February 10, 2004

Department of Productivity Growth

I think I've figured out the nonfarm productivity growth numbers (fourth quarter to fourth quarter) that underpin the Bush administration employment projection. The maroon bars are history. The yellow are the projection:

This shows the magnitude of the collapse in productivity growth--starting five weeks ago--needed in order to reconcile the Bush administration projections of employment growth with their projections of real GDP growth.

Hey. It could happen. This is a weird world. It certainly could be the case that the faster-than-normal productivity growth of 2002 and 2003 will be partially given back in 2004, and that productivity growth will be slower-than-normal (whatever normal is) in 2004.

But that's just one of the possible scenarios. I have a very hard time thinking of reasons that such a sudden collapse in the productivity growth rate down to 1.4% should be anyone's forecast, anyone's estimate of the middle of the path of possible distributions as we peer into the future.

Posted by DeLong at February 10, 2004 08:01 PM | TrackBack | | Other weblogs commenting on this post

Perhaps they're counting on all that fiscal drag, together with the incredible shrinking dollar, to collapse productivity growth, eh?

Posted by: john c. halasz on February 10, 2004 08:15 PM


I'll take "regression to the mean" for $50.00.

Posted by: jerry on February 10, 2004 08:25 PM


I should have posted this here:

"A White House official rejected the criticism, saying: "Accusations that we are deliberately fixing the forecast are preposterous." Productivity growth was expected to slow naturally from very high growth rates as the recovery proceeded, the official said."

I keep saying these people don't want capitalism to work so well for a while, that they don't want productivity going up so fast for a while and nobody believes me. Well, there, they eventually came out and said it themselves.

Posted by: bulent on February 10, 2004 08:36 PM


Professor DeLong, if with all your over-qualified help the Press Corp still doesn't get it, then they are unredeemable... Or perhaps, fewer PhD graduates should head for academic and institutional jobs, and more to work with newspapers and other media... Or is it simply that as far as our discipline in concerned, we are still living through some dark ages, somber times when the high priests still have more clout about our equivalent of planetary motion than physicists... When I am in a positive mood, though, I tend to hope / think that this kind of invaluable contribution drips down, almost subconsiously, and surely often unacknowledged, on the rest of the civil society and ultimately finds a way to influence public opinion and political outcomes. But there are days on which this feels does feel like a leap of faith...

Posted by: Jean-Philippe Stijns on February 10, 2004 08:40 PM


One would productivity growth to drop off after this recession-induced surge, no?

Posted by: fred on February 10, 2004 08:48 PM


Well, maybe the Bushies figured productivity growth will drop off because the Republicans would start spinning in circles, trying to justify the goofy budget.

Posted by: Lee A. on February 10, 2004 09:50 PM


I agree that the predictions are pretty absurd, but isn't it fairly normal for job growth to be accompanied by falls in productivity? Presumably the people you are hiring will be the least efficient workers.

This doesn't mean we'll see a collapse as they predict. On the other hand if I'm eyeballing it right, 4 of the 6 projected years are higher than 3 of the 9 actual years, so this isn't as absurd as some of their other claims.

Posted by: Brian Weatherson on February 10, 2004 10:23 PM


Productivity growth normally slows over the course of the cycle.

In the 6 recoveries from 1958 to 1992 -- excluding the aborted 1980 cycle --
productivity growth averaged 4.4% in the first year of the cycle, 2.2% in the second year and
1.6% in the third year.

On this basis you should expect 2% to 3% productivity growth this year. At 3% productivity and 4% real GDP hours worked, employment should grow 1%.

The Bush record speaks for itself and the public is recognizing it. I agree with the above post that this site is getting too caught up in the details of getting Bush and is not as good as it was when I first discovered it. And I am as big an
anti-Bush individual as you will find.

I think your record of attacking the press for its poor record of reporting and discussing economics is great. You are starting to have some success at getting the press to pay attention. Do not ruin it by becomming so extreme
that reasonable people will turn you off.

Posted by: SPENCER on February 11, 2004 05:50 AM


On the assumption that some of the recent acceleration results from a shift in Stateside employment toward more productive activities and away from less productive activities (a point anne has made here a few times, I recall?), then we must ask whether that process is running out of steam. Will we begin sending fewer low productivity jobs overseas in 2004?

