President Bush is on record as detesting historical revisionism, but in the case of today's employment report, I don't think he'll have any objections.
The headline news today was that the Labor Department reported a net gain of 126,000 jobs in October. This was about double the consensus estimate, but given the standard error on the monthly numbers -- and the guesswork that goes into the consensus estimate, that in itself isn't particularly significant.
The real surprise is what the Bureau of Labor Statistics did to the reported payroll numbers for the previous two months. The August number, which was originally reported as -93,000 jobs, then revised last month to -41,000, was revised again to show a gain of 35,000 jobs.
Meanwhile, the September number, originally reported as +57,000 jobs, was more than doubled, to +125,000.
These are substantial changes. While I'm sure there have been monthly revisions larger than the net 128,000 job swing seen in August, I can't personally recall ever seeing one -- and I've been writing about economics for going on 12 years now.
In any case, these revisions converted what had appeared to be a net loss of 41,000 jobs in the third quarter into a net gain of 103,000. So it looks like a modest recovery in the labor market actually started in mid-summer, and has been gradually (very gradually) accelerating since then.
This, of course, would be good news for the administration, since it suggests the long job-killing recovery of 2002-2003 is finally drawing to a close. Modest job growth should create at least modest wage and salary growth, supporting consumer spending, which in turn should prevent a dangerously abrupt fall off in growth from the third quarter's blistering 7.2% rate.
But there is a mystery to the revisions that I can't, off the top of my head, explain. While employment rose in the third quarter -- at least according to the revised BLS numbers -- the aggregate number of hours worked didn't budge an inch. Among other things, this means that productivity growth remained an astoundingly high 8.1% in the third quarter, despite the supposed upturn in payrolls.
What's going on? My first guess was that employers used the new hires to take some of the strain off their existing employees, resulting in a reduction in the average hourly work week, but leaving aggregate hours unchanged. But no, average weekly hours also didn't budge during the third quarter, remaining stuck at 33.7 (seasonally adjusted.)
So it appears employers hired more workers, presumably paid them, but didn't get any work out of them. This is not, at least in my experience, normal capitalist behavior.
Now I have to confess that I didn't pay a whole lot of attention to revisions in the monthly numbers when I was a reporter. Usually, you got the latest number, reported it, and then moved on to the next story -- machine tool orders, or whatever. So I don't have a clear idea where these changes are coming from.
Reading the fine print in the BLS news release, however, I see the bureau uses a seasonal adjustment process that is constantly being updated on a rolling three-month basis -- meaning that changes in the normal seasonal patterns observed in any given month can also result in revisions to the previous two months.
Perhaps that's what happened here. We did have that big blackout in the Northeast in August, which may have led to a spike in layoffs, resulting in some distortions in the seasonal adjustment that are just now being ironed out. But that's only a guess.
Interestingly enough, Wall Street didn't react much one way or the other to today's jobs report -- even though the October numbers themselves were unambigously better, with payrolls, hours worked and the average work week all rising. In late afternoon trading, bond prices were off a bit but above their lows for the day, and stocks were mostly flat to down.
It appears the street has already discounted a modest improvement in the employment picture -- revisions or no revisions -- but is unsure whether it will be enough to sustain the kind of economic growth needed to support its rosy earnings expectations. If growth downshifts to a 3-4% rate, will the job-killing recovery return? And if it does, will growth continue to downshift? These are reasonable questions to ask, especially as we get closer to the end of 2003 and investors start looking to lock in their gains for the quarter and the year.
But the voters appear to have considerably lower expectations -- at least according to the various confidence surveys. This could work to Bush's advantage. If the administration can persuade the voters the corner has been turned, and that job growth will continue to improve, even if only gradually (the modern day version of pie in the sky, by and by) they may be willing to forgive his truly abysmal economic record to date.
On the other hand, there are still many reasons to believe this economic recovery will never match the Clinton fat years of the late 1990s, much less the Reagan boom of 1983-84.
Just to point to one of the most obvious: Both of those previous growth surges were accompanied by huge declines in the personal savings rate -- or, to put it a little differently, huge increases in the marginal propensity to consume. Given the mathmatics of GDP (personal consumption accounts for about two thirds of it) such shifts, if sustained, can have an enormous effect on growth.
