Pictures of Job Market Slack, and Its Costs in Terms of #’s of Jobs

November 25th, 2013 at 4:09 pm

What with all this talk about the new Baker/Bernstein full employment book, including the musical version debuting tonight in DC (OK, it’s chin music, but still…), it seemed an opportune time to remind readers about the extent of the problem, including a new metric on the number of jobs we’ve failed to create by spending so much time in slack labor markets.

First, recall the concept of the non-accelerating inflation rate of unemployment, or NAIRU: the lowest unemployment rate consistent with stable inflation.  The figures and estimates below all accept the CBO’s version of the NAIRU, though Dean and I argue that it’s probably too high, meaning we could achieve tighter labor markets without inflationary pressures than current NAIRU estimates suggest.  But that said, we’d be very happy to reach 5.5% (their current NAIRU) sooner than later.

Consider this long term picture of the income distribution, typically referred to as “growing together, growing apart.”  It shows the growth in low, middle, and high family incomes, 1947 to 2011, indexed to 100 in the first year.  Different income classes moved together—all pretty much doubling through the late 1970s, and then diverged.  (Note: these data leave out the capital gains that have driven inequality to historical heights in recent years—thus, if anything, they understate the extent of “growing apart.”)


Now, consider the fact that if you compare the CBO NAIRU to actual unemployment, the jobless rate was “too high” only about a third of the time between 1949 and 1979, but 70% of the time since then.  Yes, the Great Recession boosts the latter result, but as the third bar in the figure below shows, it’s far from the whole story.


Source: CBO, BLS

No one is saying that’s the whole story by any means re the pattern of income growth in the above figure.  But neither is it a random coincidence.  As we show in a number of different ways in Chapter 2 of our book, the absence of full employment contributes significantly to wage and income inequality.

Since the actual unemployment rate is above or below (or equal to) the NAIRU in any given year, we can refine the analysis a bit by cumulating all the percentage points that the job market was taught or slack in these years.  For example, let’s say the NAIRU is 5%, and the jobless rate is 4% in year t and 6% in year t+1.  The unemployment gap would equal -1 in t and +1 in t+1 (4-5 and 6-5, respectively), so our cumulative measure would be 0 (-1 + 1).

The next figure shows that over the 31 years from 1949-79, the actual rate was a cumulative 15 percentage points below the NAIRU.  Since then, it has been 31 points above, though half of those points were due to slack since the GR.


Source: CBO, BLS

A final bit of fun with numbers turns those percentage points into jobs.  Since we know the size of the labor force in each year, we can say how many jobs we were either missing by having too high an unemployment rate, or added by being below the NAIRU.  For example, in 1975, the NAIRU was 6.2% but the actual rate with 8.5%, meaning the unemployment gap was 2.3 ppts.  The labor force that year was about 94 million, so 94*.023=2.2 million missing jobs.

Or, conversely, in 2000, the NAIRU was 5% but the actual rate was 4%, making the gap 1 ppt.  There were about 143 million in the labor force so we had 1.4 million more jobs than we “should have” if we were honoring the NAIRU (core inflation picked up a bit that year, but nothing too scary, and it had been decelerating before that, even though we were below the NAIRU).

The next step is to again sum up all of those negatives (missing jobs) and positives (“too many” jobs relative to the NAIRU).  Between 1949 and 1979, the job market ran hot to the tune of 11 million jobs, meaning that’s how many more jobs we had in those years counting times we were above and below the NAIRU.  Since then, the job machine has run cold such that we’re down 42 million jobs; 19 million if you stop counting in 2007.

Had we been at the NAIRU from 2008-12, we’d have had 23 million more jobs, and that’s factoring in the decline in the labor force.  That’s a lot of slack.  It’s a lot of missing pressure on employers to bid up wage offers.  It’s a lot of wage and income losses, and a lot of safety net to offset them.  And the income, production, tax revenues, higher living standards, and so on, from that missing work is gone forever.  They were left on the table and the table has passed into history.

Clearly, I haven’t discovered the argument, chart, or table that will focus policy makers on these issues.  But the evidence is there, big time, and ignoring it won’t make it go away.

Print Friendly, PDF & Email

2 comments in reply to "Pictures of Job Market Slack, and Its Costs in Terms of #’s of Jobs"

  1. urban legend says:

    Why is there such an intractable assumption that there is such a thing as a NAIRU? Shouldn’t the late 90s disabuse us of such a notion? With labor so weakened. and the likelihood of union aggressiveness a zillion years in the future, what is a plausible dynamic that would cause inflation to increase?