Paul Krugman thinks--and I agree--that "business cycle recovery" means the same thing as "falling unemployment rate." Glenn Hubbard uses a different definition in his Financial Times article, and so Krugman blows his whistle, throws his flag, and calls a fifteen-yard intellectual foul.
Posted by DeLong at October 11, 2002 09:58 AM | TrackbackJOBS, JOBS, JOBS: Of all the things to be upset about right now, a silly fallacy in Glenn Hubbard's FT article today doesn't rate very high. But it's a fallacy I hear a lot, so let me take a minute to bash it.Hubbard, like so many people, asserts that one reason for optimism about economic recovery is our continuing high rate of productivity growth. At first this sounds right: higher productivity growth means, other things equal, faster income growth, which means more spending, which means faster demand growth. So all that is good for recovery, right?
No, not really. What most people mean by recovery is job growth - an economy that is growing, but in which employment grows more slowly than the labor force, may be in recovery by some measures, but it will feel like it's still in recession. Now what does productivity growth do? It raises incomes and hence demand. But it also raises the growth rate the economy needs to achieve to create jobs, and to a first approximation it does so by exactly the same amount. That is, an economy with 3 percent productivity growth will, other things equal, have 2 percent faster demand growth than an economy with 1 percent productivity growth; but it will also need a growth rate 2 percent higher to keep unemployment from rising.
In fact, the "productivity growth helps jobs" story, if that's what it is, is just the flip side of the lump-of-labor fallacy, which says that productivity growth reduces employment - and equally wrong. (There's a different argument, about how productivity growth reduces the NAIRU - but that's about the supply side, not the demand side). You might say that I'm being too abstract; what about the lessons of history? But they all confirm this point. Terrific productivity performance in the 1920s didn't protect us against the Depression; all through the 70s and 80s Europe had higher productivity growth than the U.S., but worse job growth; you can multiply the examples.
I'm not surprised that Hubbard would make a silly argument; he is not, after all, a free man. But it's a sign of desperation, I think, that so many people have bought into this fallacy.
The point is - workers count. There is simply no reason for optimism about job growth at this point.
Posted by: on October 11, 2002 12:18 PMBy Paul Krugman
The Financial Times reports that it wasn't just Congressional Republicans and the accounting industry that blocked the appointment of John Biggs to head a crucial new accounting oversight board; the White House was "concerned at labor union backing for Mr. Biggs."
Workers, you say, who needs workers anyway?
Posted by: on October 11, 2002 12:54 PMHaving just read Hubbard's piece I do not understand Krugman's objection to it.
The title of the piece is "America Is Not In Danger Of Deflation", and it is written to counter the impression of the risk of future deflation given by "today's newspapers ... filled with claims that a ruinous 1930s-style deflation lies just ahead."
That's it. He doesn't talk about employment *at all* -- the word does not appear -- nor about any great "recovery" in progress (the word only appears once at the bottom of the piece). So how can he be accused of misusing that term, or of repeating some variation of the lump-of-labor fallacy about employment?
As far as productivity goes, Hubbard's entire comment is that high productivity counters deflation by increasing consumption. Here it is...
"[The] deflation scenario seems to make a lot of sense -- until you get out your calculator ... To start with, analysis of the productivity data over the past six quarters confirms some of the best news that economists have delivered in a generation - the acceleration in productivity growth that began in 1995 continues unabated. Thanks to this, today's consumers can look forward to real incomes that grow much more quickly than they have during the past 30 years - a good omen for current consumption."
And with this Krugman certainly seems to agree: "Now what does productivity growth do? It raises incomes and hence demand."
OK, they agree. Higher productivity increases demand, which helps counter deflation. End. Finis of Hubbard's discussion of productivity. No mistake that I see.
But Krugman seems to go on to manufacture a disagreement. He wants to talk about employment, not deflation. And since the subject now is employment, Hubbard's idea that productivity is good is a fallacy, because it need not be good for employment. Huh?
And then of course comes the obligatory personal shot (and shot at Republicans):
"I'm not surprised that Hubbard would make a silly argument; he is not, after all, a free man. But it's a sign of desperation, I think, that so many people have bought into this fallacy."
