February 09, 2004

The CEA, OMB, and Treasury Forecast: My Guess

How is it that the Bush administration's analytic agencies--the Council of Economic Advisers, the Office of Management and Budget, and Treasury's Domestic Economic Policy--issue a ludicrous forecast? Table 3-1, p. 98 of the 2004 Economic Report of the President has a reasonable 4% real GDP growth forecast for 2004. But the payroll employment number is ridiculous: an average payroll employment for 2004 of 132.7 131.9 [updated for employment shortfalls since the forecast was frozen: the administration argues that since the forecast was made last November and since there's been bad employment news since, it's only fair to allow them to lower their target; they have a point] million--a forecast target which could only be hit by job growth over this entire year of 470,000 320,000 [as per lower target] per month, faster job growth than the American economy has seen, ever, and a forecast which implies that American worker productivity (which has been growing at an average of 3.4% over the past eight years) will fall by 2% not grow at all this year.

How is it that people wind up issuing a forecast that already has private-sector economic forecasters rolling on their office floors in helpless fits of laughter?

Here is my guess, presented only because I cannot think of another alternative, and because I can think of no way to justify stagnant productivity in the next year. In the fourth quarter of 2000, U.S. payroll employment was 132.3 million. In the first quarter of 2001, U.S. payroll employment was 132.5 million. Somebody in White House Media Affairs took a look at the draft of Table 3-1, and saw the real estimate for average payroll employment in 2004: something like 131.2 million.

Whatever powerful person it was then called the forecasters in on the carpet: "We cannot publish a number saying that payroll employment in 2004 will be lower than it was at the start of the administration. That number *must* be bigger than 132.5 million. If that number is smaller than 132.5 million, there will be lots of negative newspaper stories saying 'Bush administration forecasts negative job growth over first term'. We can't have that."

And so the number is 132.7 million.

That's my guess, but it's an informed guess. And if anyone in Treasury OEP, OMB's Office of the Chief Economist, or the CEA wishes to pitch me an alternative story for how the 320,000 per month payroll employment growth number fits with the 4% annual real GDP growth rate, I'll be happy to listen.

Posted by DeLong at February 9, 2004 01:32 PM | TrackBack | | Other weblogs commenting on this post

Allow me to tee one up:

Talk to the White House, starring Greg Mankiw:


Posted by: David on February 9, 2004 01:45 PM


Alot of people are flying blind in this job market.

I think it was Lehman Brothers who correctly called the latest disappointment (the january figure of 112,000). One of the honchos was quoted as saying they got that call right because they didn't believe the figures their models were showing so they threw out the models and came up with something based on 'a finger to the wind'.

The models are broken.

Want some globalization with your unemployment check, anyone?

Posted by: camille roy on February 9, 2004 02:14 PM


Obviously the key to getting the job growth they need is to reverse the last decade's productivity gains. I expect them to start launching their own "I love you" viruses soon.

Posted by: SP on February 9, 2004 02:29 PM


They cannot have made up a number and changed it. If they just changed the jobs number without tweaking underlying assumptions, the number would not compute.

They cannot seriously believe that they can print bogus numbers and not get caught? Of course it would be caught. Then the best they could do is blame their economists who don't know how to use a calculator. Is there another fallback? Publishing a bogus number would only serve to draw much more attention to the number when the correction is made.

If any number can be corrected without political damage, look for it to be the GDP estimate. Compare: "Administration predicts even greater economic growth in 04" to "Administration predicts lower job growth in 04".

How they squirm out of this one will be interesting.

Posted by: bakho on February 9, 2004 02:29 PM


in the movie biz, you just manipulate expectations all day. if you are a producer, you manipulate the studio's expectations for your new draft of a script ("yeah, i'm not too confident on this one") so that when they read it, they think, "hey, this ain't bad at all!".

there is no greater skill than making someone forget that the improvement shown by your latest effort took you from an F to a C. "IT'S THIRTY PERCENT BETTER" is all you have to say, a mantra for the modern age.

Posted by: Robert Green on February 9, 2004 03:00 PM


Of course they believe they can print bogus numbers and not get caught. Look at the unemployment figures - does anyone really believe that the jobless rate is only 5.7%, or that it is declining?

They Just Make It Up.

Posted by: Jon Meltzer on February 9, 2004 03:52 PM


Of course they can make them up. Whaddya think this is, an open society?

Jeebus, where's the stuff that documents this crap? Where's the underlying backup?

Up is left, right is violet.

Posted by: Duckman GR on February 9, 2004 04:19 PM


Of course they believe they can print bogus numbers and not get caught. Look at the unemployment figures - does anyone really believe that the jobless rate is only 5.7%, or that it is declining?