With the personal savings rate now down to 3%, it's hard to see how the previous pattern can be repeated now -- although I probably should know better than to underestimate the willingness of the average American consumer to borrow and spend. Hell, for all we know, the savings rate could go negative. It all depends on the willingness of the Fed to accommodate such behavior, which in turn depends on the willingness of America's foreign creditors to accommodate the Fed.
There, of course, a finite limit to this kind of economics. Even if interest rates fell to zero, there would eventually come a point at which debt burdens simply couldn't be stretched any further. But -- and this is what matters politically -- we may not reach that point in this economic cycle.
Which means that next year, "it's the economy, stupid," may simply mean that the stupid candidate is likely to end up the winner.
So it appears employers hired more workers, presumably paid them, but didn't get any work out of them.
shhhh!!!
...I see the bureau uses a seasonal adjustment process that is constantly being updated on a rolling three-month basis
Heh. Sounds like what my old math prof used to call a Dynamic Fitting Factor.
When you want to do curve-fitting (i.e. go backwards from a graph to an originating formula) the [lazy] expedient thing would be to use a 'constant' or 'weighting factor' to make things work. It's the geometrical equivalent of a ball-peen hammer taken to misaligned pipe-fittings. With numbers that varied over time, you needed a more pliable 'Dynamic Fitting Factor.'
It's closely related to the lesser known mathematical phenomenon known as Pegging the Bullshit Meter.
JOB CREATION Every good socialist (or national one's) clarion call...
Hmmm. lesseee heah , now.... killbots wanted for iraq, et al, AND private security and other technicians for.... and prob neither with a union or benefits/retirement plan worth the unicorn horn written on....
JAmessons, if ya please,,, Erin, no Bra!!!
I can explain all the mysteries. All these executive branch books are being cooked like the Christmas goose, right down the line. You know the old joke about the accountant applying for a job. The CEO asks one question: "what is 2 + 2"? The accountant says "what do you want it to be?" "You're hired", says the CEO. And so it goes.
Quite the opposite of what's happening in Russia, where people work but don't get paid!
I won't be surprised if the mammoth defecit is totally ignored by the American public come election time.
After all most are happily ignoring their own debt, so why not ignore the country's too. The only way most folks seem to have been able to spend their way through the last few years has been by re-mortgaging their houses, amassing huge credit card debt and overdrafts. (totally unsubstantiated claim, but supported by much anectdotal info).
Same thing up in Canada.
An interesting link for y'all:
Cool War - Economic Sanctions as a Weapon of Mass Destruction
The article I read mentioned that 70,000 workers on strike in CA. Is it correct to assume that the strikers are not considered "unemployed", but that the scabs are counted in the "newly employed" column? If so, that alone could account for over half the gain.
"For example, the Ministry of Plenty's forecast had estimated the output of boots for the quarter at one-hundred-and-forty-five million pairs. The actual output was given as sixty-two millions. Winston, however, in rewriting the forecast, marked the figure down to fifty-seven millions, so as to allow for the usual claim that the quota had been overfulfilled. " - George Orwell, "1984," chapter 4
But there is a mystery to the revisions that I can't, off the top of my head, explain. While employment rose in the third quarter -- at least according to the revised BLS numbers -- the aggregate number of hours worked didn't budge an inch. Among other things, this means that productivity growth remained an astoundingly high 8.1% in the third quarter, despite the supposed upturn in payrolls.
What's going on? My first guess was that employers used the new hires to take some of the strain off their existing employees, resulting in a reduction in the average hourly work week, but leaving aggregate hours unchanged. But no, average weekly hours also didn't budge during the third quarter, remaining stuck at 33.7 (seasonally adjusted.)
I would really take the hours numbers in the productivity data with a huge grain of salt. There is a lot of guesstimating with this data outside of manufacturing. here is an excerpt from the BLS Handbook Chapter on Productivity:
For nonmanufacturing sectors, employment and average weekly hours are computed from the CES, CPS, and NIPA sources. Although CES data on average weekly hours refer only to nonsupervisory workers, it is assumed for the computation of hours that the length of the workweek in each nonmanufacturing industry is the same for all wage and salary workers.
For my take on the jobs number visit Economists for Dean:
http://econ4dean.typepad.com/econ4dean/2003/11/finally_jobs_pi.html
It all looks like Finagle Factors (scroll down) to me.
That high productivity has me creeped out. I'm very interested in understanding the cause of it. Any ideas anyone? Is the investment in IT over the years finally starting to pay off? What happens if it continues to grow? Does unemployment tend to rise with productivity? It looks to me that the system is wired in such a way that what should essentially be a good thing (less work required to maintain needed production levels) can turn out to be very bad for those of us who work for a living since some will no longer be needed. What factors are there that can offset this effect?