What fallacy? That higher productivity increases demand? Which is all that Hubbard said about productivity -- and Krugman said it too.
Questions:
As Krugman goes on about employment, does the reader get the impression that Hubbard maybe really did state some silly fallacy about the relationship between productivity and employment?
In Krugman's entire piece does the word "deflation" appear even once, so the reader would know that was what Hubbard was talking about -- and all that Hubbard was talking about?
Who deserves the flag for the intellectual foul?
Posted by: Jim Glass on October 11, 2002 01:27 PMDang it, where do they hide these things on the ft.com page?
Posted by: Jason McCullough on October 11, 2002 01:48 PM
>>Dang it, where do they hide these things on the ft.com page?<<
The url is a mile long, but just use the search feature for "Glenn Hubbard" or "deflation" and you'll get it.
>>To start with, analysis of the productivity data over the past six quarters confirms some of the best news that economists have delivered in a generation - the acceleration in productivity growth that began in 1995 continues unabated.<<
One of the problems here is that productivity is a poorly understood phenomenon - by all of us. The majority of the numbers being bandied-about at the moment seem to be numbers for 'labour productivity' from the BLS. And this is where the argument starts, because the 'jobless recovery' is a product of another process, the relentless drive by corporate America to cut costs, ergo increase output while not contracting new labour, or hiring young workers on short term contracts etc. It's very early to start making conjectures about these numbers because the serious analysis isn't through yet. But it is entirely consistent with the deflation/output-gap argument. US workers are trapped in a kind of modern version of the olive press, and at some stage all that can be will have been squeezed.
>>Now what does productivity growth do?<<
Perhaps this is an unfortunate way to present the story, since maybe productivity growth doesn't 'do' anything. In the present context this productivity growth is an effect of something, and it's understanding this 'something' that is the problem.
Paul Krugman himself in another piece on all of this makes the following point:
>>Brad DeLong is quite right that there's no such thing as excess capacity for the economy as a whole. I've been vociferous about that myself in the past. What I meant to say was over-investment in short-lived business capital, mainly tech, relative to other resources....<<
(http://www.wws.princeton.edu/~pkrugman/deflate.html.
This, rather than the conjuctural productivity/output-gap argument, is where the real problem lies. You see, I think it's just not credible to say that "over-investment in short-lived business capital" is pushing the global economy into a period of secular deflation (I take it as read that Hubbard just doesn't understand what's going on here). The normal expectation would be that this excess capacity would be worked-off, and on we would go again.But it's all too evident that this just isn't happening.
I've already indicated in this column (ad infinitum) some possible candidates: the continuing fall in ICT and biotech prices, the structure of world trade (especially China, but my guess is India won't be too long making its presence felt), and demographic processes in the G3 countries.
If demography is going to be important in the story then the deflation problem should be more acute in the EU countries (Italy and Spain in particular, but also Germany) than in the US, but obviously with Japan and Europe going downwards the US is bound to be sucked in.
Put this another way: if demography is not important, then why, in Stephen Roach's words is the world lacking an alternative global growth engine right now. Surely using conventional growth theory the 'convergence factor' should be offering plenty of catching-up momentum elsewhere.
Perhaps demography will turn out to be the modern economists 'pons assinorum'.
Incidentally, while the Glen Hubbard piece may be eminently forgetable, the ever-to-be-recommended Economist has a very nice (Krugman influenced?) piece on deflation, and especially the dangers for Germany right now - going 'naked into the conference chamber' - thanks to the voluntary Euro disarmament. (It's hard to remember that it was only 10 months ago that the world was cheering on this single currency 'experiment' - so why, oh why, have all the evangelists suddenly gone silent?).
http://www.economist.com/finance/displayStory.cfm?story_id=1382605
Here we go.
Looks like the bit we're talking about is here:
This modern deflation scenario seems to make a lot of sense - until you get out your calculator. When you do, the basic features of the US economy look quite good and deflation appears unlikely. To start with, analy sis of the productivity data over the past six quarters confirms some of the best news that economists have delivered in a generation - the acceleration in productivity growth that began in 1995 continues unabated. Thanks to this, today's consumers can look forward to real incomes that grow much more quickly than they have during the past 30 years - a good omen for current consumption.
