February 18, 2004

Administration 2004 Employment Projections One Last Time

According to the Bush administration's forecasting "Troika" (a team made up of the macroeconomic forecasters from the Council of Economic Advisers, the Treasury's Office of Economic Policy, and the Office of Management and Budget's Chief Economist) there was a straightforward process that generated the employment growth forecast released nine days ago--the forecast that the Secretary of the Treasury, the Secretary of Commerce, the Press Secretary, and George W. Bush are now running away from.

Here is the administration story as I understand it, and I am highly confident (but not completely sure) that I have what the administration would say (if the administration were saying anything, that is) right:

The administration story is that the forecasting process was carried out last November, when there was only a little data from the just-begun fourth quarter. The administration Troika made its forecast of demand growth, and concluded that the conventional wisdom was correct: that real demand and thus real output would grow at a roughly 4% per year annual rate between the third quarter of 2003 and the end of 2004. The administration Troika made its forecast of productivity growth, and concluded that worker productivity would grow at a roughly 1.4% per year annual rate between the third quarter of 2003 and the end of 2004. Since the rate of increase in employment is equal to the rate of increase of output minus the rate of increase of productivity, these two forecasts required that the administration forecast 4% - 1.4% = 2.6% per year for the rate of growth of employment. This 2.6% per year number was then bumped up to 3% (through some minor adjustments I don't understand).

Now a 3% per year rate of growth of payroll employment--that's a forecast of an average of 320,000 net new jobs a month. The Troika forecast that this rapid job growth would start with the number for last November (and, since it didn't, we are already some 740,000 behind the forecast). The Troika forecast that this job growth would continue throughout 2004. Thus the Troika forecast that payroll employment would grow by 3.8 million during 2004, that by the end of 2004 payroll employment would stand at 134.5 million, and that the average level of employment in 2004--taking all twelve monthly employment survey estimates and averaging them--would be 132.7 million. And that is the number that appears in the 2004 row of the last column of Table 3-1 on page 98 of the 2004 Economic Report of the President.

This 132.7 million forecast for the average level of employment in 2004 was then compared to the Troika's forecast (remember: this was made last November) of an average of 130.1 million for the average level of employment in 2003, and launched a lot of newspaper stories that said: "administration expects 2.6 million jobs to be created in 2004." That was wrong. The forecast actually expected the 132.7 million number to be surpassed in July--it is an average of the twelve monthly surveys--and expected 3.8 million jobs to be created in 2004.

But there is a huge problem with this employment forecast. The forecast's estimate of real GDP growth--4% per year--is in line with what other forecasters think now (and thought last November). The forecast's estimate of the unemployment rate--falling from an average of 6% in 2003 to an average of 5.6% in 2004--is in line with what other forecasters thought last November (and is a little pessimistic today). But the estimate of 1.4% per year for productivity growth is way, way low: almost every other forecaster is projecting that productivity growth in 2004 will be at least twice as fast. And the forecast of 320,000 per month for employment growth (already way above reality for the past three months) is way, way high. It is a good twice the average forecast (that's 2.2 million jobs high by December 2004). It is roughly 120,000 per month higher than the Federal Reserve's forecast (that's 1.7 million jobs high by December 2004). And it is a good 90,000 per month higher than the most optimistic upper tail of the distribution of private-sector forecasts (that's 1.2 million jobs high by December 2004).

As Dana Milbank writes in tomorrow's Washington Post: Bush Backs Off Job Forecast (washingtonpost.com): "Federal Reserve Chairman Alan Greenspan said Tuesday that the 2.6 million jobs forecast was 'a credible forecast' if productivity gains decrease this year. But the Fed issued a report Tuesday saying rapid productivity gains are 'likely to be sustained' this year." Dana Milbank thus concludes, combining Alan Greenspan's conditional "if... then..." statement with his staff's denial that the "if..." part is true, that the Federal Reserve is quietly saying that the administration's forecast of 320,000 average payroll employment growth a month is not a credible forecast.*

The 2004 Economic Report of the President is silent on the 1.4% rate of labor productivity growth forecast for the last quarter of 2003, for 2004 (and, I think, for 2005): no explanation is offered in the text. The CEA won't even officially confirm that number. The flow of information about the underlying details of the forecast has been shut down: senior administration officials insist on anonymity before they will tell reporters that they should carefully read footnote 1 to Table 3-1. Last year, as part of the lobbying campaign for Bush's 2003 dividend tax cut, then-CEA Chair Glenn Hubbard issued a projection of the economy's growth if the tax cut passed that--in a similar way--highballed projected employment growth, and for which no documentation or supporting materials were ever offered.

It is in this context that I feel driven to resort to numerology. When George W. Bush took office, the estimate of payroll employment stood at 132.5 million. The 2004 payroll number in Table 3-1 as printed is 132.7 million. A less optimistic employment forecast would make the 2004 number below the payroll employment number when George W. Bush took office. One can think of the sentences that would then be written by hasty reporters glancing at Table 3-1: "Bush administration forecasts that 2004 employment will be lower than at start of administration."

