Mitchell’s Musings 7-10-2017: My Guess on Sluggish Pay Increases

08 Jul 2017 9:01 AM | Daniel Mitchell (Administrator)

Mitchell’s Musings 7-10-2017: My Guess on Sluggish Pay Increases

Daniel J.B. Mitchell

There are beginning to be headlines and news items about labor shortages. A recent article from the New York Times is headlined “Lack of Workers, Not Work, Weighs on the Nation’s Economy.”[1] California farmers are complaining they can’t get enough workers.[2] Such anecdotal reports are not surprising given the fact that unemployment is down to levels that preceded the Great Recession. And the job openings (vacancy) rate is higher now than then. (See charts 1 and 2 below.) More demand; less supply.


Chart 1: Unemployment Rate: June of Year Shown


Chart 2: Job Openings (Vacancy) Rate


However, the reports on the labor market are telling us that wages have been slow to increase, despite labor market pressures. Nominal wages, measured by average hourly earnings, were increasing about 2% per annum during the aftermath of the Great Recession. Now the rate is about 2.5%. (Chart 3) So there is some acceleration in pay, but not as much as might be expected.[3]


Chart 3: Average Hourly Earnings for Private Nonfarm Employees, 12-Month Percent Change: June-to-June


It’s also interesting that employers – although they may complain about the difficulty in finding workers – are not behaving as if there is a shortage, even if they are talking about one. Here is another anecdotal report:

…Looking for a job has never been fun but lately it seems to be getting more complicated. The U.S. is at close to full employment, but companies are getting pickier and less responsive at the same time, as Kristen Shattuck found out recently. Shattuck got a shock when she started looking for a new job last year. She’s an education consultant and found the process much more involved than the last time she was on the market back in 2011. Before her interview, organizations asked her to write a personal statement, submit strategic plans, proposals, and watch videos, which she then had to give written feedback on… So why do job hunters have to go through all this? Allison Hemming runs The Hired Guns, a New York City recruiting firm. Her clients are digital companies. “They’re looking for more and more out of each individual person that they hire, and this is their way to manage the risk,” she said. “I think some of it can be ridiculous and too long.” But she added that candidates need to look at it like dating. You wouldn’t want to marry someone after the first date, right? She said these interviewing marathons are a way for companies and candidates to make sure you’re the one….[4]

There is a problem with the explanation of putting job seekers through all kinds of hoops. The rationale seems to be that the hoops are due to the fear of making a hiring mistake. But that story doesn’t make sense. Let’s put aside the issue of whether the hoops have any correlation to later job performance. Yes, there are costs of making a mistake. But these costs can be mitigated by such devices as probationary periods. Is there any evidence that the cost of making a mistake is higher than it was in, say, 2006-2007 or 2000-2001 (previous business cycle peaks)? I haven’t seen any such evidence.

It’s worth noting that there is another stream of recent news stories about robots taking over jobs. In that story, there really isn’t a labor shortage because robots are raising productivity so much that more labor isn’t needed. Workers are unnecessary, so why pay those that are hired more? But the robot story flies in the face of available evidence. The output per hour numbers don’t show a blast of productivity. The last productivity blast occurred in 2009-2010, when employers dumped labor in a panicked response to the drop in demand. (Chart 4)


Chart 4


You can always argue that maybe the real output data are somehow missing the growth in output due to all this alleged productivity. But if you take that approach, you have to explain why measured unemployment is so low. You have to explain why vacancy rates are so high. Are those numbers wrong, too?

OK. So what IS happening? All I can venture is a guess as to why wages are at best reacting very slowly to a tight labor market. The first three minutes or so of the “Marketplace” public radio program for July 7, 2017 deal with the wage issue.[5] One employer says that, yes, she is raising wages but not more than her competition. If she raised them faster, she wouldn’t be competitive, she says. A construction employer is reported to be worrying that if wages go up, newly built homes won’t be affordable. There is an element in such stories of a labor-buying cartel, even though these employers are small and competitive. How can that be?