He clearly implies the recessionary productivity growth will fight deflation, increase income, and help aggregate demand ("a good omen for current consumption").
Posted by: Jason McCullough on October 12, 2002 09:51 AMAhem, Jason, I suggest you scroll up and read Jim Glass's first comment again.
Posted by: Patrick R. Sullivan on October 12, 2002 12:19 PMDelong and Krugman are entirely right. We have an economic problem that will not be cured by productivity, rather by productivity and the proper fiscal and monetary policies.
This administration has put us in a sad fiscal state.
Posted by: on October 12, 2002 02:52 PM>> He clearly implies the recessionary productivity growth will fight deflation, increase income, and help aggregate demand ("a good omen for current consumption").<<
He doesn't imply it, he says it outright. And Krugman says it too: "Now what does productivity growth do? It raises incomes and hence demand."
OK. Krugman agrees with everything Hubbard said about productivity, since that's all Hubbard said.
Now Obvious Question #1 is: on the strength of this, just how does Krugman justify writing this entire piece to bash Hubbard for making the "silly argument" that "productivity growth helps jobs"? Where is that argument in the piece??
Please, somebody, go back to the FT story and quote the passage where Hubbard makes that argument. (Note that words like "jobs" and "employment" do not appear anywhere in the text.)
And Obvious Question #2 is: How do Krugman and our host Prof. DeLong justify criticizing Hubbard for giving in this piece a definition of "business cycle recovery" that they disagree with -- or for giving any definition of it all? Where does he do it??
Please look at the FT story and help me find the relevant passage, because I see the word "recovery" used only once, near the end: "Private forecasters expect the overall rate of consumer inflation to rise to about 2.4 per cent in 2003 as the recovery takes hold." Exactly how does that conflict with the Professors' preferred usage of "recovery"?
Before reading Hubbard the message I got from the Professors was "Hubbard, as he works for the Administration, and is not a free man, is exaggerating the strength of the recovery and using a bogus argument that 'productivity creates jobs' to do it." But that was before reading Hubbard's piece. And it's *hardly* consistent with him saying the recovery will take hold in 2003!
After reading Hubbard the message I'm getting is that the Professors thought "Since Hubbard works for the dishonest Administration he *must* be misleading people about the recovery and using bogus arguments, so let's bash him for that, whatever he said."
But doesn't anybody believe in "read before bashing" anymore? If not, maybe someone should throw a flag on the home team for unsportsmanlike conduct.
>>Delong and Krugman are entirely right. We have an economic problem that will not be cured by productivity<<
But, of course, Hubbard didn't say they would be, so it's not like they're right and he was wrong on that.
>> rather by productivity and the proper fiscal and monetary policies.
This administration has put us in a sad fiscal state<<
That may be. I might even agree. But then it should be easy to bash the admin's people for things they *do* say.
No, it's not right to bash people for saying things they didn't say (a recent theme around here).
Jim Glass writes:
>>
>> He clearly implies the recessionary productivity growth will fight
deflation, increase income, and help aggregate demand ("a good omen for
current consumption").<<
He doesn't imply it, he says it outright.<<
p => p
>>on the strength of this, just how does Krugman justify writing this
entire piece to bash Hubbard for making the "silly argument" that
"productivity growth helps jobs"?<<
Krugman never said Hubbard made the "silly argument" that "productivity
growth helps jobs".
>>Please, somebody, go back to the FT story and quote the passage where
Hubbard makes that argument.<<
Please, somebody, go back to the Krugman url and quote the passage where
Krugman says Hubbard was making "the 'silly argument' that 'productivity
growth helps jobs'"
>>How do Krugman and our host Prof. DeLong justify criticizing Hubbard for
giving in this piece a definition of "business cycle recovery" that they
disagree with -- or for giving any definition of it all?<<
Where did Krugman criticize Hubbard for giving in this piece a definition
of "business cycle recovery" that he disagrees with -- or giving any
definition at all?