Given the recent behavior of this administration--the $160 billion left out of the 2009 budget so that Bush can claim to plan to cut the deficit in half, the removal of post-2009 estimates from the budget in the (largely correct) hope that if they aren't in the budget in black and red that the press won't talk about the really lousy projected deficit numbers after 2009, the highballing of employment growth estimates during last year's push to pass the dividend tax cut, the making of a huge deal out of plans to cut $4.9 billion (five-hundredths of one percent of GDP) worth of programs from federal spending, et cetera, et cetera, et cetera--one must wonder if somebody, somewhere, sometime decided that the "Bush administration forecasts that 2004 employment will be lower than at start of administration" should never be written, and that in order to keep them from being written the forecast had to have a 2004 employment number above the start-of-the-administration 132.5 million. Statements by this administration are simply not credible, and cannot be naively taken at face value.

Of course, an optimistic payroll employment growth forecast--a highly optimistic payroll employment growth forecast--a ludicrously optimistic payroll employment growth forecast--carries other political dangers that were perhaps not so apparent last November when the forecast was being made. To make a forecast means that one will, in some sense, be disappointed if reality falls short of it. And this week a number of people in the administration have been thinking about whether they want to be on record as saying, "Yes, I believe that 2.6 million (or, even worse, 3.8 million) jobs will be created in 2004." And they have been concluding that the answer to this question is "No."

Hence this week's spectacle: the Secretary of the Treasury, the Secretary of Commerce, the Press Secretary, and George W. Bush running away as fast as possible from the employment growth forecast that the administration issued a week ago Monday, and that it then pointed to as evidence that the economy would be strong in 2004 and that prosperity is just around the corner. Yes, prosperity is just around the corner, they still say. But just--not that much prosperity.

It's tough luck on the Council of Economic Advisers to find itself left holding the bag as everyone else in the administration rapidly flees the jurisdiction. If they did (through some mental process that I cannot follow) conclude that there was solid evidence that productivity growth was about to collapse to a low level we haven't seen since 1995 and thus that we would see 320,000 a month employment growth on average starting last November, they deserve backup from their Political Masters. If they tweaked the forecast in obedience to the commands of their Political Masters, they deserve backup all the more. Remember: the Treasury Department and the Office of Management and Budget helped make and signed off on the employment forecast.

*It is a possible scenario--it could happen that firms begin hiring like gangbusters, that the pace of work they demand from their employees falls, and that as a result productivity growth is anemic in 2004. The economy is a strange and surprising beast. But nonfarm productivity growth in the fourth quarter of 2003 has already come in, and its initial estimate is that productivity growth was twice as fast as the administration projection. Nobody I have found save the administration Troika and Yale University forecaster Ray Fair regards such a sudden collapse in productivity growth as a reasonable forecast.

Posted by DeLong at February 18, 2004 09:01 PM | TrackBack


Your guide to "Faith-Based Economic Forecasting"

1. Figure out how many jobs you'd need to create in order not to have a net job loss from the time you took office.

2. Determine what assumptions would be needed to get to this number (e.g. drop in productivity).

3. Say it's a forecast and pray real hard.

4. Rinse and repeat.

Posted by: Kosh on February 18, 2004 09:51 PM


I've been Google-News-Reports'ing "employee outsourcing" for several months now, gathering news from around the globe as the issue was first just an occasional management tickler, until recently, when the list of media stories on outsourcing approaches a dozen a day.

Who knew!? Consider....

If a company with ten employees making, on average, $25 an hour, costing the company some $50 an hour, when you include their workspace, lighting, safety, ergonomics in addition to taxes and benefits and HR, then the cost for that company to produce, say, 1,000 units an hour is $0.50 per unit, and say the units sell for $1, then employee productivity is gross $100 and net $50 per employee-hour. Basic math.

Say that company outsources one of its workers to overseas at $12 an hour, or less, but let's say $12.50 an hour, half their former employee's base pay, and a full 75% less then the total cost of an employee.

Now, skipping the rehash, employee productivity has grown to that same 1,000 units per only 9 employees costing the company $0.463 per unit, assuming the product is digital, say, one customer service call, ten lines of computer code or processing one tax return form, with no transfer cost other than bandwidth.

Instantly, the employee productivity has jumped to $53.75 per employee-hour, an incredible +7.5% increase. But the numbers of employees dropped by 10%, while at the same level of production. Recapping, for each 7.5% productivity increase, employment drops by 10% in a global economy.

Which is a long way to say, when you consider foreign outsourcing, which is expected to effect 1 out of every 10 Americans, the whole notion of (economic growth - productivity = employment) goes out the window. The old formula only applies to a closed domestic system. We are now edging into a global economy, where the Chinese can produce the same unit for $0.35 an hour, plus shipping, or (India)ns can produce the same unit for $7.50 an hour, with no transfer cost, between 5,000% to at least 500% productivity, instantly. (The Phillipines/ASEAN are said to offer somewhere in between those limits.)

Which brings me around to the other side of the coin, where we should pay attention. The effect of untariffed US agribusiness products, produced in an inverse productivity to manual farm labor on tenant farmer plots, is so staggering to the 3rd world, relentlessly overwhelming, a Niagara of cheap, subsidized product pouring over their borders, that millions, even tens of millions of rural peasants are being driving off their land, forced to abandon agriculture for a life in the slums, and a life at the bottom at that.

