Inside the March Unemployment Report

The March unemployment report was a bit of a disappointment, and shows the job market skipping like a stone across the bottom. Since October we have been popping just above or just below new job creation. 

The report claims 162k new jobs, seasonally adjusted. Without the seasonal adjustment, only 136k new jobs were created. Without 81k statistically-generated jobs from the birth/death model, only 55k jobs were actually counted. Without the 48k census jobs, only 7k jobs were created. Or so it could be argued. 

More hopeful is the alternative method of counting, the BLS Household Survey, which says over a million jobs were added by small business in Q1. That survey is often volatile, and is considered less reliable, but it seems to be pointing to an underlying strength belied by all the negativity surrounding the main survey. As I have said, every W starts as a V, and we may have upside surprise in employment over the next quarter. 

GS thought BLS would report 275k jobs (revised down to 200k the night before), and TrimTabs, which seems to be able to track better than the BLS, expected 280k. The "consensus" had been 184k. Much of the shortfall was due to fewer census workers (48k) than expected (100k). It may be the census hires have been grossly undercounted (irony abounds). Most likely the 162k gets adjusted upwards later. Again, I expect a positive surprise in the next few months. 

The rate stayed at 9.7%, although it would have been 9.8% without the census workers, and a good argument can be made that they should have been skipped from the seasonally-adjusted number since they go away in a few months. (The BLS subtracts them from the non-seasonally adjusted number before applying the seasonal adjustment, then adds them back.) Inside the 9.7% number, rounding masks an increase of 0.5% (9.69% to 9.75%). 

We need around 200k new jobs to employ all the net new workers coming in (eg college grads). Lower than that and the rate still falls, or should fall, except that the BLS has been dropping the total workforce fairly steeply through this period. I discussed this in some detail in Feb. 

The way to avoid the noise is to look at the totals not the rates. We have lost 2.3M jobs in the past 12 months, and 8.2M since the start of the Great Recession. It had been 8.4M. We are skimming the bottom (see chart from Calculated Risk), and the critical question is whether we bend up rapidly as in prior recessions or stay down in the U shaped bottom for an extended period. Ominously, the most recent recession in 2001 and the prior one in 1991 both are the U shaped curves in this compendium chart, and suggest a similar valley of death awaits us this time:

Tragically, the stats underneath the numbers are not promising. An astounding 44% of unemployed are long-term, and the number of these chronic unemployed (meaning unemployed for more than 26 weeks, or half a year) is at record levels (6.5M). The number of part-time workers (9.1M) is also at record levels, and after a few good months has has gone back up from 8.3M in Jan and 8.8M in Feb to the horrendous level of last Oct (9.2M).

Weather effects in Feb may account for about 100k of the total. Weather adjustment plus census workers add to 148k, very close to the total. Temp help rose 40k, healthcare 27k, state & local govt held constant; with the census (48k), a major part of the increase was of the wrong sort of workers (115k) instead of productive private sector positions (41k), such as manufacturing (17k). If you expect the private sector temp positions to roll into permanent, the report suggests 123k growth in private sector jobs, albeit 81k of them are statistical, meaning created by the BLS's questionable birth/death model for small businesses. 

Analyst reactions were mixed. 

Looking forward, the blogosphere reports that job postings are increasing, but a new survey by the BLS says the opposite: the number of unemployed per opening has trended up to 5.5, and the total number of job openings has dropped a bit. This chart is a bit complex, but compares the unemployment rate to the job opening rate. As you can see, it ticked down a bit in Feb, but this might be due to weather effects. Hence the blogosphere may be a better leading indicator than the backwards-looking BLS survey.

 

As we bounce off the bottom, we might see several months of a real uptick, partly driven by census workers and also driven by good private sector hiring. Then the Summer of Disappointment could hit: census jobs go away, and a huge wave of state & local layoffs are expected to commence this summer. Whether the net increase of real private sector jobs will exceed the drag of govt jobs is unclear. 

The Onion had the last word: it reported that the Dept of Labor had created one special new job for $40B of stimulus. The position would be based in the US Virgin Islands and be paid $500k salary plus multi-million bonuses to study multiple orgasms in a hot-tub environment. (Kudos to Mish for finding the cartoon above and this Onion story.) 

Comments

[Mamma Boom Boom]

Fed-speak changed slightly "Pledge of low rates for 'extended period' contingent on economy, not calendar"

Forked tongue????

[yelnick]

Mamma, Fed did not raise discount rate yesterday, which is what I expected since it would have been too fast a path to 100 bp higher than Fed Funds, the marker of raising both in tandem. Short term rates have ticked up a bit, which is probably the better indicator of when the Fed is going to raise.

I thin the Fed Speak today is an indication of higher confidence some sort of recovery is in the works, giving the Fed the headroom to raise.

The next shoe to drop is the Q1 GDP rate, where expectations still hover around 4% +/-, which is below Q4. I expect a rate at the high end of the range, which should creep higher as we get closer to the pronouncement.

[Mamma Boom Boom]

>I thin the Fed Speak today is an indication of higher confidence some sort of recovery is in the works, giving the Fed the headroom to raise.<

Oh, I agree. But, also, I read where the change in fed-speak will be a secret message to some of the big players to start to pull out of their carry trades. And there it is.

[JT]

So many conspiracy theories... so little time.

[Yury]

Hi everyone, made a new EW analysis, ES min. extension from now over 1295 until Aug/Oct, http://tinyurl.com/yfwkmvu

[DG]

We'll know by April 20th if this is right or not. The price target is only a minimum. If I had to guess at a maximum, I'd say 1.618X wave-.E, which would be at 121.84. Fitted into the larger context of the wave count from the March 2009 low, this would potentially be a top, but it could still be only part of a larger structure. There's just no way to know at this point.

http://yfrog.com/6zspydailyapril2p