Heres a hypothesis that may help explain the failure of wages to take
off during this expansion: workers have accumulated unprecedented levels
of personal debt, and this restrains their ability to negotiate for
higher wages.
Now for the story behind this hypothesis. It all goes back to 1986,
when I was working on my doctoral thesis. I was reviewing the
literature on the age-wage profile when my son Nick, who was 12 at the
time, walked into the office. What was I working on? he wanted to
know. I explained that there is a tendency for workers to make more
money the longer they have been in the labor market, and that this had
perplexed economists, since the wage trend continued long past any
evidence of improved skills or performance. I then explained efficiency
wage theory, which was the model I was working with at the time.
Workers bargaining position, I went on, is based on their cost of job
loss, which in turn depends on how long it would take to find another
job and how that job would probably compare to the one theyre in.
Employers may pay higher wages to increase the cost of job loss (making
the job more valuable compared to alternatives) in order to elicit
greater effort or obedience. The question is, why should they be more
inclined to do this as workers grow older?
Nick thought for a moment and then said that one explanation would be
that workers through their lifetime are accumulating savings. Since
they can live off their savings for a while, this reduces their fear of
losing their job and increases the pressure on the boss to pay more
money. I dont know whether there is any basis to this idea in
practice, but it is certainly theoretically consistent, and it merited a
citation in my dissertation when it was finally done.
Which brings us to the present. The facts indicate that, rather than
accumulating savings, workers are accumulating debt. The aggregate
savings rate is barely positive, negative in fact during long stretches,
and the growth of leverage is occurring across the income hierarchy.
Perhaps there is truth to this cost-of-job-loss story after all. I
owe, I owe, its off to work I go is an old joke, and there have been
many flip comments about the effect of a mortgage on a formerly wild
youths work ethic. But workers now face an unending stream of credit
card payments, car lease payments, and other forms of consumer debt
service that leave them a paycheck away from insolvency. Perhaps this
new economy debt peonage is the force that shattered the Phillips Curve.
How would one test this proposition? Obviously a time series wont do,
since so many other factors share the same time trend. Nor will a
straightforward cross-section work, since an individuals indebtedness
is as likely a consequence as a cause of changes in wages, stability of
employment, and other labor market characteristics. Perhaps a fixed
effects model might do the trick: we could follow a sample of workers
over time and distinguish between the accumulation of debt, which is
caused by wage and employment outcomes over the same period, and the
level of debt, which should be exogenous with respect to other
contemporaneous variables.
So all we would need is a set of panel data on workers that includes a
debt measure. PSID has wealth, but Im not sure about debt.
(Incidentally, net wealth is not a valid proxy for debt in this story,
since a large portion of wealth is not liquid. Net financial assets
might work...)
There you have it, a hypothesis and a suggestion for research. I have
more than enough to do, so Ill hand this off to anyone who wants to
play with it. Do the tests and let us know how it comes out. Just
remember to cite my kid.
Peter