When we talk about economic growth in America, whether productivity is real, or statistically manipulated (stock overvaluations, unaccounted stock options and so on) and employment 'growth' (or merely overall subsidence to a series of crappy minimum-wage jobs), it's prudent to read some international press coverage of foreign trade issues and the impact of WTO.

Then we can look at the Federal Reserve's forecast with a clear eye, and see the storm of bull-hockey they hide behind their statistics. Even in the best scenario, we already know their statistics lie, and can't predict employment even one month forward! Then consider the hidden factors, the variables not even shown, you can see how worthless these statistics really are.

Like blind sports announcers calling the horses from the drumming of their hooves over the roar of the crowd as they close on the finish line.

Gibberish, in a fancy FR label. I'll **bet** that a statistical analysis of their reports back 20 years would show the standard deviation predicted actuals exceeds the actual variation.

So why is everybody laying money on those odds?

And presuming the sages at FR are at least as smart as these observations, then it's natural to presume *they* don't believe them either, which makes you wonder just what the hell all this whiteout of bull-hockey has to do with the upcoming election, international currency exchange and the prime rate.

A scam? A Ponzi? There is no doubt, whatsoever.

Posted by: Moe Clips on February 18, 2004 09:56 PM


They worked the numbers just like any poor suffering self-underemployed slob works the business expenses & income on his tax return so he won't get audited.

Not that I know anything about that.

Posted by: John H. Farr on February 18, 2004 11:53 PM


An excellent essay but it is 1624 words long. You should be editing it to 700 because it is probably heading for dead trees.

The troika is being treated like the CIA, first pressed to exaggerate then blamed when they are proven wrong. At least this time it didn't take a war to see that the claims were way off.
My guess is that there are a lot of furious economists at the moment and that someone should leak soon.

It is weird isn't it that, in this day and age, the projections in the Economic Report of the President are frozen 2 months before it is published. I mean drafts could contain a spreadsheat that could be updated in a minute.

I think this shows that the mere numbers are considered less important than spin and that the message people insist that the numbers not change while they are polishing the spin.

Posted by: Robert Waldmann on February 19, 2004 12:38 AM


Wouldn't it be safe to say that Dubya's posse was interested only in the initial headlines when the CEA announcement was first made? Everything else would be buried on page 42 of the national papers. After that, Dubya's Wur-lie-tzer can take care of the discrepancies, as we've seen with Scott McClellan's statements.

Posted by: PrahaPartizan on February 19, 2004 04:21 AM


Shorter Moe Clips:

"I have absolutely no clue how productivity is calculated but I do know that the hundreds of Ph.D's working for the Fed are wrong"

Posted by: strawman on February 19, 2004 04:22 AM


"Since the rate of increase in employment is equal to the rate of increase of output minus the rate of increase of productivity, these two forecasts required that the administration forecast 4% - 1.4% = 2.6% per year for the rate of growth of employment." Well, kinda. In fact, this assume that average hours worked (in each sector, probably) are fixed, so that any new effort comes through new employment. That is rarely the case. Hours growth leads jobs growth.

Thus "This 2.6% per year number was then bumped up to 3% (through some minor adjustments I don't understand)" is all the harder to understand. The 2.6% figure should be bumped down, not up.

Maybe its just a typo - 2.6 mln rather than 2.6%.

Posted by: K Harris on February 19, 2004 04:33 AM


One thing to keep in mind is that productivity is calculated by dividing output by the payroll jobs figure, which is significantly below the household figure. If, as I and other economists believe, the payroll number is too low due to measurement error, then the productivity number is too high. Indeed, I do think that recent productivity numbers are implausibly high, which is one reason why I think that the payroll jobs number is too low. Therefore, it is conceivable that some combination of real job growth and the discovery of missing jobs that are not now being captured in the payroll survey may give us numbers like those that the CEA is forecasting.

Posted by: Bruce Bartlett on February 19, 2004 05:16 AM


Shorter Strawman: " "

Posted by: ManStraw on February 19, 2004 06:05 AM


Imagine the WaPo citing this website as an authoritative source.


That says a lot about credibility. Kudos to DeLong for setting the record straight and educating the press.

Here is what Mr. Bush is now saying:


Q Mr. President, do you think the economy is strong enough to produce 2.6 million jobs this year, as your economic report projected?

PRESIDENT BUSH: I think the economy is growing. And I think it's going to get stronger. I do think there are some things we need to do. We need to make sure the tax cuts are permanent. I look forward to continuing to talk about this issue. Uncertainty in the tax code could affect small business planning. Uncertainty in the tax code will make it harder for our citizens to make rational decisions about spending money. We need to have an energy policy, open up markets for trade, we need less regulation. We need tort reform. There are things we can do to make sure the economy grows.

I'm pleased by the fact that since August there's been 366,000 new jobs, in one survey. There was another survey called the Household Account that's been more optimistic. But I'm mindful there are still people looking for work, and we've got to continue building on the progress we've made so far.

