wages in the wholesale trade sector are growing at a 4.7% rate. Wages in the broad service sector are rising at a 4.9% rate (their fastest pace of the cycle); in the narrow services sector, they are rising at a 5.3% rate (also their fastest pace of the cycle). Wages in the retail trade sector are rising at a 5.2% rate. And wages in the finance, insurance and real estate sectors are rising at a 5.7% rate.
These accelerations are quicker than either the rate of productivity increase or the rate of inflation -- or both put together. What to make of this, especially in an environment wherein firms have absolutely no pricing power? Think productivity (which rose only 0.1% during the second quarter). Think labor costs (which surged 3.9% during the second quarter). Think profit margins (which have shrunk for three straight quarters).
Think the Fed is going to ease against that backdrop? Forget it.
The index of aggregate hours
0.1% last month. It will post a 2.3% third-quarter increase if it turns in only a trend (0.2%) gain in September. That means GDP is very likely to rise at least that much -- hours-worked plus productivity equals GDP -- and, if productivity is still rising at the pace it did last year (1.4%), then GDP could easily come in as big as 3.7%.
And do note this: The people who scream that productivity is now so high that it will keep inflation dead forever are the same ones who turn around and attach teensy productivity estimates to the hours-worked figure. Sure, it's disingenuous, but that's one of the best ways to torture a slowdown out of the data. Perhaps turnips can ooze blood.
The manufacturing sector, you ask? Well haven't you heard? It sucks. It has chopped 149,000 jobs since February. The index of aggregate manufacturing hours is on track to plunge 3.9% during the third quarter following a 4.2% dive during the second. The factory workweek has fallen to 41.7 hours from 42.2 hours in eight months. Overtime hours sit at their lowest level since April 1996.
And those things make the three bullets above that much more impressive.
The Asian crisis has been hanging over the economy for more than a year. Trade (with a little help from the strike) inflicted unprecedented damage during the first half of 1997. The economic expansion is ancient against any yardstick.
So how come no one finds it amazing that employment
continue to print the way they do even in the face of all that?
There exists tremendous pressure on the economy to show a material slowdown soon. It may well do so later this year (even my first-grade nephew has a 50-50 shot at getting that one right). But there are no signs that it is happening yet.
going on and on about
this morning? One good thing came out of it. We can be pretty sure that those who criticize the notion most are the ones who understand it least.