Let’s look at the idea that you can’t raise wages more than your competitors or you won’t be competitive, or that your product won’t be affordable. Presumably, if you can’t hire all the labor you need, you are giving up business. Your labor demand is a reflection of product demand; no one hires workers for the sake of it. So, apparently, you have customers lined up and a margin for (profitably) raising wages to get the labor you need.

Nonetheless, if all employers in your market think that way, collectively profits might be reduced. A strategy of everyone just doing what the others do – even if all experience a labor shortage – could be net beneficial for the group. Regardless of how the behavior is rationalized, everyone doing what everyone else does is a form of labor buyers’ coordination that can benefit employers at the expense of workers (monopsonistic behavior). And there is one thing we know from surveys of employers about how they set wages. The universal answer is that they find out what others are doing. They may do their research informally. They may get reports from trade groups or other sources. They may hire pay consultants who have survey data. But imitation of other employers is the key feature.

In that sense, current behavior – grousing about labor shortages but not competitively raising pay in response – is not abnormal, at least for a time. So the question is whether there is something now that is different from past recoveries. The most comparable recovery was the episode in the 1980s. It started with two back-to-back recessions, the second of which was quite severe. These recessions, however, were in some sense less dramatic than the Great Recession. They were essentially engineered by the Fed to combat inflation. In contrast, the 2008-9 event was characterized by a collapsing housing/mortgage market that led to a collapsing financial sector which, in turn, had to be bailed out to avoid further disaster.

The profile of the recovery was also different. Unemployment shot up in the second of the two back-to-back episodes, but then came down relatively fast after it peaked. (Compare Chart 5 below with the earlier Chart 1.) That profile is in contrast with the Great Recession in which unemployment peaked, and then only slowly came down. During the Great Recession, there was endless talk of a “new normal” in which the economy would forever be sluggish and workers would just have to adapt to their diminished prospects. In human resource/employer circles, the new normal talk after the Great Recession was especially pronounced. I don’t recall anything comparable to it having occurred in the 1980s.[6]

Wage behavior was different in the 1980s episode. (Chart 6) Pay increases were high in the inflationary late 1970s, but in the face of the second of the two back-to-back recessions, they started down.[7] By 1986, however, the downward pressure on wage had reversed. Thereafter, pay increases rose.


Chart 5: Unemployment Rate: June of Year Shown


My guess is that the accumulation of all the new normal talk among employers is now showing up in a more gradual adaptation to a tight labor market than might have been expected, based on historical data. How long will that retarding effect last? We don’t have past experience to go by to answer that question. There was only one Great Recession and it has left economic and political scars.


Chart 6: Average Hourly Earnings for Private Nonfarm Production and Nonsupervisory Employees, 12-Month Percent Change: June-to-June





[3] Confession: Back in 2015 in a Mitchell’s Musing, I interpreted a blip up in wage change as a sign the Phillips curve had returned. [] The curve isn’t dead, but it is less healthy than I had thought.



[6] Indeed, the 1980s saw the beginning of the overheated “end-of-the-job” talk that continued into the 1990s and after (caused by the growth of temp agencies and related practices). “End-of-the-job” postulated that the labor market was becoming a spot market. Such chatter was an earlier version of the contemporary “everything-is-turning-into-a-gig economy” talk. Pay in a spot labor market should be highly responsive to labor shortages. It presumes low costs of hiring and firing and no real employment relationship. There are no “employers,” just users of labor in the same sense that there are users of copper.

[7] Chart 6 uses the less comprehensive measure for production and nonsupervisory workers rather than all employees, since the all-employees series was not available in the earlier period. 

Employment Policy Research Network (A member-driven project of the Labor and Employment Relations Association)

121 Labor and Employment Relations Bldg.


121 LER Building

504 East Armory Ave.

Champaign, IL 61820


The EPRN began with generous grants from the Rockefeller, Russell Sage, and Ewing Marion Kauffman Foundations


Powered by Wild Apricot Membership Software