Posted by: bakho on February 19, 2004 06:05 AM


So, the administration was caught in "jobs creation related program activities?" The actual jobs weren't of course created, but that is enough to put out this report?

All this coming from our first MBA president...

Posted by: yam on February 19, 2004 06:07 AM


MOE CLIPS: your explanation of how outsouring impacts the productivity data is in error.

Your explanation assumes that the output of the new foreign worker is included in the output of
the domestic firm and that is incorrect.

The output of the new foreign worker will be counted as an import. Output is calculated as domestic consumption less net trade and change in inventories. It is standard practice that imports are an input into the production of many domestic goods. But the part of the domestic production
attritutable to imports -- the output of the new outsourced worker in your example-- is netted out of the domestic product.

Output of firm in your example will be reduced by
what the replaced worker previously produced, that is in your example. But the output of the foreign worker is not added back into the output of the firm as in your example.

The error in your analysis would be like triple counting the value of iron ore in a car by counting the iron ore and adding to it the value of the steel and adding that to the value of the car. That is not the way final output is calculated.

I've seen several people use your line of thinking, but it is not correct.

Bruce Bartlett comments that the payroll data may be in error may be on to something that is quite possible. But I doubt it. One, when the survey data is adjusted to be comparable to the
payroll data there is not a significant difference. Greenspan talked about this in his latest testimony and the JEC has reached similiar conclusions. Two, on a cyclical basis the last two years productivity record is very good by historical standards, but not so far out of line as to be unacceptable or unbelievable. Since 1995 -- when the productivity rebound started productivity growth has averaged 3.1%, what it averaged in the pre-1974 era before productivity growth slowed. the only thing extremely unusual about productivity over the last couple of years is that the normal cyclical productivity surge lasted two years rather than the more normal one year.

Posted by: SPENCER on February 19, 2004 06:12 AM



Either side of the ratio could be wrong. If one starts from the assumption that productivity growth is too high, one could as easily conclude that GDP is being overstated as that job growth is understated. The issue of which jobs tally is right has been hashed over a good bit, here and elsewhere. The end result, while it doesn't absolutely capsize the notion that the household survey is doing a better job of capturing actual employment trends, certainly undercuts earlier strong statements in that vein. Have you evidence that contradicts the defense of the payroll data offered by Chairman Greenspan, for instance?

Not to pry, but, are you the same Bruce Bartlett who is not advocating, but rather predicting, a tax increase in today's WSJ?

Posted by: K Harris on February 19, 2004 06:22 AM



Sorry Spencer. I was writing while you were posting.

Posted by: K Harris on February 19, 2004 06:24 AM


Now if we can just get this down to a sound bite for the Democratic candidates, we will have something. Somewhow I think the general voting public is not going to get into the subtleties of this whole problem. Although Bush is an announced non statistician, I must assume that he had some fairly significant exposure to statistics, and maybe even Economics, at Harvard University.

I hope the Dem campaigns are keeping in contact with you, Brad. Your input could be devastating to the W reselection campaign.

Posted by: tstreet on February 19, 2004 06:32 AM


Might the administration's change in overtime rules play into this? If fewer people get overtime pay, and that leads to a substantial number of people working fewer hours (instead of those people working the extra hours for free), then more workers need to be hired to work those hours. K Harris above said that hours growth leads jobs growth, but maybe jobs growth will grow faster if hours-per-worker goes down. I'm reminded of the French policy of reducing worker hours to 35 per week in order to increase the number of jobs. (I don't know if it actually worked.)

Posted by: David Weigel on February 19, 2004 07:06 AM


note that Froomkin of WP WhiteHOuse Briefing gets the job number (end of year increase of 3.8 million) right:


"The forecast issued by the president's own Council of Economic Advisers in the annual Economic Report of the President predicted a huge payroll growth by the end of 2004 -- 2.6 million jobs by some accounts, but actually more like 3.8 million."

Posted by: David on February 19, 2004 07:26 AM


The damn productivity number. Do I have this right?--its precariousness allows even the sharpest of macroeconomists to differ wildly on employment forcasts?
I appreciate (some of) the political context (that CEA estimates seem pre-meditated on defeating the statement 'Bush term shows net loss of jobs').
But I refuse to be bullied by this 2nd or 3rd rate derivative. I mean this little wort: productivity.
Surely-to-Christ we can get an estimate of employment without recourse to this fuzzy little...bastard.

Posted by: calmo on February 19, 2004 07:29 AM



I'm looking at the claims -

First notes on the series since 1967

- As you would expect, there is a hiring cycle.
- The Quarter on Quarter numbers exhibt the expected heartbeat pattern.
- The NS continuing claims, howver, is where the action is - the heart beat pattern is in place - with only two sbtantial breakdowns since 1967 - the late 1970's series of recessions, and 2001.

The normal cycle has a sharp peak, and then a gradually reducing series - the normal bull/bear paradigm where a market adjusts rapidly and then rebuilds slowly.

In every cycle, before a large uptick in continuing claims, there is a year which srongly breaks the descending pattern, the next year is the year that the situation rolls off the table - beginning the destructive phase of the labor market cycle.

