Wednesday, February 2, 2011

Estimating Friday's jobs number

I have always hated the fact that the monthly jobs number ends up being so important. Whatever the number is that is announced this Friday by the BLS is almost sure to be revised significantly, up or down, in the months and years to come, as BLS eventually reconciles its estimates to actual data based on IRS tax receipts. Sometimes the revisions can be of an order of magnitude in size.

It's also the case that the ADP monthly estimate of the number of private sector jobs gained or loss has lost a lot of credibility in the past year or so, since it has ended up grossly underestimating the number of jobs that the BLS has reported. You can see this in the chart above, where the red line has been consistently higher than the blue line for most of the past year.

So are both these numbers so flaky that it is not worth paying attention to them? Only time will tell, but in any event you have to take them with several grains of salt at least. Both can be wrong.

Nevertheless, it is tempting to get a little excited about the relatively large January jobs gain that is being projected by ADP in their release today (+187K), since estimates for the BLS number are only +140K, and 200K jobs per month is an important threshold number since that is what is required—if it persists for a sufficient period—to bring down the unemployment rate. ADP overestimated job gains last month by a lot, but that was the only time they had overestimated the BLS number in over a year. What if ADP once again is underestimating the BLS number? If they are, then Friday's payroll number could be a welcome surprise to the upside.

I wouldn't want to put much stock in these speculations, however. As anyone knows who has followed at this data over the years, trying to predict monthly jobs numbers is fraught with difficulties, and the value of the exercise, even if one is right, can be negated months later by revisions to the original data. I only mention this because it is tempting, and because there are other indicators which are pointing to a substantial pickup in jobs numbers going forward (e.g., declining unemployment claims, stronger-than-expected corporate profits, stronger leading indicators, strong growth in capex, strong growth in commodity prices, strong export growth, strong car sales).


Benjamin Cole said...

"Nevertheless, it is tempting to get a little excited about the relatively large January jobs gain that is being projected by ADP in their release today (+187K)"

I'll admit it, I am excited. Dang, I think we are full into a bull market and on the cusp of a boom economy.

This is the first leg.

What wonders, what inventions, what technologies will man bring forth in this next long global business upcycle?

The speed of innovation is breathtaking. Profits will be huge--as Scott Grannis has pointed out, profits are already huge, and the recovery has hardly started.

Boomtimes, baby, boomtimes.

Remember prosperity? It is making a comeback. Get used to it.

Lori said...


You are absolutely right. One never knows when and where technological improvements will come.

Colonoscopys are a thing of the past.

Jeff said...

I've noticed lately that your outlook for inflation has subsided...using phrases like "trend slowly higher".

This is in contrast with language such as "an over supply of dollars", "rising inflation expectations", "inflation alive and well", etc. over the last few months.

What is the biggest reason for the change in tone? Is my continued fear of significant inflation overblown?

Scott Grannis said...

Jeff: My inflation concerns may have been tempered a bit over the past two years, given that I have way overestimated the gains we have seen in the CPI and related indices. Apparently the lags are longer than I thought, and there is more inflation "inertia" than I had expected, but I have not abandoned the idea that inflation should be increasing. I am acknowledging that it won't increase as fast as I originally thought.

Benjamin Cole said...


The world has never had so much venture capital, and so many scientists and engineers at work. The Internet speeds technical data globally instantly.

I wish I could live several more decades. I think I have two or three left. Dang.

We may also see a revolution in our understanding of money. If, as I suspect, deflation is the next big monster, then we must be eternally vigilant against deflation.

That is going to require a monetary policy very unfamiliar to what has worked in the past--and a lot healthier, by reducing the belief in deficit spending, and increasing the belief in monetary stimulus, that does not create debts for our children.

QE may be permanent, as well as very low interest rates. Gluts of capital may be normal.

brodero said...

I have always thought that the payroll number should be announced
as a 3 month moving is
by far a more accurate indicator of what is going on as opposed to the one month number...oh well
i guess it is human nature to only
want the most current (inaccurate?)

Frozen in the North said...

Everything you say is absolutely correct, commodity prices continue to rise, oil at $92, Copper at $10,000/t, all correct and the signs that the labor market is finally turning the corner is very positive.

My question is why is the Feds still talking QE, if everything is so good? Maybe if the gov't ease off the gas pedal America would be better off!

Still, that's the problem, so many classical indicators are saying: America is in recovery, but the agencies that have non-public data are behaving as if America is still on the edge?

Scott Grannis said...

Frozen: the Fed is between a rock and a hard place. The data always come out with a lag, and unemployment is always the last thing to improve. If they were to pull the plug on stimulus before the unemployment rate starts declining they could be subject to huge criticism, however injustly. This is one of the reasons to be worried about inflation: the Fed is going to be slow to react.