Recent employment landscape, sequel

This is a corollary to our successful article done earlier this week, regarding the heterogeneous differences in employment growth, across the top 12 U.S. metro areas.  One can also see this Wiley news for the Significance article, which was coincidentally published just prior to the June labor report.  How do we reconcile the strong June report (nearly 0.3 million jobs gained), with the shallow labor data we also see when looking against the broader population?  

In order to resolve these differences, see the actual government data below.  This is a clear view and the default screenshot from the Bureau of Labor Statistics (BLS) for the employment-population (emp-pop) ratio.

That illustration should look disturbing to any reasonable human.  And the flattish trend at the bottom of the hill, since 2010: that's the recovery from the financial crisis.  Those values don't look anywhere near the ones we saw pre-crisis (and those 2006-2007 data themselves never fully recovered from the levels we enjoyed during the 1990s internet technology boom).

Now it's easy to throw out some common hypotheses that present this in a positive way.  Perhaps people are suddenly living lengthier lives?  Perhaps they too suddenly have the desire to retire?  These hypotheses would alter the value of the denominator and numerator, respectively, of the emp-pop ratio.  And if true, then sure this would suffice.  We could even use this to support the other debate as to why the labor force participation rate also hasn't rebounded from the recent financial crisis.  

But the empirical data we have just doesn't support these hypotheses very well.  We can begin with our analysis of the labor force.  This is the same data that Forbes had interviewed me about for one of their publications at the time.  Additionally, when this is looked at on the pension side it also doesn't make as much economic sense.  Americans have not waited until just after the financial crisis when their asset values were badly cut, to then suddenly feel financially secure enough to retire.  The labor data discussed with the BLS also can't support that for those aged 65 and over.  Also checks on the actuarial longevity tables for the U.S. (or even across most developed countries) do not show enough of a longevity expansion in recent years to fully explain the lack of emp-pop recovery.  This is not to mention that population aging never necessitate a fixed-age retirement, as people had been able to enjoy working more past age 65 as seen by their re-entry rates back into the labor force after an initial retirement.

The much more important variable, for young urban professionals, is still a lack of enough meaningful jobs.  Of course this is frustrating to hear after seeing an average of >200k monthly jobs gains, so far during 2014.  But just take another look at the top chart above.  There are simply not enough jobs; that's the end of the story.

Leaving the demographic explanation aside, we can mathematically use an approximate rule of thumb to estimate the marginal change in the emp-pop ratio.  This can help us make sense of the moving parts behind various explanations which come together to help explain labor statistics we often see (e.g., the unemployment rate).

- change in population longevity

- organic population growth
- net immigration

+ change in labor force participation

- change in the unemployment rate

~ marginal change in emp-pop 

Now clearly it is not at all that simple, but it should give one a basic estimate of how the labor math (directionally) flows.  We would have to be practical and know that in a healthier economic state, we would not have had such a large sudden drop in the labor force rate; and we would have also still retraced half of the drop in the emp-pop chart above (this also accounts for actual population aging or "voluntary" retirement or both).

This research isn't designed to blame a particular political policy (and certainly no government stimulus six years ago would have made the current situation much worse), but rather to shed light on a glaring problem that is simply shuffled away under the headline numbers.  So let's explore another hypothesis and see if there are strong relationships of the emp-pop, to differences in employment changes shown across the top 12 metro areas.

In the chart below we show the employment growth in the top 12 city regions we have looked at previously, and here we cumulatively cover the past 4 years (through the recent May 2014 regional data).  We have also computed the difference in the emp-pop ratio using the same BLS establishment survey, so the closes population proxy is slightly different though directionally identical to the one shown above.  As an example, the U.S. saw emp-pop drop from nearly 63% in 2007, to 58% in 2010.  This is a -5% difference (58%-63%).  And the change in emp-pop since then has been only 1%, to 59%.  While the New York combined metro area saw emp-pop change by (a statistically similar) 2% difference, and hence this is the vertical chart coordinate for New York.

If demographic patterns occur slowly across the entire country, then we should not see any strong relationship here, other than by spurious luck.  But there is no evidence that those in Washington and Boston are suddenly living longer and retiring sooner, while those in Houston and Atlanta are suddenly dying broke.

Instead there appears to be a strong relationship between growth in employment, and any rebound we have in the emp-pop.  The bottom line is that we need to have government policies that help dramatically stimulate job growth, for half the major U.S. cities and other parts of the nation.  Among other issues, recall those are 4-year growth numbers in the chart here, so the annualized growth is a little less than ¼ what is shown.  Government policies are suggested here, since surging record business profits and executive compensation (here, here for examples) don't seem to have respectably trickled down on its own to a rebound in the topmost emp-pop chart above.  Nor down to other measures such as real median wage growth (here, here for examples).

Also, note that yellow sunny markers are used for 6 of the top 12 metro areas, while blue northern-arrow are used for the other 6 metro areas.  If there were no pattern in the types of cities (with their respective industry base) and their economic performance, then the yellow and blue markers would overlap.  Instead this chart immediately above only statistically reinforces our argument (in the prior articles) that professionals in some of the top northern cities continue to suffer even as most of the national growth occurs for different type of work elsewhere (e.g., some of the sunnier southern metro areas have some narrow, super-fast growth industries).

For U.S. readers: happy July 4^th!

Incidentally just found out in July 5 Bloomberg, or here.