If you don't have enough patience to read through to the end of this post, the answer is I don't know. But, as I noted last week, Daniel Gross doesn't share my confusion:
... the most recent jobs report from the Bureau of Labor Statistics? The employment-population ratio is 62.8 percent, lower than it was for all of the second half of the 1990s. This is not what full employment looks like.
That thing about what full employment looks like is a bit tricky -- and not easy to resolve.
In related items last week, Barry Ritholtz -- reinforcing observations made some time back by William Polley and pgl, to name just two -- documented the recent decline in overall labor force participation rates, and the particular patterns of participation across different age groups. Here, as a reminder, is what the record looks like:
The big stories, of course, are the rising participation rate of workers 55 and over -- a trend that began in 1996 -- and the more recent, but equally dramatic decline in participation by workers in the 16-24 age group. Despite the fact that labor force participation trends of men and women have been quite different in other age groups, they have been remarkably similar for the oldest and youngest workers:
Ones view on what employment growth should look like depends on how these trends are interpreted. A few months ago, the Boston Fed's Katharine Bradbury made a splash with a study that calculated what employment (or the unemployment rate) would be, if only participation rates would revert to their pre-2001 levels. Jim Hamilton roundly criticized the study -- or at least the implications others were drawing from the work -- noting the hazards of picking a particular point as the benchmark for what is "normal" in a series that is trending in one direction or another.
Yet the insistence persists that falling participation rates (or employment to population ratios) reflect fundamental weakness in labor markets. Certainly that explanation is a tough one to make for workers in the 55+ group, since their participation rates have risen right through the post-2000 period. In fact, the timing of the take-off in the labor force participation for this group coincides nicely with the 1990s employment booms. So, what if we do the Bradbury experiment in reverse -- ask what employment would have looked like in the latter half of the 90s if the participation rate of the over-55 bunch had remained at its average for the period from 1983-1995. Here's your answer:
If the actual employment growth in that picture seems a little different than what you are accustomed to seeing, it is because it is based on the Bureau of Labor Statistics' household survey, rather than the establishment payroll series. I generally prefer the latter, but using the household series has the advantage that it comes from the same source as the participation data.
Although it may not look like much, the experiments in the picture above imply that rising participation rates added roughly 200,000 jobs per year over the period. (The calculations in the graph assume that additional entrants find jobs at the same rate as the entire population of 55+-year-olds in the labor force.) That' s about an extra month's employment every 365 days; a full year's worth of jobs for the entire span from 1995 through 2000.
Was it the exceptionally good economy of the latter 1990s that prompted such a burst of later-life job-market activity? Not likely -- unless you want to concede that the last four years have been particularly attractive for job seekers as well (or at least older ones). Better, I think, to look for a smoking gun. Here's a candidate: The increases in the eligibility age for full social security retirement benefits (legislated in 1983), which affected its very first set of folks in the 55-and-over club in 1993, and every subsequent wave of AARP-eligible workers since (in increasing measure).
But wait, Dave (you might say), if rising participation rates contributed to higher employment growth in the 1990s, wouldn't that be all the more true after 2000, making the performance of job creation since then look all the more dismal?
Yes, indeed. But enter, now, the declining participation rates of the youngsters. What if we repeat the experiment above, adding the assumption that people in the 16-24 age-bracket joined the labor market in the same percentage they did in the period from 1992-2000 (over which time the participation rate for this group was relatively stable). Here's where we would end up:
Nothing changes the fact that no jobs were created during the recession year 2001, or that things got worse in the first year of the recovery. But the story is very much different thereafter. If not for the declining participation of the young, the plot may well have been about how robust the employment turn around has been.
This, of course, is just a different way to state the Bradbury results, albeit with a radically different interpretation. If you believe that the trend participation rate of the youngest workers is captured by its pre-2001 -- the theme of the Bradbury piece -- the gap between the green and blue bars is a measure of how weak the labor market remains. If, on the other hand, you believe that trend has permanently declined, then we are at "full employment" (by definition), there is little cause for anxiety, and one might even conjecture that the job market has been pretty darn strong in the past three years.
So this is the question: Is there any reason to believe that young workers have, of their own volition, and for whatever reason, delayed their entry into the working life? Is there a smoking gun that explains the labor force behavior of 16 to 24 year olds?
More to follow.