Actually, yes, given the way that figure is defined, I find it believable. Discouraged workers don't show up in it, and are increasing tremendously -- apparently the drop to 5.6% was more due to 300,000 discouraged workers than the 112,000 new jobs.

Posted by: cmdicely on February 9, 2004 04:22 PM


Brad, you are exactly right. Check out the handy chart over at Billmon's for validation of your theory of origins for this fucked up forecast.

The pattern is very clear. Each year the administration has in mind a number that they have to say they'll reach by '04. The number is slightly higher than the status of employment when Bush took office. They then forecast a linear trajectory to reach this end cap goal for '04. Each year their forecast doesn't pan out and is wildly off the mark. So what do they do? Turn around and do the exact same thing the next year. Check it out. Truly amazing stuff.

Posted by: manyoso on February 9, 2004 04:32 PM


Actually, it's not totally unthinkable that so many jobs could be created so fast.

Mostly, but not totally.

Massive demand-side tax cuts combined with calling in a few favours and getting some of the biggest employers to increase their rates of growth a lot..

A lot.

Maybe the latter is what they have planned. That they plan to call in a few favours and reverse it that way. If this is the plan..the question is..why not sooner?

Don't answer that. Stupid question.

Posted by: Karmakin on February 9, 2004 05:22 PM


One can safely presume they assumed this kind of exercise would be conducted. Therefore, they don't care. Favorable first-glance headlines are more important than credible economics. Blow-dried news anchors are more important than pointy-headed economists.

What's annoying is, behind this smokescreen, which eventually will be blown away, they're busy undermining our sources of economic growth and strength for the short-term benefit of their cronies. And to the extent they're called on anything, they're called on the smokescreen, rather than on their catastrophic mismanagement.

Posted by: bleh on February 9, 2004 05:26 PM


Productivity goes from 3.4% to ZERO??? LOL What a weasel!!!

They figure that productivity has been increasing so rapidly it cannot continue. Productivity gains stop tomorrow!

Posted by: bakho on February 9, 2004 05:28 PM


Now we just need you or someone to track the actual against the projection (graphically) from now until November ...


Posted by: Pilot on February 9, 2004 06:02 PM


I noticed today that something has changed about the way the news media report this -at least drive time radio news. After announcing the forecast the main Bay Area station had short comments from four economists -from liberal and conservative think tanks, Wall Street, and an academic. All four had the same reaction: a confusing forecast, the most charitable interpretation of which still seems very unrealistically -maybe laughably- high.
This approach might keep up until BushCo produces some economic forecasts that are someplace in the ball park.
Maybe the US Admin has dipped into the "How DARE you!? Of course we are credible" well one time too many. Anyway, the change was very noticeable, and -for me at least- bracing.

Posted by: jml on February 9, 2004 06:24 PM


Has Okun's Law beem repealed?

Posted by: Knut Wicksell on February 9, 2004 06:26 PM


Has Okun's Law been repealed?

Posted by: Knut Wicksell on February 9, 2004 06:28 PM


I remember some 2003 posts here saying that Okun's law is out of for this recovery. So what would have to happen with Okun's law for the CEA analysis to be accurate?

I guess it would take at least four or five scenarios to answer that, but I am curious about it.

I think Okun's law requires that there be a stable background growth rate in productivity that stays the same on average, and that is a big unknown right now.

Posted by: jml on February 9, 2004 06:41 PM


It could happen. Now that the G-7 has agreed to let the dollar plummet and with Wal-Mart squeezing even its Chinese suppliers to decrease their costs, we could soon see millions of new subminimum wage jobs stocking shelves and cleaning floors being created.

Posted by: Stan on February 9, 2004 06:42 PM


"Check out the handy chart over at Billmon's for validation of your theory of origins for this fucked up forecast. The pattern is very clear. Each year the administration has in mind a number that they have to say they'll reach by '04. The number is slightly higher than the status of employment when Bush took office. They then forecast a linear trajectory to reach this end cap goal for '04. "

Whoa there -- be careful how you interpet the chart. It shows forecasted PERCENTAGE job growth, not forecasted ABSOLUTE job growth, which means that 2%+- growth projected for 2004 equals progressively smaller payroll totals in each successive forecast -- because of the failure to meet PREVIOUS projections.

In reality, the job growth forecasted in the first Bush budget for the first Bush term was even more ludiciously high than the current forecast (although we didn't know it at the time because we weren't yet three years into the most jobless recovery in American history.)

It is true, however, that no Bush administration forecast has EVER projected negative job growth for his entire term -- which was a reasonable assumption two years ago but is looking pretty, ah, stretched right now.

Posted by: billmon on February 9, 2004 06:48 PM


Great, great stuff.

The crux is the point made in the posting that it came down to a political choice: either (a) tell the truth and say at the beginning of the 2004 campaign that Bush expects to end his first term with fewer jobs than he started. Herbert Hoover all over, with a $200 billion war in Iraq and fat cats with tax cuts to boot. Dead-solid certain election year loser.

