Bruce C. Fallick, Michael Lettau, and William L. Wascher write,
Given the relative magnitudes of the various measures of rigidity at the 1- and 2-year horizons, it seems clear that nominal rigidities are less important when one takes the longer view of wage changes over more than one year, suggesting that time is, indeed, an ally of wage flexibility. Even at the 2-year horizon, however, operative wage rigidity appears to have increased in the low-inflation environment of recent years.
Pointer from Mark Thoma.
The paper thus goes against my own views. The excerpt I pulled out is the one that comes closest to giving comfort to my way of thinking.
As you know, I think that many macroeconomists rely much too heavily on the sticky-wage story, which says that real wage rates are strongly countercyclical. I do not think of the job market as a single firm, laying off and hiring workers. Instead, I think in terms of PSST.
The data on Job Openings and Labor Turnover show millions of jobs being lost and found each month. When the creation of new jobs is 10 percent less than the flow of lost jobs, you see a large increase in unemployment. If the creation of new jobs is too slow, I do not attribute much of that to the stickiness of wages in existing jobs.
Given that it took until 2015 for labor utilization to recover from its collapse in 2009 (and it might be argued that labor utilization remains below trend), I think that the sticky-wage story is hard to defend as a main causal factor. It is even more difficult to argue that the 2015 recovery was due to a dramatic upsurge in inflation and consequent decline in real wages.
You mentioned the JOLTS report so I had a look at some numbers. Between November, 2007 and June, 2009 there were a total of 93.365 million separations. The average size of the labor force was 154.353 million. That means during the recession 60% of the labor force separated from their employer. I’m not sure how a wage “sticks” when an employee is no longer being paid by an employer.
“I’m not sure how a wage “sticks” when an employee is no longer being paid by an employer.”
There are ways that can happen. Reservation wages of the unemployed is one. But even when sticky reservation wages aren’t a problem, there are various difficulties in paying new hires lower wages:
http://econlog.econlib.org/archives/2013/09/why_dont_wages.html
Reservation wages are largely employers not wanting to hire someone who might be unhappy. Long term there is considerable hysteresis due to long term unemployment, currency and skill loss, and again, not wanting to hire someone who might be unhappy, and the deeper and longer, the greater the damage. The longer out of the workforce, the less likely to return.
Was sticky wages ever really meant to be taken seriously? If wage prices are sticky is the problem then, trivially, raise the money level or increase government spending. So it is a hand wave to get from unemployment to the desired intervention.
It’s funny that another way of saying sticky reservation wages is voluntary leisure. I don’t really believe that, but I suspect wage stickiness might be more real than the people who use it as a rationale actually believe.
A lot of the lost jobs are zmp (or negative marginal product) positions speculating on growth serving the prior future. Remaining jobs making more than the new jobs of the new normal does not prove wage rigidity.
Sticky wages are consistent with psst. You don’t know your own worth after a disruptive event. What drives you to change industries is realizing your best wage opportunity.