It seems there is some controversy over the NFP report this morning, which I reported on earlier.

Among the howlers are Barry Ritholtz, who is basically calling Santelli a liar, and Time's "Swampland" blog which claims that the report is in fact all good news.  Both predicated their analysis on the benchmark revisions that came with the 2010 Census data, which resulted in adjustments that are "one time" changes and are not reflected back into the previous month's household data.

Incidentally, this is why I use the NSA numbers -- BLS doesn't change them later.  If you've followed my analysis on this report for any length of time you'll remember that I've noted this repeatedly -- the only way to get an honest view of what's up is to not have a series that changes out from under you constantly, otherwise you wind up having to revisit everything that is changed and, well, change your mind.  Of course a month or three or six later nobody cares that the counts were wrong.

This very sort of distortion is what makes the weekly claims data very difficult to use with any sort of authority -- it's literally all over the place with the revisions and often they're quite material.  That is why I use the "big table" at the bottom of that report for my analysis -- it's two weeks late, but it also has the revisions in it so what I report on there doesn't change.

The screech owls all claim that the revisions are responsible for everything and thus the report is strong. 

They're incorrect.

It is true that the 1.685 million people that were claimed to be added to the population is about 10x more than the typical monthly increase in population, and is obviously the Census adjustment (even though the BLS didn't break it out for the non-seasonally-adjusted figures.)  But what's not established is how the NILF adjustment was determined, given that the Census data is more than a year old.  Allegedly NILF is a count off the survey (the population is obviously not as it comes from the Census; employed is.)

So one can try to explain away the exit from the workforce figures with simple ratio adjustment but then we have a problem if we do -- the participation rate dropped in a big way, from 58.5% to 57.8%.  That's two tenths off the low from last year (January 2011) and more-importantly, basically a flat-line since the recession's depths.  There has been no material improvement in this metric since the end of 2009 when it hit the floor.  Yes, it's rattled around, but it's not getting better, and it is this number that matters to fiscal sustainability, as only employed people pay taxes.

As I noted the employed number went down from 140.681 million to 139.994, a drop of 737,000 on the month.  This, into an adjusted (higher) population.  Yes, I know, this figure is non-seasonally adjusted and people are usually (always, for all intents and purposes) fired in January.  Nonetheless it is what it is, and it, along with the population change, was responsible for the drop in the participation rate.

Note that irrespective of when the population change was recorded (all at once or across the last year) the end point for the participation rate remains the same; in fact the inclusion of the population adjustment this month means we've been overstating the participation rate for the last year.  Six of one, a half-dozen of the other; the end point is the same.

Finally, if you look at the not-in-labor-force as a percentage of the total population, you discern the problem for the broader public, as that tells you who can work but isn't even bothering to try to find a job.  It's on a tear, showing that indeed, people are exiting the workforce, and the claim of those who say this did not happen last month are factually wrong. 

Again, if the population adjustment was "late" then this means we've been understating this value for the last year (lower is better from the standpoint of the workforce), so it's been worse than reported.  But the important aspect of this chart is the slope, not the absolute value.

I had to go over to FRED to get the data series back before 1991; they have it back to 1976.  Note carefully that the upward rise began in 2009 with people exiting the labor force as a percentage of the population, that we are back to where we were in the late 70s and early 80s as women entered the workforce en-masse, and that absolutely nothing about the revisions or last month's data has done a thing to change the slope of that data when one smooths the normal seasonal fluctuations.

In other words this report sucked both from the standpoint of government fiscal sustainability and from the standpoint of the labor force (that's you and I, folks) itself.

I stand with my originally-reported view -- this is not a strong report.