Today’s employment report calls for posting Aerosmith’s “Same Old Song and Dance”:
**** ORIGINAL POST ****
In light of conflicting data about how well or poorly the economy has been doing during the past month or so and the stock market’s down year thus far, today’s employment report is getting even more attention than usual. It’s too bad, as will be seen later, that almost no one will focus on what really happened in January.
- On Wednesday’s ADP conference call, Mark Zandi of Moody’s predicted 170,000 seasonally adjusted job additions and the unemployment rate dipping from 6.7% to 6.6%.
- Bloomberg has 180,000 jobs and the unemployment rate staying the same.
- Christopher Rugaber at the Associated Press has 170,000 jobs adds and the unemployment rate staying the same. Rugaber seems to be getting excuses ready if he needs them, which is why I’ve saved it for future reference if needed.
- This Reuters report has 185,000 jobs added with no unemployment rate prediction.
Raw (not seasonally adjusted) data benchmarks
This is a particularly important report for the raw data (i.e., the government’s estimate of what really happened) as opposed to the seasonally adjusted information almost everyone will report, as seen in the following graphic:
The question about January isn’t how many jobs were added. It’s how many net workers lost their jobs.
Based on the track record since the turn of the century, I’m going to say that the economy isn’t meaningfully improving unless January job losses overall were limited to 2.53 million or fewer, and private-sector jobs losses were 2.05 million or fewer. I’m using those benchmarks because they’re 100,000 or so below January 2012. If the economy can’t do significantly better than that, it’s probably going to continue to be a very long time before we see a legitimate jobs recovery. I also considered the fact that December Christmas season hiring was lower than usual, so the post-season job losses should also be lower than usual.
Many other items bear watching, including but not limited to the labor force participation rate, the part-time worker stats, and what industries see outsized or lower than expected job declines on the ground. It will also be interesting to see the size of the revisions to November and December — especially December, since last month’s result was called bogus and too low by a lot of the “experts.”
We’ll find out what happened here at 8:30.
Last-minute note: For those who think I’m setting today’s bars too high — “Getting headline unemployment down to 6 percent, while employing those folks at the margins of the labor market, would require about 365,000 jobs each month for three years — about double the pace accomplished since the economic recovery began and during the George W. Bush expansion.”
HERE IT IS (permanent link to full HTML) The snap judgment is that it’s the same old song and dance —
Total nonfarm payroll employment rose by 113,000 in January, and the unemployment rate was little changed at 6.6 percent, the U.S. Bureau of Labor Statistics reported today. Employment grew in construction, manufacturing, wholesale trade, and mining.
Household Survey Data
Both the number of unemployed persons, at 10.2 million, and the unemployment rate, at 6.6 percent, changed little in January. Since October, the jobless rate has decreased by 0.6 percentage point.
… After accounting for the annual adjustment to the population controls, the civilian labor force rose by 499,000 in January, and the labor force participation rate edged up to 63.0 percent. Total employment, as measured by the household survey, increased by 616,000 over the month, and the employment-population ratio increased by 0.2 percentage point to 58.8 percent.
… Establishment Survey Data
Total nonfarm payroll employment increased by 113,000 in January. In 2013, employment growth averaged 194,000 per month. In January, job gains occurred in construction, manufacturing, wholesale trade, and mining.
… The change in total nonfarm payroll employment for November was revised from +241,000 to +274,000, and the change for December was revised from +74,000 to +75,000. With these revisions, employment gains in November and December were 34,000 higher than previously reported. Monthly revisions result from additional reports received from businesses since the last published estimates and the monthly recalculation of seasonal factors. The annual benchmark process also contributed to the revisions in this news release.
Mark Zandi, who was completely convinced that December’s original 74K was bogus and way too low, must be gritting his teeth today. There’s another revision to December coming in February, so he could still end up being right.
Now let’s look at the revised raw and seasonally adjusted results:
- The total nonfarm payroll job loss was 2.870 million — the worst since 2009 (barely, but still worse), and 340,000 below my pre-release benchmark (yes, I’m looking at the numbers after the population control adjustments.
- The private-sector job loss was a nearly as awful 2.346 million — the worst since 2009, and a sliver less than 300,000 below my pre-release benchmark.
This is a very, very weak report. More detail is forthcoming.
UPDATE 1: Chris Rugaber at AP is already furiously spinning (report will be revised again, but I also saved this iteration) —
Hiring was surprisingly weak in January for the second straight month, likely renewing concern that the U.S. economy might be slowing after a strong finish last year.
Chris, December had 75,000 job adds. That’s not “a strong finish last year.”
UPDATE 2: The Household Survey says that seasonally adjusted January employment increased by 638,000. That seems to be a long-overdue catch-up to the Establishment Survey trend of the past year. UPDATE 2A: The breakdown, which doesn’t add to the total because of independent seasonal adjustment calculations, is 378K full-time and 168K part-time.
UPDATE 3: The civilian labor force increased by a seasonally adjusted 523,000. But even with that increase, the labor force is 239,000 lower than January 2013, and 13,000 below September 2013.
UPDATE 4: The unemployment rate for African-American men went from 11.5% to 12.0%.
UPDATE 5: Seasonally adjusted temp employment grew by 8,100. December’s temp adds were revised to 30,100 from 40,400. 40% of December’s 75K in seasonally adjusted net job adds were temps.
UPDATE 6: Average weekly hours in retail dropped to 31.1, which I believe may be an historic low.
UPDATE 7: Commenter Scott tipped me to this paragraph in today’s report —
This revision incorporates the reclassification of jobs in the QCEW. Private household employment is out of scope for the establishment survey. The QCEW reclassified some private household employment into an industry that is in scope for the establishment survey–services for the elderly and persons with disabilities. This reclassification accounted for an increase of 466,000 jobs in the establishment survey. This increase of 466,000 associated with reclassification was offset by survey error of -119,000 for a total net benchmark revision of +347,000 on a not seasonally adjusted basis. Historical time series have been reconstructed to incorporate these revisions.
The cynical response would be, “Well, if your economy can’t create new jobs, you invent them.” But at least at first glance, that seems unfair.
Assuming the people involved do work and get paid for it, they probably should be considered employed. If someone has a counter-argument for that, I’d like to hear it.
Earlier text in the related section of the report indicates that these “new” employees were spread across a period of either 21 months or the past 5 years (it’s hard to tell), and not just dumped into a single month. This does increase reported Obama-era employment in comparison to previous administrations, because the adjustments don’t go back to any previous administration. If (emphasis if) the people in this category were being detected but not reported all along, I would question why that wasn’t done.
Note that these jobs were added into the Establishment Survey, where this month showed 113K seasonally adjusted job adds, but not into the Household Survey, which showed an increase of over 500K in total employment.
UPDATE 8: Things might have been worse but for the fact that January’s Birth-Death adjustment of -307,000 was less than January 2013′s -314k and January 2012′s -367K — which is odd, given that the job losses which BLS could estimate from its survey were clearly higher this time around (specifically: 2012 without birth/death was -2.225 million; 2013 was -2.550 million; 2014 was -2.563 million)