How'd that happen?

Wednesday, May 02, 2007
All the squawking about the lousy the U.S. job market seems finally to have died down, but not being one to refrain from kicking a dead horse, here are several indicators of just how strong the labor market really is:

1) First the obvious - the unemployment rate sits at 4.4%. That's low.

2) The 12-week trailing average of the Job Insecurity Index (TM) - a measure of the proportion of the "insured workforce" (people who are eligible to file for unemployment) that actually files for unemployment each week - sat at 0.245% as of the end of March, just above its all time low (since 1971) of 0.221% recorded in May of 2000, and lower than at any point between 1971 and July 3, 1999. Oh, for the good old days, when the JII could reach as high as 0.832% (April 5, 1975), and never dipped below 0.448% (January 1974 through December 1983). [Note: I've modified the JII slightly from the previous post linked above, now using the total insured population rather than total civilian employment as the denominator. I'll probably modify it again in the future to use insured population less continuing unemployed, but hey, this is an iterative process.]

3) The 3-month trailing average proportion of the Civilian Noninstitutional Population that says it wants a job now but is not looking for one (primarily because of discouragement - "there aren't any jobs out there!") hit a new record low in March, 1.95%, for the third month in a row. The previous record was in June of 2006, and before that April of 2001. That should put to rest the idea that the unemployment rate is low because people have given up looking - a favored canard of those with a vested interest in creating the perception of a weak labor market. In fact, a smaller percentage of people have given up looking than ever before in recorded history. Or at least since 1994, which is the limit of the data I have available. Oh, and in March of 1994, it was 3.26%, which is a lot higher than it is now, so you can almost certainly go back to the 1980s before you have any chance of finding a lower one.

4) The ratio of people who have a job to people who want a job stood at 93.0% in March. The record (again, since 1994) was 93.4%, recorded in October-November of 2000. In March of 2000, at the height of the boom, it was... 93.0%. Before September of 1999, it never reached 93.0% - even with rounding.

5) As of March of 2007, the labor force participation rate (66.2%) sits 1.1 points below its high of 67.3% recorded in January-April of 2000. The proportion of the Civilian Noninstitutional Population not participating because they simply don't want a job right now (31.9%) is 1.3 points higher than it was in January-April of 2000 (30.6%), explaining the entire drop in the participation rate (and a little bit more).

The drop in the participation rate is part of a long-term demographic trend toward more people who are "of age" to be included in the Civilian Labor Force not seeking a job because they don't want to work. Just to be clear - it is not part of a nefarious plot by the Bush administration and their cronies at the Bureau of Labor Statistics to make the unemployment rate look lower than it actually is. I know that's been a worry of some people out there, and I hope this eases your minds.

The real reason, and I don't have numbers to back this up, but I'll eat my hat if it isn't the real reason - as the Baby Boomers have aged, some proportion of them have also retired early. So yes, the participation rate is lower. It's lower because a greater proportion of people have retired. Early. Heaven forfend.

That's one of my favorites, because people have been so fond of saying "well, if the participation rate were as high as it was then, unemployment would be a lot higher". In fact, though, if the proportion of people who had decided they'd rather not work had stayed the same the participation rate would be slightly higher now than it was then, and they'd be making the opposite adjustment. If they were intellectually honest, that is.

So somehow, while the punditocracy was shouting about the soft labor market and the jobless recovery, we managed to rocket back to conditions remarkably similar to those at the height of the 1999-2000 boom. Astonishing, isn't it?


Rick Ballard said...

That's very nice work, Morgan. Completely understandable and well documented.

There are a couple of other interesting situations to consider (actually side issues) wrt employment. The first is the fact that employment is holding so well despite the slump in home starts. The year over year data for March show a drop of over 400K units. That's a net hit of a minimum of 1.2M jobs. If you plug in a modest multiplier the actual loss is over 2M. That translates to over 1% of the workforce. While I agree with you completely that early retirement accounts for a great deal of the decrease in participation rate I wonder if the rise of the parasitic (sort of) trustafarian class doesn't enter into the equation.

The second item of interest is that if you tie the numbers for those entering the workforce to the current employment numbers you come up with a decent scenario for some stiff wage inflation over the next decade. There aren't enough people available to maintain growth in the total number of employed. I really think that's a very large reason why Bernanke and the Fed are holding firm on rates - they just don't need to worry about the job creation side of their mandate much at all.

Morgan said...


Great thought wrt Bernanke, and I absolutely agree about coming wage inflation - anyone under the age of 18 is going to walk into a labor market like nothing we've seen in recent history.

In early 2006, the KC Fed came out with an analysis that put the trend number of jobs needed to maintain full employment at 120,000/month over the next 10 years (as opposed to the 150,000 that had been the rule of thumb) - so they're wise to the demographic side of things. But I honestly think that's too high. Maybe we need 120,000/month today. In 5 years, we'll need half that if the Census bureau's projections are right. And that's assuming that "payroll employment" is the only kind that counts.

Europe is seeing record low unemployment right now, and I think it's evidence of the same demographic trend playing out there.

It looks like any job losses from the housing slump are being swallowed up by an otherwise booming economy. In a labor market like this, just about anyone smart enough to hit a nail with a hammer can find other employment, and I suspect that a lot of people formerly employed in constuction (and real estate) were greeted warmly when they walked in a prospective new employer's doors.

WRT trustafarians - I hadn't really considered that, and hope not to have to eat my hat in the near future.