February 14, 2004
Payroll Unemployment Figure
The blue line on the figure below is the actual path followed by nonfarm payroll employment since early 1999. You can see payroll employment peaking in early 2001, then declining sharply until the end of 2001 during the recession period proper, then continuing a slow decline until the middle of 2003, after which it begins a slow advance.
The red line is the "projection" of employment that Glenn Hubbard made when he was Chair of the Council of Economic Advisers as part of the administration's propaganda campaign to convince the Congress last year that the then-proposed dividend tax cut was a "jobs and growth" program. Staff at the CEA, the Treasury, and OMB have hastened to assure me that this was not their forecast, that it was "not arrived at through the official Troika forecasting process," and was "Glenn's baby, and only Glenn's baby."
The orange line is the current Blue Chip consensus forecast for employment growth for the rest of this year. The green line is the "projection" released last Monday in the 2004 Economic Report of the President--the "projection" that was, as of the moment of its release, already 750,000 too high.
I remember seeing earnings estimates for telecomm stocks that looked like that, circa early 2001.
Posted by: Billmon on February 14, 2004 10:40 AMNothing like a graph to paint a picture of optimism/pessimism.
And this one tailored for the general ( I mean us) public(?)
Does anyone go to the trouble of painting ones for individual sectors that might have some utility for, say, those young people making career decisions? or IT workers? or manufacturing workers? or healthcare providers? or...
Is pragmatism passe? Lets do something (more)helpful than crush WH garbage.
Prof DeLong, care to draw your own line on the tail end of that graph?
Posted by: Alan on February 14, 2004 12:36 PMTake the slope achieved in the last administration and add it to the end of the curve, presto, the same performance. When you wish upon a star, makes no difference who you are...
Posted by: wolf on February 14, 2004 12:48 PMWhy did the NBER pick November 2001 as the end-date for the recession?
Posted by: Julian Elson on February 14, 2004 12:57 PMThe CEA projections are simply beyond belief. The discontinuity in the rates of growth are much too large to pass the giggle test. There has to be some induction time.
What would also help, and I hesitate to ask for more work from Brad DeLong, is taking the graph back to say 1988 or so, so we could see how the payroll employment curve changes as one emerges from a recession into a boom. Looking at this curve it appears that the inverse process (boom to bust) took a bit over a year (00-01.5) to appear in the employment figure
If ERP forecasts were made by professional economists, then I begin to understand your takes on outcountry outsourcing of jobs.
Posted by: Eli Rabett on February 14, 2004 02:31 PM>>There has to be some induction time.<<
Hmmm... Good question... It took about a year in the early 1990s to build from essentially no job growth to the pace of a full expansion...
In 1982-1983 it only took five months--but that was because the Federal Reserve kicked the economy with a big sudden interest rate reduction...
Posted by: Brad DeLong on February 14, 2004 08:09 PMYou know, due to the controversy over the household and establishment surveys, why doesn't the government just drop employment surveys altogether and have third parties do employment surveys? I mean, seeing as how various news organizations can do opinion polls every week... replicating the household and establishment surveys shouldn't be too hard. And this would give us several employment numbers rather than just one.
Of course, the government would still release numbers for those number in which it had exclusive bureacratic information for, such as jobless claims.
Posted by: Schism on February 14, 2004 10:28 PMMax Sawicky posted a good graphic of a similar phenomena regarding actual versus forecast job growth earlier this week here:
http://maxspeak.org/mt/archives/000146.html
Unfortunately, this is not an isolated incident. We are watching the politicization of all things economic.
This is important -- our credibility with our creditors( ie, the rest of the world) is at stake. (See http://bigpicture.typepad.com/comments/2004/02/more_on_job_cre.html for more on this)
Well, if we get back to the slope that we had in '99 over the next month or so, the CEA estimate isn't that far off. The odd thing is that the slope on the blue chip projection is historically strange over the course of this graph. I see a steep increase, a steep decrease, and a basicly flat area. Could the slow increase projected be a creation of some people projecting continued flat job creation, and some people projecting something along the same lines of the CEA?
Posted by: Dan on February 15, 2004 07:40 PMFine graph, Brad, but I think it's missing a "long term sustainable full employment" estimate. This is important because the dot.com boom was overemployment, and unsustainable -- and therefore an employment contraction was inevitable.
Unless the bubble pop, and it's second & third order ramifications, are more fully explored, it's hard for me to credit a lot of criticism of Bush's macro handling -- a big deficit of tax cuts and spending increases DID avoid a depression, despite the huge evaporation of paper wealth.
Posted by: Tom Grey on February 17, 2004 06:45 AMI also want to heartily congratulate you on trying to point out the political motivation that seems to be dominating forecasts.
Seems a lot like the deep green bogus science on global warming. There's prolly a good post there in using the global warming critics' arguments, against Bush's forecasts--but it's also prolly a lot of work to do right.
Posted by: Tom Grey on February 17, 2004 06:48 AM