Marginal Utility

by Tom Bozzo

In my morning troll of business headlines yesterday, I came across this item from USA Today in Yahoo's news aggregator: "Wages grow at fastest [sic] in three years." It just happens to be totally wrong.

American workers are beginning to see long-awaited wage gains, though increases remain well below the levels prior to the 2001 recession...

Hourly wages rose 3.1% in the 12 months ended in December, the fastest pace since spring 2003. While improving, wage growth appears to remain below consumer inflation, which advanced 3.5% in the 12 months ended in November.

The 3.1% gain in hourly wages compares with a 2.6% rise in the year ended in December 2004 and 1.7% in 2003. Still, wage growth is lower than some economists expect in an economy with a 4.9% jobless rate.

"When you're looking at wages, it's clear that they are moving upward. It may be from a low base, but the trend is clearly up," says John Silvia of Wachovia.

If you read through to the point of seeing that the "long-awaited wage gains" still don't keep up with measured inflation, you might be underwhelmed.

But it's actually worse, as the actual data suggest that the inflation gains have so far managed to outstrip the nominal wage gains, so that "real" (inflation-adjusted) wages have been declining, not rising — exactly the opposite of what the story's lede suggests.

Here's the wage data with the inflation data:

Year	Wage Increase (%)	CPI-U Increase (%, NSA *)	CPI-U Increase (%, SA *)	Annual Real Wage Increase (%, NSA/SA)
2003	1.7%	1.9%	1.9%	-0.2%/-0.2%
2004	2.6%	3.3%	3.4%	-0.7%/-0.7%
2005	3.1%	3.8% (Nov.)	3.5% (Nov.)	N/A

Just how bad 2005 will end up being for wage earners depends on the December CPI release. If the headline CPI change ends up being zero, so the price changes through November are the full-year change, the real wage change rounds (up) to -0.3% using the seasonally adjusted data, or to -0.7% using the not seasonally adjusted data. My guess is that zero CPI is too optimistic, though I was too distracted with the holidays to have noticed exactly when gas prices started back up after the post-spike decline that dragged down the November CPI.

It would only take an 0.3% increase in the seasonally adjusted CPI for December to ensure that the 2005 real wage decline is just as bad as 2004's by either calculation. It would take a positive surprise just to get real wages "up" to the 2003's modest decline, when labor market conditions were much more widely viewed to be cruddier. So John Silvia of Wachovia gets the Reality Distortion Award's coveted rippled glass pane for spotting a trend in these data.

Don't even get me started on what screwy CPI measurement of house price inflation means for what inflation really was over the last few years.

I also found this closing quote amusing:

[Ken] Mayland [of Clear View Economics] says the wage gain "is not a worrisome situation for the inflation situation."

I would suggest pitchforks and torches if negative real wage growth were "a worrisome situation" for monetary policymakers in the intended sense.

Addendum: Barry Ritholtz has a very nice post, with graphs, on the subpar labor market recovery. The post was edited to clarify that the USA Today had buried the lede and then some.


(*) December CPI-U, U.S. City Average, all items, % change over December of previous year except as noted.