One of the more astonishing things I’ve come across recently was an utterly disingenuous article in Sunday’s Times: Navigating the Fog in Jobs Data.
The article is a discussion with (former Prudential) Strategist Dr. Ed Yardeni. It seems that he has been consistently over-estimating the BLS new job creation each month. As a group, Economists have been consistently over-estimating job creation this entire post-recession cycle; taking the "Under" on NFP has been the winning bet over the past few years. But the conclusion Yardeni reaches — which is not only repeated by the paper, but practically endorsed — is that since this cannot possibly be the case, something must therefore be wrong with the BLS Data.
I practically choked when I read the short column. The author opens with:
"When the numbers stop making sense, maybe it is time to stop believing them.
The Bureau of Labor Statistics reported far fewer net new jobs in May and June than Wall Street expected. If it happens again when July data are issued on Friday, Edward E. Yardeni, chief investment strategist at Oak Associates, will assume that there is something wrong with the report, not the economy. (emphasis added)
That’s denial, plain and simple. Let’s continue:
A Bloomberg News poll of economists estimates that 145,000 jobs were added in July. Mr. Yardeni, who said that he had overestimated the figures announced in the two previous reports, is aiming high again.
“I’m going to take a chance on making a third mistake and go with 180,000 to 200,000,” he said.
Most evidence aside from the government report suggests that jobs are plentiful, Mr. Yardeni said in explaining his prediction. If fewer people are getting hired than economists expect, he is inclined to attribute it to too few prospective workers, not too little work. (emphasis added)
Its not my job to factcheck/proofread the NYT business section — but isn’t somebody‘s
job? There’s so many things wrong with that statement I do not know where to begin. And as a pundit, I’m as wrong as anyone, so I prefer to point out flawed methodologies and/or statistical errors; I really don’t like to point
fingers and say "Hey! This guy/gal is wrong." People who live in glass houses,
and all that.
Let’s review: Why might we think these numbers make so little sense to begin with? We know (at least those of us who have been paying attention) that the economy has been unusually Real Estate dependent this cycle. We have seen a vastly disproportionate amount of new job creation come from the Real Estate complex (including agentsm, mortgage brokers, etc.) We also know that the Housing market has cooled dramatically — those of us paying attention also saw the beginnings of this as long ago as August 2005.
Further, we know that by just about every historical measure, this has been the weakest job creating recovery cycle in the post WWII era. We have previously noted that the unemployment rate is down due to NiLFs: Not In Labor Force. Barron’s described it as "the incredibly
shrinking labor force, a phenomenon that’s largely responsible for the
deceptively modest unemployment rate." The labor participation rate touched is near 15-year lows.
Using other measures that include these slackers, unemployment is actually much higher than the reported 4.6%: The well respected Liscio report noted the actual number could be much higher of a jobless rate. If one includes the so-called
marginally attached workers and part-timers who really want to be
working full time, the unemployment rate weighs in at a formidable 9.4%.
Jobs that we are creating by and large have been paying less and offering weaker benefits than the jobs they have replaced. (This is a factor in the negative savings rate, but let’s save that discussion for another time). One notable exception to the weaker pay has been the home construction industry. These jobs have been relatively high paying. The sector has been a major source of new job creation, from real estate agents to mortgage brokers to Loews and Home Depot.
It should come as no surprise that as the entire sector has cooled, so too has the jobs creation associated with it. We now have the biggest inventory of new and existing homes for sale we have seen in nearly a decade. Realtors have said that home sales are now a ‘buyer’s market’. New home builders are witnessing the rise of a Ghost Housing Market, where specualtors walk away from their down payments, rather than close on a property which has declined significantly in price.
Let’s look at what Mr. Market is saying about the sector: In the face of slowing sales and building inventory, the homebuilders have gotten shellacked. The chart shows they gave up nearly 50% of their value over the past year:
ISE Homebuilder Index, May 2005 – July 2006
The strongest engine of Job creation (Housing) was created by ultra low interest
rates; As mortgage rates have risen back towards their historical
average, that engine has gone into an expected cyclical decline. The market correctly anticipated this, as the chart above reveals.
Although the numbers disagree with Yardeni’s forecasts, the conclusion may not be that the numbers make no sense; perhaps it simply means that Yardeni’s forecast — like those of so many other economists — is failing to comport with economic reality:
“If I’m wrong, I’m going to start to wonder if we’ve run out of able-bodied people to hire rather than worry about whether the economy is slowing,” he said. “In this economy, we need more skilled workers and knowledge workers, and they’re getting harder to find.” This is true “even in China and India,” he added. “It’s not just a U.S. phenomenon.”
This confuses two different issues: The lack of highly qualified technical workers with the low new job creation. I assume no one is suggesting that we are short 9.3 million qualified technical workers — thats the shortfall Liscio report noted.
Mr. Yardeni supported his case with several pieces of evidence, including the unemployment rate, which has remained at 4.6 percent despite the tepid announced job creation, and responses to confidence surveys indicating that consumers consider jobs to be plentiful.
A weaker-than-expected number of new jobs on Friday may stir concern about economic sluggishness and provide a lift for defensive sectors of the stock market such as health care, utilities, financial services and consumer staples, Mr. Yardeni said. Shares in industries where earnings are tied more closely to economic swings may rise if the job creation figure is surprisingly high.
Either way, he encouraged investors not to read too much into the jobs report, which has often required subsequent revision. Whatever it says on Friday may not be the last word.
“If the number is weaker than expected, pay no attention to it,” he said. “It’s probably wrong and likely to be revised up. Revisions tend to rise in an economic expansion.”
This is more than cognitive bias — this is denial, plain and simple.
I have been a critic of various data assembly methods, hedonics, birth
death adjustments, and many other numerical sleights of hand. I have
rationally explained the differences between the Household and
Establishment surveys. All it takes is a little elbow grease and cognitive functions.
I suggest others do the same.
NOTE: For a markedly different approach, compare the above story with a column in the Times a few months ago by Dan Gross: When Sweet Statistics Clash With a Sour Mood. Instead of rationalizing, there is a genuine attempt to make sense of the statistics in context of what we know — rather than merely hope — to be true.
Navigating the Fog in Jobs Data
CONRAD DE AENLLE
NYTimes, July 30, 2006