More broadly, are pure managerial efforts (as seperated from capital spending or training) running out of steam? Inventory management seems a likely part of productivity gains. Inventory ratios remain rock-bottom. Will they fall further in 2004? Is the broadening of productivity enhancements through already well understood technology (quicker learning so less initial productivity loss) running out of steam?

The notion that productivity gains will slow to the 43-year average seems to give up on any effort to understand the sources of productivity gain and declare the result to be in the hands of the gods. Add some workers, put up with low-productivity periods of capital equipment installation and training on the new equipment, and down goes productivity, inevitably. Is that all there is to the rise in productivity in recent years?

Posted by: K Harris on February 11, 2004 06:26 AM


But if productivity drops, what will Wall street think?

Posted by: big al on February 11, 2004 06:28 AM


I'm doing my part to help the economy: I've done nothing but surf the internet from my desk for the last 5 weeks.

You're welcome, Mr. President.

Actually, I think that some of the commenters here are on track, that productivity should drop if and when companies start hiring again - I know my office is now busier than we've been in probably 18 months, but with 3 fewer employees than we had then. We're about to start hiring again, and may replace all 3, but I wouldn't expect (much) more total production once they're hired.

Nonetheless, Bush's forecasts are quite unrealistic, since they suggest a 6 year period of slow growth, whereas the previous array include both slow and very rapid growth - much of it occurring during a boom, I might add.

Posted by: JRoth on February 11, 2004 07:00 AM


The discontinuity going forward is that we have entered a period in which a weaker dollar will sustain growth and have left a period in which a strong dollar compelled a lot of restructuring. US firms had to be tough as nails to prosper in that environment, and productivity benefitted substantially. Maybe we'll have less creative destruction in the next several years.

Posted by: Jim Harris on February 11, 2004 07:31 AM


Thanks to K Harris for comment on productivity! My thoughts exactly. The productivity statistics are being used as a mystery meat fudge factor to be pulled out of the can when needed to explain why other numbers are out of whack (nice mix of images there -spam and Grandma's chocolate, but what I am talking about here is just as intellectually appetizing, so that is fine with me).
If we had a discussion of what the last year or so has taught about the sources of the current productivity growth, the whole debate would be more appetizing. Last time I checked the debate seemed to revolve around several theories about the increase in measure productivity:
1. a statistical artifact that didn't really exist
2. confined to a IT and computer sector
3. real and revolutionary for economic growth
So how do these three alternative stand up now, and what are implications of each for behavior of productivity over the business cycle?
I would love to hear thoughts on this from our fearless leader, and the other regular macro posters as well. (Where is that anne, anyway?)
The figures for change in productivity growth over the business cycle are interesting, but it would be more interesting to see how they would look related to other indicators of what stage cycle was in, rather than arbitrary calendar periods. For example, excess capacity, real income growth, etc. The arguments using those numbers here seem to assume that we are close to a significant upturn in activity consistent with the late stage of a boom. That seems to beg the question.

Posted by: jml on February 11, 2004 11:13 AM


Of course what interesting about this graph is the years chosen. Include years preceding 1995 and we have productivity growth of

1991 1.9
1992 3.9
1993 0.3
1994 1.2

In which case the projection looks, well kind of normal. A high period of productivity growth 1993 and 2002-3) followed by a remission 1993 and 2004 as new hires are integrated into the workforce. The graph as drafted is very much an attempt at comparing apples (boom productivity growth) with oranges (upswing productivity growth). Just doesn’t work and worthless for making projections.

Posted by: Giles on February 11, 2004 12:08 PM


Well, I haven't found anybody who thinks we're still in the 1973-1995 "productivity slowdown" period...

Posted by: Brad DeLong on February 11, 2004 03:55 PM


Hi All!I love your site! Mark painting

Posted by: painting on April 25, 2004 12:11 PM


You know what's interesting about Washington? It's the kind of place where second-guessing has become second nature.

Posted by: Green Matthew on May 3, 2004 06:07 AM


Fashion exists for women with no taste, etiquette for people with no breeding.

Posted by: Sciortino Paul on June 30, 2004 08:56 AM


i love the whole of this siteoooooooo

Posted by: mugu on July 14, 2004 03:38 AM


Post a comment