>>>> He clearly implies the recessionary productivity growth will fight deflation, increase income, and help aggregate demand ("a good omen for current consumption").<<<<
>>He doesn't imply it, he says it outright.<<
Reading again the Hubbard article, Jim Glass is right by the letter that Hubbard's article asserts only that productivity growth leads to income growth, but wrong in the spirit of things, since Hubbard clearly doesn't understand the problem, and neither, unfortunately, does Jim Glass. You see this is the 'silly argument'. Silly because it's easy to make demagogically, and silly since it's so simplistic that no right-minded economist would be congenitally capable of making it.
As I have already indicated above productivity doesn't 'do' anything. Rising productivity can be accompanied by either rising incomes, or falling prices, it depends. And it's this 'depends' that really is at the issue here. Hubbard, following Glass, clearly believes that this increasing LABOUR productivity will lead to increasing real incomes, increased consumption, and thus, presumably increased investment (although this is my deduced mechanism since - thanks in advance Jim - it isn't made explicity), because he says so.
"the acceleration in productivity growth that began in 1995 continues unabated. Thanks to this, today's consumers can look forward to real incomes that grow much more quickly than they have during the past 30 years - a good omen for current consumption."
The deflationist argument is that the other possibility exists, that the increased productivity leads to falling prices as firms compete to sell. This is the role of the output gap argument. This argument is also backed by the idea that current consumption has been maintained by borrowing. It's not the wealth loss caused by the drop in the financial markets that's the problem, but the increased debt encouraged by the false sense of security provided by rising assett, and subsequently rising property, values.
Now either you believe this story or you don't. But Hubbard is wrong. There are TWO possible outcomes, and this time next year we'll have a better idea which one of the two it is.
>>Policymakers in America and Europe have been quick to dismiss any fears of possible deflation. Bond markets, on the other hand, reckon that the risk is mounting: bond yields have fallen to historical lows.<<
Economist print edition "Of debt, deflation and denial" October 10th 2002
http://www.economist.com/finance/displayStory.cfm?story_id=1382605
Is this a debate over basic economic relationships, or simply the politicization of language ("what has the right to be called a recovery")...?
Surely Hubbard is right in suggesting that productivity growth increases income and stimulates employment in the long-term (don't advocates of a link between changing NAIRU levels over time and productivity growth make the same claim on somewhat the same grounds?), and DeLong and Krugman are right when they point out that taking productivity growth as a gauge for the state of the economy forces one into "irrational exuberance" in the face of persistingly high unemployment figures.
>>One of the problems here is that productivity is a poorly understood phenomenon - by all of us<<
Poorly understood, because we continue in the assumption that it is a phenomenon at all!
"The first move in the conjuring trick has been made, and it is the one which we all thought was innocent"
L. Wittgenstein, "Philosophical Investigations".
................................
On the whole, I think I come down on Krugman's side. The key passage:
>>To start with, analysis of the productivity data over the past six quarters confirms some of the best news that economists have delivered in a generation - the acceleration in productivity growth that began in 1995 continues unabated. Thanks to this, today's consumers can look forward to real incomes that grow much more quickly than they have during the past 30 years <<
... only makes sense if one reifies "productivity" -- makes the assumption that what the "productivity statistics" are measuring is the actual productive capacity of a unit of labour, rather than the quotient of output and the workforce. It is this reification which is needed to make the assumption that incomes will grow as a result of "increased productivity", and it is this reification which Krugman is protesting about.
Both Hubbard and Krugman are horrendously unclear and contorted in their actual arguments, but this is always going to be the case when your subject of discussion is a philosophically suspect concept like "productivity".
Incidentally, it is now *thirty* *years* since Joan Robinson put the boot into the use of this kind of productivity talk, and twenty-five since Paul Samuelson admitted she was right. It is still not taught in American graduate schools, which since the topic of today appears to be "sad failings of the educational system", seems to be rather salient.
Posted by: Daniel Davies on October 14, 2002 03:54 AMLump of labor fallacy? Egads! Krugman has revised the ol' lolf in a way that I approve of. Based on what Krugman sez here about recovery and productivity, the appropriate policy response should be reduction and redistribution of working time. I agree.
Posted by: Tom Walker on October 15, 2002 06:11 PM