November 17, 2003

Actual and Potential Output

When industrial capacity utilization--as estimated by the Federal Reserve Board--falls one percentage point below its average of 82%, real GDP falls about 0.6 percentage points relative to the economy's productive potential. The industrial sector is a cyclically-sensitive one, and its relative cyclical movements are more than half again as great as those of the economy as a whole.

We can use this statistical relationship to get an idea of the relative movements of real GDP and the American economy's productive potential over the past generation and a half:

The naked eye, gazing at this figure, can easily discern two things:

  1. The acceleration in the rate of GDP growth in the mid-1990s as the technological revolutions in information and communications technology reach critical mass and significantly affect the growth rate of the economy's productive potential.
  2. The two great shortfalls of output relative to potential in the past generation and a half--the Volcker disinflation recession of 1980-1983 and the current shortfall of output relative to potential--are of roughly similar magnitude.

I remember the 1980-1083 Volcker disinflation. This current episode--although the shortfall in capacity utilization and in output relative to potential is as great and has gone on for as long--does not feel nearly as bad as that one did. One reason is the much faster current growth of potential output: even with the large shortfall in capacity utilization and the large output gap, real GDP today is some six percent higher than its pre-recession peak. With such rapid potential growth, our current depressed state of the business cycle is a waste of opportunity rather than (as 1983 felt in that slow-potential-growth era) an absolute decline.

A second reason is the differential performance of the labor market. Something like 1.5 percentage points of the rise in the unemployment rate that we would have expected to see from such a decline in capacity utilization has not been there--has been taken up in a rise in discouraged workers and a reduction in labor force participation much larger than one would have expected given the shift since the end of 2000 in the unemployment rate.

Posted by DeLong at November 17, 2003 12:50 PM | TrackBack

Comments

When you say that "this current episode does not feel nearly as bad as that one did," in part because of "a rise in discouraged workers and a reduction in labor force participation," do you mean that the statistics don't feel so bad, or that the zeitgeist is not reflecting the effects on these workers, or that the workers themselves are better off somehow? Because it seems to me that it's not much better to be unemployed and hopeless, as opposed to unemployed and not looking for work.

Posted by: Tyrone Slothrop on November 16, 2003 05:05 PM

____

A 10% lower inflation rate coupled with 10-20% lower interest rates might have something to do with it, let alone about 5% difference in unemployment.

Posted by: Josh Halpern on November 16, 2003 05:25 PM

____

So here's a question to which I don't know the answer: if there's so much excess capacity (as can also be seen from a post of mine here: http://www.angrybear.blogspot.com/2003_10_19_angrybear_archive.html#106665864158427290.), then why do firms keep adding more capacity?

By implication, this makes me wonder (yet again) how sustainable our current economic growth is. Business investment is doing pretty well right now, but why should firms continue adding capacity when there's so much to spare? Shouldn't we expect them to reduce their investment spending again?

Posted by: Kash on November 16, 2003 05:49 PM

____

It would be more informative to plot the difference of the two curves. The trend followed by the curves confounds the strength of the relationship, and can be misleading.

Could you plot the difference, or provide a link to the data so I can download it and plot it myself?

Thanks.

Posted by: Raul Martinez on November 16, 2003 06:21 PM

____

I've read about the discouraged-workers effect on the labor market a number of times in recent months. It seems like a good point, but I've never seen anyone compare today's level of discouraged workers with the level that existed during the 80s recession. Surely there were discouraged workers back than. Was it a smaller percentage than today, and if so, how much smaller? Have I simply missed it, or has anyone made or seen such a comparison?

Posted by: Maiden Lane on November 16, 2003 07:15 PM

____

Back a few years ago they changed the definiton of "unemployed" to remove nany of those discouraged workers from the unemployment numbers. So back then, the discouraged workers were just some early retirees. Those like the modern discouraged were counted as unemployed.

Posted by: Dick Thompson on November 16, 2003 08:11 PM

____

Do the "1992 chained dollars" include the technology deflator that count a $2000 computer this year the same as a $4000 computer last year?

Posted by: Michael Robinson on November 16, 2003 08:14 PM

____

two words: log scale.

please.

Posted by: wcw on November 16, 2003 11:06 PM

____

If US demographics look anything like ours (Australia's) then one reason you get a big discouraged worker effect (preventing a rise in measured unemployment) is that you now have many more women working than in 1983. Women's elasticity of labour supply is, in general, much higher than men's. Also these days you have many more older workers now than in 1983 who take premature retirement rather than keep looking for a job - ie another group whose labour supply elasticity is higher because they have some alternative.

To the extent that non-employment is occurring amongst groups with a higher supply elasticity the slump in labour demand will cause less welfare loss than an equivalent job loss in 1983 would have. Doesn't mean the aggregate welfare loss isn't still very substantial though.

Posted by: derrida derider on November 16, 2003 11:56 PM

____

Could it be that someone who remembered the Reagan recession was at a stage of career when their job was more vulnerable? and that now their own job is more secure? Just asking. Also, it depends on location. The job loss is not spread evenly across the country.