Will post on bopnews tomorrow on this - it is very interesting

Posted by: Stirling Newberry on February 19, 2004 08:14 AM


from Milbank

"White House press secretary Scott McClellan, repeatedly asked about the forecast, played down its significance. "It's an annual economic report that is put out by the administration based on the economic modeling and the data that's available at that point in time," he said. "What the president stands behind is the policies that he is implementing." "

Question: why would the economic modeling McClellan refers to as the basis for the forecast NOT be expected to include the anticipated effects of implementing the President's policies? Is the forecast part of the rationale for supporting the policies, or is the forecast the baseline against which the policies should be evaluated?

Posted by: David on February 19, 2004 09:21 AM


SPENCER: re accounting for foreign labor

One important point Moe made was that the labor output was supposedly "digital", i.e. no physical import is taking place.

Consider the case of a software company that produces, let's say, a number of software tools that are sold predominantly to US customers. The company is headquartered in the US, and most of its employees are based in the US. Recently it started hiring people in its India/China office, so that now 10% of the employees are based in Asia.

Initially, "low-key" activities like testing, bug investigations, or support are farmed off to Asia, but as people there gain proficiency with the product, more substantial R&D work follows. The labor is delivered via the corporate network, and the product is still sold predominantly in the US.

I would imagine that it is at least hard to attribute who did what work, and to recognize revenue/productivity in the US and Asian parts "fairly". Supposedly the Asian employees are paid lower salaries, so the revenue of the Asian subsidiary would probably be relatively smaller.

Can you explain on which basis the work is attributed? I would think there is quite some leeway for the company, unless there is strict oversight of the internal reporting.

Moe's somewhat extreme proposition was that all of the Asian labor output is reported under the name of the US company. In a public company, the Asian subsidiary and the number of its employees would have to be disclosed, so this kind of fudging should not be possible. How about a weaker version, where the US company can maintain product price but lower product cost, thus recognizing Asian labor input at a higher "grade"? (That still wouldn't say anything about productivity numbers.)

Posted by: cm on February 19, 2004 09:22 AM


Good morning Mr. DeLong:

You can grind numbers until the decimal points are worn off, but one thing remains:

There is a very good chance that there will be more than a million new jobs in America by the time the election comes around and the stock market will have cracked the 11,000 barrier.

Your mission, should you decide to accept it, is to make this look like bad news.

You have skills, intelligence, and the economic background required. I have no doubt you are up to the challenge.

This blog will become a Pong game in five seconds. Good luck.

Posted by: tbrosz on February 19, 2004 10:06 AM


There is a very good chance that there will be more than a million new jobs in America by the time the election comes around and the stock market will have cracked the 11,000 barrier.

There may be a "good chance", but I wouldn't bet on the first part, and the second part is largely meaningless to most of the country without the first.

Your mission, should you decide to accept it, is to make this look like bad news.

Well, even granting your predictions as true, that's pretty easy. All you have to do is talk about the whole term rather than the 9-10 months befoere the election. 1 million new jobs would still leave Bush with the only net job loss for a term since Herbert Hoover.

Posted by: cmdicely on February 19, 2004 10:22 AM


"There is a very good chance..."

Good chance? Good chance?! How many more good chances do I have to hear before we get a real chance?

Bush should just do the right thing instead of the chancey thing. And, you're right it doesn't take a lot of additional number crunching to figure that out.

His economic policy supporters are dwindling yet now and again one pops up.

Posted by: dennisS on February 19, 2004 10:52 AM


Has anyone here addressed the odd inconsistencies between household employment and non-farm payroll employment? If you went by the former, there are almost 800,000 more jobs now than there were in January of 2001. I have yet to see a workable explanation for this. Charts show that self-employment only accounts for about a quarter of this.

Posted by: tbrosz on February 19, 2004 11:19 AM


I am not sure, but I am guessing that the economic data, with regard to company labor/materials imports/exports, come from firms' accounting data. hard to imagine an alternative source. The problem here is that, to the extent possible, companies will allocate profit among international units to minimize taxes. Ireland, for example, is a hugely profitable region for many large firms, because of the relatively lower taxes charged there. I think these managerial accounting constraints will tend to confound any neat productivity analysis.

Posted by: John on February 19, 2004 11:26 AM


"Dow 11,000" - measured in heavily devalued dollars

"a million new jobs' - which works out to 550,000 dollars of deficit for each new job. Ouch - even extravagantly low import quotas are a better deal - Stiglitz calculated that in the 1980's protectionism cost 100,000 dollars in higher costs for each job saved in the auto industry.

This is the worst recovery in the last 30 years for the jobs picture - other than the 1980 hiatus before things really got rolling in 1981-1982.

Posted by: Stirling Newberry on February 19, 2004 12:00 PM


"Has anyone here addressed the odd inconsistencies between household employment and non-farm payroll employment? If you went by the former, there are almost 800,000 more jobs now than there were in January of 2001. I have yet to see a workable explanation for this. Charts show that self-employment only accounts for about a quarter of this."

you'll find those jobs right next to the Iraqi WMD.

Posted by: Stirling Newberry on February 19, 2004 12:02 PM


Bruce Bartlett

Alan Krueger has written on the seeming anomally of 2 surveys and 2 different employment date outcomes. As far as I can tell, when statistical differences between the surveys are accounted for both surveys show a serious lag in job creation. Will there soon be a change? I hope so, but see no reason for a change. Add to the lack of job creation, the lack of growth in wages and benefits and we have a problem that will be sadly slow in resolving.

Posted by: anne on February 19, 2004 12:17 PM



Resources Form the Basis for Economic Growth

A POPULAR notion in economics today is that an abundance of natural resources is a "curse" for developing nations. Such an endowment, it is argued, encourages corruption, undermines institutional development, pushes the value of a currency uncompetitively high and cannot support long-term growth because the reserves eventually run out. Small wonder, then, that oil-rich nations like Iraq and Venezuela are poor or in decline.

Gavin Wright will have none of this. Mr. Wright, an economic historian at Stanford and long a specialist in the role that natural resources play in economic growth, agrees that overdependence on a single resource can lead to poor policies, but it is by no means inevitable. To the contrary, many developed and developing nations have used their mineral resources as springboards to wealth and broader-based development - not least the United States itself.

Mr. Wright and a colleague, Jesse Czelusta, have written a fascinating study (at www-econ.stanford.edu) on the subject that should be required reading. The lessons to be drawn are especially pertinent for countries like Iraq.

The economists start their analysis by looking at the evidence compiled by advocates of the resource curse. The seminal study was done by Jeffrey D. Sachs and Andrew M. Warner in 1995 and showed a strong statistical relationship between resource abundance and slow growth.

Many follow-up studies using the same method draw remarkably sweeping conclusions about the inevitable disadvantages of resource abundance. One recent study explicitly concludes that poor institutional development, including weak governance and property laws, is "intrinsic" to nations with oil and other minerals. Mr. Sachs and Mr. Warner have recently concluded that the curse is a "reasonably solid fact."

But Mr. Wright and Mr. Czelusta point out that almost every one of these studies uses the proportion of exports of the particular natural resource as a proxy for a nation's mineral abundance. Among other obvious problems with this measure, a high proportion of resource exports may simply reflect a lack of other kinds of exports, which is almost a definition of underdevelopment in the first place....

Posted by: anne on February 19, 2004 12:18 PM



Indonesia Takes a Tortuous Path to Oil

Deep underground near a town called Cepu on the Indonesian island of Java lies more than 600 million barrels of crude oil, $20 billion worth at current prices and enough to increase daily output in that impoverished country by 15 percent.

But Exxon Mobil, which has an exclusive contract to extract the oil at Cepu (pronounced CHEH-poo), will not pump a drop. Since finding much more oil than expected in 2a001, Exxon Mobil has been locked in negotiations with Indonesia's national petroleum company, Pertamina, which has demanded that Exxon Mobil share more of the bounty.

So it goes with oil investment in Indonesia, Asia's only member of the Organization of the Petroleum Exporting Countries. Indonesia this year assumed the rotating presidency of OPEC, which on Feb. 11 announced production cuts to stabilize prices. Indonesia will not have to cut: it already falls short of its OPEC quotas, and analysts say that falling investment is hastening an end to Indonesia's days as a net oil exporter....

Posted by: anne on February 19, 2004 12:48 PM


"So, the administration was caught in "jobs creation related program activities?" The actual jobs weren't of course created, but that is enough to put out this report?

All this coming from our first MBA president..."

the Harvard School of Business might consider issuing a blanket disclaimer which states that most graduates do not have such a fuzzy understanding of business, employment, and economics. Business school graduates usually are smarter than the Presidential example.

Posted by: bcinaz on February 19, 2004 01:15 PM


tbrosz, with respect to your second comment - about the delta between the household and the payroll surveys - yes, this matter has been hashed over here and elsewhere endlessly.

In short, the BLS and anyone else you'd care to ask will tell you that overall, the payroll numbers are more likely to be accurate than the household numbers, although it's not impossible that either: a.) this time it's different or b.) that the household survey is a leading indicator.

See, for instance, Bruce Bartlett's and K Harris's comments above (i am assuming that this is THE Bruce Bartlett, meaning one of the last honest supply-siders left).

As for your first comment, 1M new jobs this year is more than a might underwhelming, considering that labor force growth is typically somewhere in the 125K - 150K/month level.

As for the dow being at 11,000, what of it? Sure, my portfolio will be better off, but we already know that profits are increasing (since no one is hiring!) and that, thanks to low interest rates, PEs are increasing (whether ben graham would like them or not), but it doesn't say anything about jobs increasing.

In fact, i'll go further: if the economy were producing 3M new jobs this year, the market would react negatively, since it would suggest costs are increasing, profits aren't increasing, and interest rates will be increasing....

Posted by: howard on February 19, 2004 01:26 PM


IN case anyone wants to hav a look, here's the link to Ray Fair's model: http://fairmodel.econ.yale.edu/

Posted by: Charlie Martin on February 19, 2004 02:22 PM


Am I right in believing that Fair's model is simply calibrated to very long term historical data, and doesn't incorporate any assumptions about a secular trend in productivity growth? If so, the model would essentially assume that the surge in productivity last year was mostly cyclical, and the forecasts should be interpreted as telling you what will happen "if this cycle is like previous cycles". We know it probably isn't, but the model does not seem to be built to take that kind of knowledge into account.

Posted by: doctorwes on February 19, 2004 06:15 PM


Msg to Spencer re. the productivity numbers

Your comment on calculating the output (and thus the productivty) seems correct but misses the point on outsourcing. When one job is outsourced the ex-worker still purchases goods and service (either credit or savings) so domestic consumption remains steady (for now) while the tally of hours worked is cut. Yes, there is an offsetting entry in the form of imports but given the cost differential between countries the drop in domestic work hours far exceeds the increase in import costs. Thus driving the productivity numbers high. Given the poor level of infrastructure investment over the last 30 years the numbers should be far lower. Or am I mising something here?

Posted by: MSBCEO on February 19, 2004 06:19 PM


"This 2.6% per year number was then bumped up to 3% (through some minor adjustments I don't understand)"

The most likely source of the "minor adjustments" is that GDP growth is for the whole economy, while the standard measure of productivity growth is for the nonfarm business sector. The part of the economy outside the nonfarm business sector has (mainly by assumption) very weak productivity growth. As most people who've put together a macro forecast know, this kind of thing typically pulls a forecast of GDP growth down by about 4 tenths relative to total hours growth in the economy plus productivity growth in the NFB.

Also, somebody brought up hours per worker. If the scenario was one of a cyclical fall in productivity, like it apparently was, then it would be natural to forecast a fall in hours per worker as well. This is another reason to bump up the employment forecast a little.

Is a little retreat in productivity growth unreasonable? I don't think so. The last couple of years have been very high and likely owe at least a little to unsustainable increases in effort.

I actually think the forecast of emloyment was at the high end of reasonable for when it was made. And given what I know of how the process works, it was probably a nearly completely technocratic product. Of course, given all the fuss and accusations, we'll never get an unpolitical forcast again.

Posted by: pi on February 19, 2004 09:14 PM


Regarding the posting of Alan Krueger's article referencing the houshold - payroll gap, it was right then, but not now. The gap after adjusting for self employment, multiple job holders, etc. has opened up quite a bit in the last 6 months.

Posted by: pi on February 19, 2004 09:56 PM


Pushing Spaghetti Uphill In World War IV

[from Lacy Hunt, of Hoisington Asset Management, from John Mauldin's, 'Barbarians at the Fed']

“Dr. Hunt, points out that M2 and M3 contracted in the fourth quarter at the fastest rates since the end of World War II (6.8% for M3), and that it’s even more alarming that the six-month change in M3 hovers close to zero, since sustained stagnation in the aggregates ‘foreshadows slower economic growth.’ In fact, Dr. Hunt adds, ‘while the drop in the Ms may be a reaction to the rise in real interest rates in the second half of ’03, it may also be an early sign that the Fed is now ‘pushing on a string’ in the words of Keynes. Which would mean that demand for credit is falling because there is already too much debt and borrowers are uncomfortable with this much leverage—thereby raising the specter of debt deflation.

“What’s more, advised Dr. Hunt, don’t breathe easy because M3 rebounded in January, apparently growing at close to an 11% annual rate—because an accounting change (FASB’s FIN 46, which basically makes banks consolidate the assets and liabilities of special purpose entities) is messing up the comparability of the money supply stats. It’s not something the Fed has been advertising, Dr. Hunt says. He had to notice a footnote in the stats, and Ried Thunberg’s Bill Jordan had to chase down Fed officials for a clarification of some muddy reporting, but when all is said and done, this pair of gimlet-eyed Fed watchers reports that rather than rebounding sharply in January, M3 was essentially flat, or maybe up 1%. Not a bullish portent.”

But The Fed under Bush (if presidents have any affect on that secret and private Fed anyway) has so backed us into a corner, fiscally speaking, there is no exit strategy. Just as in Iraq, it was a whole lot easier to drop into rock bottom interest rates and deficit spending and massive trade imbalance than it will be to claw our way back out again.

Rather than thinking of our current economy as a car made nitro-fuel injected, where all you have to do is step on the gas, it might be better to think of our economy as a car at the top of the Rockies, engine shot, running in neutral and the brakes faded off to 1%, rocketing faster and faster down the hill, out of control. Will we ever get to the bottom? Oh, most assuredly!

Look at what Andy Xie says in Mauldin's newsletter: "“Macroeconomic policies should aim for stable growth rather than "full" employment, in my view. The integration of global labor forces through outsourcing makes the former difficult to attain. Too much stimulus only inflames asset markets, shortening and amplifying economic cycles. Thus, macro policy should aim at capping credit growth and fiscal deficits."

“The sustainability of globalization depends critically on whether the West is willing to embrace structural flexibility as the main tool for absorbing adjustment, rather than relying solely on stimulus. If the West continues to look to macro stimulus to solve economic problems, I believe it will lead to more bubbles that could eventually end with a major crash, a prolonged recession, and possibly trade protectionism that would derail globalization.”

"[and Mauldin's recap:] What Xie seems to be suggesting is that we allow the unemployment rate to rise because the global labor arbitrage makes full employment difficult if not impossible, and that by using massive amounts of stimulus, we create another bubble. Thus, the US should tolerate jobs leaving our shores and adjust slowly over time."

Xie is right about bubbles, but patently absurd about labor. There is no global labor arbitrage. There is only 1st world high-tech labor vers 3rd world high-tech labor, then there is 3rd world wage slavery. A unit-cost of one is up to 100x's the unit-cost of the other. WalMartization of labor does not "adjust slowly over time", it runs away like a forest fire. There is no "stable growth" in labor arbitrage when foreign outsourced labor units are selling for $2.50 a DAY, only bankrupcy for the 1st world laborers. Who are these free-loading corporations with a US beachhead and foreign workers, enjoying our feast!? If a child kidnapper was white slave trading to some sheik, would we call it 'structural flexibility'? When there are 100,000,000 Chinese manufacturing will-work-for-food's, and possibly 1,000,000 Indian high-tech wage-slaver's, what we are talking about is World War IV, not laissez faire arbitrage!

The better alternative is to turn massive deficit spending inwards and away from Defense, with its abyssmal productivity (how much, really, is a Star Wars system worth, that wastes a couple $T, that can't be exported, and that creates a permanent maintenance and support blackhole?). Invest instead in public works. They may be wasteful, compared to private construction financed by tax incentives, but not nearly as wasteful as Defense. Besides, the nation, especially the East Coast, is crumbling fast. We need a complete quadruple bypass.

Put the brakes on government and defense outsourcing, one, that saves the good jobs, then cut all the Star Wars - Hypersonic Space Plane crap, the International Space Station fiasco, the Moon-to-Mars hogtrough, and put those $T's into a program to retrofit or replace or upgrade the nation's bridges, oil & gas pipelines, rail, sewer and water distribution lines. Just our arteries. Forget about the overbuilt roads, overbuilt buildings, plenty good treatment plants, that can come out of State budgets, once people have decent trades jobs and are buying houses and paying taxes.

And put a 0.1% transaction tax on stocks, funds and bonds. That cuts international computer day-trading out and stabilize the US economy. Otherwise, things are going to get more and more choppy until the bottom falls out of the bucket.

Then gradually put on the brakes with higher interest rates, using 3% as the maximum allowable housing increase. (Think of how easy Greenspan's job really is, with perfectly transparent stock market & real estate price reports!)

1) Outlaw All Government Foreign Outsourcing, including Its Contractors; Pause H-1B Program
2) Cut All Defense Spending on Science Fiction (Star Wars) and War Policy Fantasy (AF 2025)
3) Put Those Savings in Instituting US Public Works on Arterials, Bridges, Pipelines, Rails
4) Institute a 0.1% Transaction Tax on Every Stock, Fund and Bond Trade, and Cut Margins
5) Gradually Increase the Prime to Compete with Euro & Yen Basket, Holding Housing at +3%

It's as if we all of us were on Apollo 13, heading for the moon in 2000, when boom, there's a hole in the WTC, and suddenly the moon seems very far away, and getting back to earth seems even that much farther.

Only the Fed and the Bush Administration are saying, "Apollo 13, we are studying solar flares, and trying to get the moon and planets into conjunction. Just hang in there, as soon as we decipher Jupiter's Red Spot, we'll move the heavens and the earth and burn all our treasure in a big pile so you can find your way back."

Here all we wanted was fiscal conservatism, cutting out fraud and porkbarrel, and a shot at some decent jobs! Does it take an MBA from Harvard to figure that out? Of course, I'm assuming they all aren't just traitors, holding 250,000,000 Americans hostage while they play at world trade and labor arbitrage and currencies. I've always been more partial to incompetence than conspiracy theory, but that's scarey either way, eh?

Our alternative is a Central American nightmare of government plutocrats, oligarchs and peasant campasenos, the guts of the middle class ripped out and spread across the village square for the vultures to feed on, and a massive disinflation on everything, to where you'd be forced, as I heard one old DustBowler shrug, "We had to work like hell to make a (single) dollar, then struggle hard to find somewhere's to invest that dollar so's we'd make sure to earn another one."

China is the lynchpin. They have pegged their currency to ours, stuck tight as a tick, while at the same time they're pouring their assets and their reserves into Australian real estate and commodities, hedging the decline of the US dollar by buying rising AU dollar instruments. It's a wicked three-way pump handle, pouring a Niagara of US treasure inexorably and irreversably across the Pacific.

Well, I'll close with an anecdote. When I went to San Francisco in 1996 to see ComDex, my taxi driver was studying Java programming at night school. There was high-tech gold rush fever in the air! Today, 8 years later, my peers are taking night jobs anywhere they can, hedging against imminent loss of their careers. All you have to do is give people a green light and a road ahead. The rest will take care of itself.

Rock the vote, or play with your noodle. Everyone out into the trenches and back to work.

Posted by: Madame Zhiu on February 21, 2004 02:11 AM


Post a comment