Thanks for the link Pedro
Josh, the tax cuts & deficit together form a huge Keynesian stimulous. Somebody (China?) is agreeing to loan the US government money at 5% to give away to various entities (Halliburton et al, Medicare, you name it.) Thus China (and others) have invested in the US productivity.
The issue (as described in such left-wing House Organs as The Economist) is that this is not a sustainable model for growth. Something, sometime, has got to give. And the longer we go without the event, the more painful the event will be.
It used to be that Republicans believed in a balanced budget. They took over congress in 1996 on exactly that platform. At some time in the last 5 years, the Republicans got taken over by Pod People.
I'd like to know how the 176K Challenger, Gray & Christmas recorded layoffs figure into this mix. I have no idea as to evaluate this and rectify the numbers, not having written about economics. Are they saying we really created 300K jobs and just absorbed these layoffs?
Color me confused.
The republicans used to be the party where libertarians fit in, not comfortably, but they did fit in. But in the last few years, they've been taken over by people who say stuff like this. The libertarian ideal is based on "objectivism". Statements like Hale's are in explicit rejection of objectivism "There is no precise way to quantify the impact of the Bush administration's tax policy on the third quarter growth rate." And "The ultimate verdict on the administration's policy is still many years away."
The last is a coded attempt to call the 90's boom a result of the Reagan tax cuts, despite the fact the 90's boom came after a series of tax hikes to balance the budget.
Anyway, here's the bogus nonsense in full:
here is no precise way to quantify the impact of the Bush administration's tax policy on the third quarter growth rate. The income contribution from the personal tax cuts was very modest. But there has been a major stock market rally since April which has bolstered household wealth and confidence in general. The administration's tax program played a role driving the bull market. The administration's program reduces capital gains and dividend tax rates to only 15 percent, or the lowest level for many decades. Before Reagan, dividend taxes exceeded 70 percent while capital gains taxes have fluctuated in a range of 25-40 percent since the Second World War. These tax changes should have a benign long-term effect on growth by creating more incentives for savings and investment. The change in dividend taxation could also enhance corporate governance by improving resource allocation within companies.
It is true that the administration could have provided more short-term stimulus by skewing its tax cuts towards low-income people and providing more revenue sharing for state governments. Both would spend every dollar they got as quickly as possible. But the administration was not focusing solely on short-term stimulus in either 2001 or 2003. It wanted to alter the incentive structure of the tax system in order to promote long-term growth. As with Reaganomics during the 1980s, the ultimate verdict on the administration's policy is still many years away.
I know Hale. He's a sleaze bag and a bit of a nut -- at least where the Clintons are concerned. I can remember him swearing up and down that the Vince Foster "scandal" was going to blow sky high.
Hale isn't quite as big a supply-side fruitcake as, say, Larry Kudlow or Jude Wanniski, but he's still not a particularly credible economist.
I'd like to know how the 176K Challenger, Gray & Christmas recorded layoffs figure into this mix. I have no idea as to evaluate this and rectify the numbers, not having written about economics. Are they saying we really created 300K jobs and just absorbed these layoffs?
Color me confused.
The payroll number is based on a survey and produces an estimate of "net employment". So if IBM has 1000 workers in Sept and laid off 100 workers and hired 150 in Oct, it would report 1050. The change is 50, but at the same time 100 workers were laid off.
Both numbers can be true. BLS is starting to produce numbers that show total job creation and job destruction of which the net employment change is a tiny fraction.
lerxst
an Economist for Dean
The economic data represents a sample, not actual numbers. Therefore, there is a margin of error. 100,000 jobs represents somewhat less than 1/1000 of all people working. Therefore, recent incremental changes from month to month could be nothing more than how you round the numbers. If Bushco will politicize the intelligence, they certainly would be willing to politicize the economic data.
Besides, the employment numbers are overrated and not reflective of the public mood. Nobody's going to vote for or against Bush because there are 100,000 more or less shitty jobs a month. The bloom will drop of the Bush Presidency when real estate values start to soften and my sources say that day may not be far off.
p mac --
"Bogus nonsense" is a double negative, of sorts.
Does this mean you think it's the truth?
ohmygod! the administration is revising information?? no...
stupid is as stupid does...