Or (b) lie our asses off. Make up a target number to get a net gain in jobs, no Herbert Hoover, the tax cuts are working. See? Don't you want some more? The people that get bent out of shape are the wonky econ types, because hey, middle America believes the President, and NOBODY ELSE UNDERSTANDS THIS CRAP. How is Kerry/Edwards/Dean going to explain "the job projections will be off the mark nine months from now"?

In this environment, with this electorate jumping from flavor of the week to flavor of the week, I would do the same thing.

The Mayberry Machiavellis are in charge, and they know people that vote know more about Mayberry than they do about Machiavelli.

Posted by: 537 votes on February 9, 2004 07:20 PM


Odd. Aren't these employment predictions from the CBO, that same agency whose figures on predicted deficits over the next decade have been treated by Democratic "deficit hawks" like they came from the lips of God?

How do you decide which figures are crap and which aren't?

I don't know where the numbers in the table came from either. The text on employment on page 94 of the same document:


seems quite reasonable.

And I wonder how many commenters here actually referred back to the original document.

Posted by: tbrosz on February 9, 2004 07:44 PM


O/T, but it looks like Sullivain himself just joined the ranks of the shrill

Posted by: agrajag on February 9, 2004 08:30 PM


Yes, I did look at the Report, and did again after reading your post. There is no CBO data in there relating to an annual employment forecast -there are for tax incidence, insurance coverage, and some stuff on tort cases filed, but nothing on employment. As far as I can tell, the sources for the forecast that Prof D and others have been fussing over are just what he says they are. What CBO employment forecasts are you talking about? If there are CBO employment forecast figures in there a searches on "Congression Budget Office" or "CBO" certainly do not find them.

And since when has the CBO been considered infallible by people posting here?

But let that pass -I would be interested to know what CBO numbers you are referring to. They might be helpful.

Posted by: jml on February 9, 2004 08:46 PM


Krugman on employment:
-A disappointing column for me -too much commentary and too few numbers. And I think average people who read the NYT would understand the relevant numbers and the column would have been more convincing. So see, I can criticize the Great Krugman! I shake a tremble at what might happen if he take Revenge from his Olympus, but.....

Anyway, here it is:


Posted by: jml on February 9, 2004 08:58 PM


I must say, I did get interested in what other employment growth forecasts are out there, and if the CBO has anything similar to this. The CBO server seems to be down right now, though. I would be interested if anyone has info on the range of credible employment forecasts that are currently out there, by the CBO, or others.

Posted by: jml on February 9, 2004 09:08 PM


How is (any Democrat) going to explain "the job projections will be off the mark nine months from now"?

By producing and repeatedly broadcasting a t.v. spot consisting entirely of archival news footage of Bush 41 OMB director Richard Darman's testimony to a committee of the U.S. Senate immediately after the 1992 election, in which Darman candidly acknowledged that the administration had cooked federal budget numbers during the 1992 election campaign in an unsuccessful effort to bolster Bush 41's re-election prospects, concluding with an end title that reads:

"Like Father, Like Son."

Posted by: nightwriter on February 10, 2004 12:49 AM


tbrosz- is really confused. The report is from the WhiteHouse, not Congress. tbrosz links to a copy of the report on the WaPo server, (which is slow) and not to the original that is on the WH server. The WH server contains only official WH documents.

It is not clear that there is much communication between the WH and CBO given the recent discrepancy in the drug bill costs.

The bottom line is that Bush is in a political hole with the economy. He is in a hole because he could not be convinced to sign onto a true jobs package that would have at least given him a shot at having positive employment numbers during his term. Instead, he is stuck with the bad employment numbers as part of his record. His dismal economic record reinforces what we know about Mr. Bush as a failure in the business world.

As for how this will play out, it will become obvious before the November elections, that employment will not meet the projections in the report. Then there will be a lot of press coverage of the bad news and a lot of administration spin to try to counteract it. It would be better for the administration to get the bad news out now and try to recover before Nov than to get hit in October with bad employment numbers. The line about Bush being the first job loss president since Hoover will still be true and stick. It will be reinforced when the 04 employment picture becomes more obvious.

Posted by: bakho on February 10, 2004 06:41 AM


Some nonsense now and then is relished by the wisest men.

Posted by: Haas Levke on March 17, 2004 05:10 PM


Against boredom even the gods contend in vain.

Posted by: Taylor Stephanie on May 2, 2004 12:28 PM


Gratitude is the most exquisite form of courtesy.

Posted by: Espinola Steve on May 3, 2004 12:23 AM


The shifts of Fortune test the reliability of friends.

Posted by: Fischer Paul on June 30, 2004 05:52 AM


Post a comment