Jun 04, 2012 at 2:18 pm
Yesterday, I posted on where the jobs are, or aren’t. As promised, I wanted to get back to this wrongheaded idea that the problem isn’t lack of demand, it’s unmet demand as employers with job openings can’t find the workers they need.
As is usually the case in economics, there’s some truth in arguments like this, but the evidence suggests to those of us who’ve been looking into this explanation that any skills mismatch/unmet demand theory of our unemployment problem is very much incomplete.
My main point here is about wages. I’ve been in this biz for decades, and I don’t remember a time when employers didn’t complain about the quality of workers. So the fact that you hear people say this isn’t evidence. They have to back it up with action, and the action an economist would look for here is in the wage offers.
That is, you hire more workers because the alternative leaves revenue on the table. Given productivity levels, you’ve got more demand for whatever it is you produce than you can meet with your current staff so you add new workers to meet the unmet demand. And if you can’t find the workers you need at the going wage, then you raise it.
So, has wage growth accelerated? Nope. It’s generally followed the weak business cycle, as shown in the first figure.
The three lines are yearly growth rates for average hourly, weekly, and manufacturing (hourly) wages over the last few years. The two hourly wage series both slump down to around 1-2% and stay there—no evidence of wage pressure. I included the manufacturing series because that’s where you frequently hear the claim made that employers can’t find workers. Well, at least by this evidence, they may be saying that but they don’t seem to mean it.
The weekly wage series does seem to grow faster for awhile back there around the second half of 2009. But that’s purely a cyclical effect of average weekly hours. Average hours per week fell with demand in the heart of the downturn. When output began to rise, 2009q3, hours went up too, and that showed up in faster weekly wage growth, which has since flattened, so no evidence of unmet demand there either.
In a recent (excellent) paper, Berkeley economist (and good pal) Jesse Rothstein goes at this wage question in some detail and comes to the same conclusion re lack of evidence of unmet demand. Jesse points out the aggregate wage trends like those seen so far might suffer from a bias relating to composition changes over the downturn:
If the least skilled workers are the most likely to have lost their jobs in the Great Recession, changes in average wages will overstate what is experienced by individual workers. This could explain why wages appear to have risen quickly during 2008 and 2009 [you can see that a big in the figure above], when employers were shedding workers and there is little other evidence of labor market tightness.
He adjusts for this by a) following the same workers over time (so holding composition constant), and b) by looking only at the wage trends of workers with new jobs. His figure, shown below, fails to find any wage acceleration even when controlling for this potential source of bias.
(Observers of my first figure will not find this surprising—if the composition bias favoing skilled workers were present, we should see aggregate wage growth creeping up towards the end of that figure—and we don’t.)
Finally, you might reasonably ask whether the relative wage of more highly educated workers has been rising (see figure: it’s the ratio of weekly earnings of college/HS grads). And in fact, in was for awhile as the downturn hurt non-college grads more than college grads, but since the unmet-demand story has gained traction, the college/high school ratio’s actually been going the other way. (Rothstein take a much more careful look at this in his table 1 and finds no relative wage gains by more highly educated workers.)
Lots of commentators find all the above hard to believe, in part because the “we-can’t-find-smart-enough-workers” story is so pervasive. But in this case, money talks louder than words, and the money isn’t seein’ it.
Thank you for joining the conversation. Comments are limited to 1,500 characters and are subject to approval and moderation. We reserve the right to remove comments that: