Friday, February 3, 2012

Nonfarm Payroll Growth During Economic Expansions (Musical Tribute)

Click to enlarge.


Source Data:
St. Louis Fed: Total Nonfarm Payrolls
NBER: US Business Cycle Expansions and Contractions


Mr Slippery said...

I wonder why Calculated Risk doesn't see it this way. Every media outlet was screaming about how the recovery was real and blue skies had returned.

Another 16 months of job growth like January should get us close to where we started in 2007 for total number of jobs.

fried said...

I am somewhat confused as well. The whole commentariat at HCN was very chipper this morning. So was Wall Street. Oh well.

Jazzbumpa said...

Mark -

I'm not sure what you are graphing here. I looked at your sources, and it's not obvious how you brought this together.

Could you describe how you arrived at your data points?


TJandTheBear said...


Now go to your room before anyone sees you!!!


Stagflationary Mark said...

I think I can clear up some of this confusion.

Everyone loves an improving short-term trend. The employment situation relative to previous months was pretty good. Get out the confetti!

I don't see how anyone can love the declining long-term trend (as seen in this chart) though.

The long run is a misleading guide to current affairs. In the long run we are all dead. - John Maynard Keynes

See? Nothing to worry about. Keynes said so himself. ;)

Stagflationary Mark said...


An example is probably worth a thousand words so let me describe the 2nd to last data point on the chart (November 2001).

As seen in the NBER link, November 2001 was the start of the expansion. It went until December 2007. That's 6.08 years (73 months).

There were 130,898,000 nonfarm payroll jobs at the start. There were 137,982,000 nonfarm payroll jobs at the end.

That's 5.4% more jobs at the end than there were at the beginning. Since it took 6.08 years, that's an average annual growth rate of 0.87% per year (during the economic expansion).

(137,982,000 / 130,898,000)^(1/6.08) = 1.0087

Here's something else worth considering. There are two data points in the chart that are worthy of additional notes.

1. The first expansion officially started in June 1938. The payroll data didn't start until January 1939 though. The data point is therefore a reasonably close estimation using just the data that I did have.

2. The last expansion isn't over yet (at least in theory). It is therefore also an estimation using only the data we have so far.

Some would have us believe that this latest expansion will be permanent of course. However, they might want to consider this.

We are 2.58 years into this expansion.

3 of the 12 previous expansions lasted 3 years or less.

6 of the 12 previous expansions lasted 4 years or less.

Stagflationary Mark said...

tj and the bear,

There are no bad charts, just bad chart owners.

I'm bad. ;)

Stagflationary Mark said...

Bonus thought.

Perhaps I should do a chart of job growth during economic contractions at some point.

Nah, I don't want to be a party pooper.

It is just a silly theory anyway. Just because the economy has had contractions in the past doesn't mean that the government hasn't miraculously managed to remove them permanently.

Stagflationary Mark said...

Yet another bonus thought.

Should the red trend line in this chart be an accurate indication of what is to come in future years (possible but by no means guaranteed), then it is not all that hard to imagine why the 30-year TIPS yields just 0.75% today. In my opinion...

Low Employment Growth -> Low Growth -> Low Real Yields

It also works in reverse. The better employment report did drive up real yields a bit today.

That said, over the long-term I consider this short-term news to be similar to what a gnat hitting the windshield does to the forward momentum of my car. Just an opinion of course.

Jazzbumpa said...

Mark -

I think I've got it.

You're plotting at the beginning of the recovery, the average jog growth during the recovery.


Your chart has to relate in some way to this one, I would suspect.


Troy said...

I just posted at Delong's, thought I'd come and post it here.

Somebody else created, which I thought was interesting too.

Then there's this one I also made: . .

The baby boom is aged ~50 to 65. Generation Y (echo boomers) is aged ~17 to ~32.

Gen Y has to get a job AND pay the boomers' retirement. AND interest on a $20T national debt. And $100+ oil. AND $2T+/yr on health care.


dearieme said...

One of the early economics books I read had a chapter called something like "And will the trees grow to the sky?" This was an attempt to persuade the reader not to make endless extrapolations of exponential growth.

Stagflationary Mark said...


Your chart has to relate in some way to this one, I would suspect.

There's just no getting the genie back into the lamp. Sigh.

Stagflationary Mark said...


Very discouraging charts you've got there. Blergh is right.

Stagflationary Mark said...


From the book Random Walk Down Wall Street...

Caveat 1: Expectations about the future cannot be proven in the present. Remember, not even Jeane Dixon can accurately predict the future. Yet some people have absolute faith in security analysts' estimates of the long-term growth prospects of a company and the duration of that growth.

Predicting future earnings and dividends is a most hazardous occupation. It requires not only the knowledge and skill of an economist, but also the acumen of a psychologist.

The knowledge and skill of an economist? Why do I feel the need to heckle?

Jazzbumpa said...


Gen Y has to get a job AND pay the boomers' retirement.

Actually no. SS was designed as a pay as you go program. Back in the 80's it was recognized that the boomer gen would draw down too much, so the system was modified and the SS trust fund was built up.

I paid into the SS fund for over 40 years. I paid to build up that fund so that I could draw on it now. Nobody in their 20's is subsidizing me. I receive a pension - deferred salary from my long time employer, and SS - an annuity I purchased over decades, on the installment plan.

I don't have the data at my finger tips, but I don't believe there has ever been a year in which SS payouts exceeded premiums received.

Consider Steve Roth's post at AB.