As for adding more capacity while utilization drops, Isn't this the hallmark of restructuring? In the Midwest, numerous big dinosaur steel have bankrupted or consolidated and are replaced by mini-mills.

A question. What effect does energy price have on capacity? I know that because of energy conservation, refinery utilization went from almost 100% to under 80% in the 1980s. It doesn't seem like conservation effects are a major factor this time. Does the lack of effort put more pressure on manufacturing because of higher energy costs?

Posted by: bakho on November 17, 2003 05:25 AM

____

Maiden Lane wrote:

"I've read about the discouraged-workers effect on the labor market a number of times in recent months. It seems like a good point, but I've never seen anyone compare today's level of discouraged workers with the level that existed during the 80s recession. Surely there were discouraged workers back than. Was it a smaller percentage than today, and if so, how much smaller? Have I simply missed it, or has anyone made or seen such a comparison?"

Yes, Brad posted the following link under a different thread from the Economic Policy Institute which addresses just this question:


http://epinet.org/content.cfm/briefingpapers_bp146

Posted by: Kosh on November 17, 2003 07:24 AM

____

Bakho and Dick Thompson have identified a couple of reasons to be wary of leaning to hard on either discouraged worker or jobless data in making comparisons to 20 years ago. Another thing making such comparisons suspect, I suspect, is that high levels of home ownership and high refinancing rates has kept the wolf from the door for lots of families. There is little reason to look actively for work when the odds are that work will not offer much that you want, in the way of pay, benefits or stability. Till things get better, you can fire the nanny, ditch the afterschool program, and take several grand in cash out of the house, lower your mortgage payments, or both, to make up for lost wages.

Posted by: K Harris on November 17, 2003 07:25 AM

____

here's a neat trick, add the unemployment rate to unused capacity (one hundred minus the capacity utilization rate), invert the scale and you have a good approximation of the output gap. twenty-two corresponds to the zero line for the output gap, fifteen to about two percent over the the zero line and thirty (about where we are now: twenty-five plus six) matches up with three percent below the zero line... and voila! we now have a fairly accurate monthly read on the output gap :D

Posted by: doho darat on November 17, 2003 10:02 AM

____

"Till things get better, you can fire the nanny, ditch the afterschool program, and take several grand in cash out of the house, lower your mortgage payments, or both, to make up for lost wages."

That is assuming that you own your home and even have a nanny. Most Americans don't have that level of income.

Posted by: bakho on November 17, 2003 10:04 AM

____

doho darat- no french please we're americans.

Posted by: big al on November 17, 2003 10:54 AM

____

I have a good equation for average hourly earnings growth that has worked for the last
almost 40 years. It makes wage growth a function of inflation expectations -- a moving average of the last 3 year CPI --, unemployment rate and capacity utilization, It now says average hourly earnings growth should continue slowing to about 1%, as compared to 2% now. This is consistent with a very large output gap and sub-par economic growth. Despite all the comments about how strong the economy is the year over year growth in real GDP is 3.4% -- the trend growth rate of actual real GDP since WW II.

Even the real averge hourly earnings argument that those still employeed are doing well is not that strong. Real wage & salary income is now 4% higher than at the economic bottom. In the last 5 recoveries, 2 years from the bottom this measure was up 18% on average. Moreover, the year-over-year change of real wage & salary income is still negative.

The question still remains, does the consumer have the ability to continue strong spending and is deflation a much bigger risk than inflation?

Posted by: SPENCER on November 17, 2003 11:22 AM

____

I was also thinking log scale is clearly appropriate.
The other factor in play here is the massive overinvestment now happening in China. Overcapacity will get worse, and disinflationary pressures will continue. And now, even some services' prices are being pressured, by cheap skilled labor from India and E Europe.

Posted by: Uber on November 17, 2003 01:58 PM

____

The "log scale" comment is apt.

In addition, a key difference between the Volcker recession and the current recession is the state of the economy in the preceding 6 years. To oversimplify, the dominant word of the late 70s was "stagflation", but in the late 90s it was "exhuberance". The Volcker recession started when the economy was already poor and made it worse. In the summer of 82 there were literally competent, educated people who couldn't get jobs at McDonald's.

Posted by: Z on November 17, 2003 02:40 PM

____

I FLATLY do not believe the top tail. Where is the boom??? A boom should show up as real GDP above the potential. If you connect the 95 point on a curve to the 2001 (?) inflection point, the current economy matches, and has been tracking actual & potential, and shows the extraordinary long dot.com boom more honestly.
Brad, the data certainly can't justify your point 1 in accelerating the rate of GDP growth as much as is shown -- there has to be some lag where the new real rate makes the curve diverge from a lagging prior rate before acceleration rate changes are justified.
I feel you're cheating here.

Posted by: Tom Grey on November 20, 2003 02:58 AM

____

Post a comment
















__