On Friday, the government's Bureau of Labor Statistics reported that the economy created 175,000 seasonally adjusted jobs in February, with 162,000 of the additions occurring in the private sector.
That result exceeded expectations of roughly 150,000, and caused the business press to sing odes of high praise to an economy that was amazingly overcoming this year's difficult winter weather. Unfortunately, as readers will see after the jump, February's raw results demonstrate that it was all an illusion.
Shortly after I began blogging in 2005, I began to grasp the importance of looking at the underlying data before the government and others who generate economic reports "seasonalize" them. Too bad the business press never has.
The BLS says that "Seasonal adjustment is a statistical technique that attempts to measure and remove the influences of predictable seasonal patterns to reveal how employment and unemployment change from month to month."
Much of the problem lies in the word "predictable."
Since the economy began to go into a tailspin in the late spring of 2008 (Note: that's several months after the poobahs at the National Bureau of Economic Research say that the recession began), the only thing that has been predictable about the jobs data is that it has been unpredictable.
The sharp recession significantly disrupted normal seasonal patterns. Reported data has continued to be erratic in the nearly five years since it officially ended. This makes sole reliance on seasonally adjusted data even more dangerous than usual, as the following graphic of actual and seasonally adjusted job gains and losses shows:
For all nonfarm payrolls, the government's best estimate of February's actual results (i.e., 750,000, the not seasonalized number) came in 288,000 jobs below February 2013, 204,000 below February 2012, and represented the worst February result since 2010, the final month of the economy's "jobs recession." Based on comparisons to the two most recent years (the BLS goes back five years in its seasonalizing, but says that it weights recent years more heavily), one would have expected a seasonally adjusted result of about 34,000 jobs added. Instead, at 175,000, it was 141,000 jobs greater.
Similarly for private payrolls, the government's raw number of 300,000 came in 253,000 jobs below February 2013, 193,000 below February 2012, and was also the worst February result since 2010. Again based on comparisons to 2013 and 2012, one would have expected a seasonally adjusted result of 40,000 jobs added (corrected from zero when originally posted). Instead, it was 162,000.
These are the largest upward or downward differences I recall seeing in comparing raw and seasonally adjusted data in the employment report during the past seven years.
How different everyone's mood would be right now if the seasonally adjusted results had come in at a level which genuinely reflects what actually happened.
These seasonally adjusted results which are so obviously at odds with reality sure came at an awfully convenient time for the Obama administration. The press, which has almost never seen fit to look at the raw data, dutifully went over the top in praising the results (bolds are mine):
Reuters, via Lucia Mutikani
U.S. job growth offers upbeat sign for weather-beaten economy
U.S. job growth accelerated sharply in February despite the icy weather that gripped much of the nation, easing fears of an abrupt economic slowdown and keeping the Federal Reserve on track to continue reducing its monetary stimulus.
... "It reinforces the case for the economy being stronger than it's looked for the last couple of months," said Bill Cheney, chief economist at John Hancock Financial Services in Boston. "It makes life easier for the Fed and feeds into continuing the tapering process."
Bloomberg, via Jeanna Smialek
Employers added more workers than projected in February, indicating the U.S. economy is starting to shake off the effects of the severe winter weather that slowed growth at the start of 2014.
... The pickup following the weakest two-month hiring gain in more than a year shows employers remain confident the economic expansion will recover after winter storms slowed consumer spending.
... “The U.S. is on a stable trajectory for moderate growth,” said Julia Coronado, chief economist for North America at BNP Paribas in New York.
Associated Press, via Christopher Rugaber
US EMPLOYERS ADD 175K JOBS DESPITE HARSH WEATHER
U.S. employers stepped up hiring in February despite a blast of harsh winter weather, renewing hopes that the economy could accelerate this year.
... Friday's figures from the Labor Department were a welcome surprise after recent reports showed that harsh weather had closed factories, lowered auto sales and slowed home sales. Along with an increase in wages last month, the report suggests that some employers are confident that consumer spending will pick up in coming months.
"If the economy managed to generate 175,000 new jobs in a month when the weather was so severe, once the weather returns to seasonal norms ... employment growth is likely to accelerate further," Paul Dales, an economist at Capital Economics, said in a note to clients.
Unfortunately, economists Cheney, Coronado, and Dales are offering the public and their clients assessments which don't reflect what really happened in February.
As I wrote in a column elsewhere on Saturday:
Do not be fooled. Especially be wary of “the prevailing view of economists,” which is that “when the weather warms up, so, too, will the U.S. economy.”
All of the “surprise” in Friday’s employment report lies in how BLS could possibly have converted truly awful raw data into decent-looking numbers ... We had better hope that the raw numbers came in as weak as they did due to February’s rough weather. If they didn’t, we may be facing the worst year in the job market since 2010.
Cross-posted at BizzyBlog.com.