On Friday, in its January Employment Situation Summary, the government's Bureau of Labor Statistics served up a stack of lemons disguised as lemonade. President Barack Obama declared in a tweet that "We've recovered from the worst economic crisis since the 1930s," and the press dutifully fell in line.
The BLS reported that the economy added seasonally adjusted 151,000 payroll jobs and that the unemployment rate fell to 4.9 percent. As has been their habit for years, business reporters failed to label either key data point as "seasonally adjusted," even though they routinely apply that label to most other government data in their dispatches on the economy. The business press almost never looks at the raw (i.e., not seasonally adjusted) figures in any area. If they had looked at last Friday's raw jobs data, they would have wondered how the BLS could possibly have reported such a large number of additional seasonally adjusted jobs.
January is a month when companies let go of a large number of Christmas season and other employees. In recent years, January job losses have typically been 2.5 million or more — sometimes much more.
January 2016 was one of those "much more" months, as the economy shed 2.989 million payroll jobs. That's the third-largest monthly job loss in the 77 years the BLS has reported monthly data — eclipsed only in January 2008 and January 2009.
As far as I can tell, this genuinely disturbing news, only slightly mitigated by the fact several figures from prior Januarys, if adjusted for workforce size, would come in at over 3 million, has gone totally unreported.
The government's seasonal conversion to 151,000 jobs is, to say the least, highly suspect, based on the conversions of far smaller January jobs losses during the preceding five years:
The table above shows two alternatives for interpreting the data:
- Converting previous years as January 2016 was converted, average seasonally adjusted job gains would have been 346,000 instead of 156,000.
- Converting January 2016 based on how the past five years were on average converted, this year's reported seasonally adjusted number would have shown 44,000 jobs lost.
Either way, January 2016's seasonally adjusted figure clearly doesn't reflect the ugly underlying reality. January 2016's actual job loss is a disturbing sign that the economy is slowing down significantly.
One of the outlets which failed to report the shocking raw loss of jobs was, as would be expected, the Associated Press.
In reporter Josh Boak's defense, quite a few people were fooled, or pretended to be:
"The January report is a solid report in disguise," said Douglas Holtz-Eakin, a former director of the Congressional Budget Office and president of the conservative American Action Forum.
That seems especially true given that the job gains come at a delicate time for the U.S. economy. Analysts have warned that the economy faces a growing risk of another recession within a year or two after having recovered only gradually from the Great Recession. Economic weakness overseas and reeling financial markets have slowed growth and squeezed manufacturers.
Last month's pay raises and the addition of retail jobs suggest that consumers have the resilience to bolster growth.
Most readers here will be totally unsurprised that the American Action Forum, which Boak described as "conservative," calls itself "center-right." Most readers here will also be unsurprised that this supposedly "center-right" group, among other things, opposes Senator Jeff Sessions's common-sense immigration initiatives, and believes that Middle East refugee immigration won't hurt the U.S. job market. At best, AAF is in the squishy middle.
"Last month's pay raises" follow a flat December, when there was no change in average hourly pay. An increase of 12 cents an hour in 2 months is far from impressive. Though the raw data appears not to be available in this area, it's more than a little possible that January's raw job losses mostly hit people who weren't particularly well-paid, leaving those who still had jobs to bring up the average, and that the seasonal adjustments didn't fully remove this impact.
Boak also claimed that "January's hiring, though modest, followed robust job growth of a seasonally adjusted 280,000 in November and 262,000 in December."
The raw numbers make a mockery of this "conventional wisdom" assertion:
As seen above, the net three-month payroll employment change of 2.566 million jobs lost barely escaped being the worst performance in the past five years. The only reason it looks all right on a seasonally adjusted basis is because, as seen earlier in this post, the January 2016 figure of +151,000 is so bogus.
As to the presidential tweet referenced earlier, readers need to be reminded that in 2009, when Congress passed the stimulus bill, we were told that its passage would enable the unemployment rate to drop to 5 percent by the third quarter of 2013.
Additionally, there's little doubt that several million people who should be considered truly unemployed are instead not considered part of the workforce, making the reported unemployment rate not credible.
Cross-posted at BizzyBlog.com.
