Ed Glaeser on Paul Krugman

Ed Glaeser on Paul Krugman's contributions to economics:

Honoring Paul Krugman, by Edward L. Glaeser: ...Paul ... Krugman’s fame as a public intellectual should not lead anyone to think that they understand his contributions to economic research just because they regularly read his columns.

The Nobel Prize citation highlights two distinct but connected contributions: Mr. Krugman’s development of the “new trade theory” and his work on the “new economic geography.”

International trade has a long history in economics, and for the bulk of the field’s history, patterns of trade have been explained by factor endowments and comparative advantage. Why does England export wool and Portugal export wine? The cold winters of Yorkshire produce really fluffy sheep and the banks of the Douro produce splendid grapes. Yet comparative advantage does little to explain much of modern international trade, especially not trade within industries.

Mr. Krugman published two seminal papers in 1979 and 1980 that made sense of the fact that Toyota sells cars in Germany and Mercedes-Benz sells cars in Japan. Mr. Krugman started with a variant of Edward Chamberlain’s model of monopolistic competition. In this model, every firm sells a slightly different good — an Infiniti is not exactly the same thing as a BMW. There are fixed costs of production, which means that producers get more efficient as they sell more. Finally, consumers like variety...

These ingredients came together and provided a framework than can match the world’s trade patterns better than the 19th-century framework of David Ricardo, or the mid-20th-century models of Eli Heckscher, Bertil Ohlin and Paul Samuelson. ...

Mr. Krugman’s trade models became the standard in the economics profession... His models’ combination of realism, elegance and tractability meant that they could provide the underpinnings for thousands of subsequent papers on trade, economic growth, political economy and especially economic geography.

Mr. Krugman’s 1991 Journal of Political Economy paper, “Increasing Returns and Economic Geography,” is the first article that provides a clear, internally consistent mathematically rigorous framework for thinking simultaneously about trade and the location of people and firms across space. It is one of only two models that I insist that Harvard’s Ph.D. students in urban economics be able to regurgitate, equation by equation.

The model begins with the same basic elements as the new trade theory: monopolistic competition, scale economics, love of variety. To these elements Mr. Krugman adds free migration of workers across space and industries. ... The paper provides economists with a clear framework that can make sense of where we all live. Firms and workers are pulled toward the same location to reduce transportation costs of shipping goods. ...

Of course, we don’t all live in the same city. A good model of geography needs both a centripetal and a centrifugal force. In Mr. Krugman’s model, populations are pulled apart by the desire to be close to natural inputs, like land or coal mines. ...

In his public role, Paul Krugman is often a polarizing figure, loved by millions but also intensely disliked by his political opponents. ... Within the less divided world of the academy, Mr. Krugman’s economic research has generated plenty of light, but far less heat. His papers are universally acknowledged to be immense contributions that helped to create two distinct fields. His Nobel Prize is extremely well deserved and not unexpected. I, for one, had bet on him in Harvard’s Nobel Prize winner pool.

Comments

[Barkley Rosser]

So, I have been going around to a lot of blogs clarifying Krugman's exact work and role, with my own post on the matter on econospeak. Guess I should do so here as well, given what a frequent contributor I am here.

I happen to think Krugman deserves the prize, but Glaeser's discussion is overly hagiographic to the point of inaccuracy. The Nobel committee was more precise and accurate. Krugman got it because he applied and extended the application of a useful model of increasing returns (the Dixit-Stiglitz model, 1977) to both international trade and to location theory/urban and regional economics/economic geography. Where is Glaeser off?

He implies that Krugman was the first to do this (and is a bit vague about what it was he did) in each of these fields separately. He was not. The Dixit-Stiglitz model was first applied to international trade theory by others, notably Brander and Spencer and also Helpman. Krugman cited all these folks, and arguably he did it better and certainly more extensively and with more public impact.

In location theory the first to do was Masahisa Fujita in 1988. Krugman also cited him in his later work and would later coauthor with Fujita, as he would also with Helpman. But he was not the first to apply Dixit-Stiglitz in either area. Glaeser is quite misleading here.

What stands out is his linking of the two areas, making him similar to the last recipient of a prize for international trade, Sweden's own Bertil Ohlin in 1977, whose most famous book was _International Trade and Interregional Trade_.

While Krugman is a worthy recipient (and probably more so clearly in the area of foreign exchange rate modeling), there are some wrinkles here, besides the non-recognition of any of those who applied Dixit-Stiglitz prior to Krugman. Probably the most serious is that the unrewarded developer of the model that Krugman won for applying has not received the prize, Avinash Dixit, who must be applauded for his generous piece about Krugman on his receiving the prize (Stiglitz got one for asymmetric information, of course, but the paper with Dixit has been unrewarded). This is sort of reminiscent of Lucas getting it for applying rational expectations to macroeconomics while the main developer of the idea of rational expectations, John Muth, remains without a prize.

I must also note that while Krugman himself has praised himself for being the first to provide a mathematical analysis of agglomeration economies in economic geography, in fact there was a large literature in the 1980s by physicists and others mostly in regional science, environmental, and geography journals that did so, although they did not use Dixit-Stiglitz. Some involved in this included Peter Allen, a student of Ilya Prigogine at the Free University of Brussels (now at Cranfield University in Britain) and Wolfgang Weidlich of the Stuttgart Institute of Theoretical Physics. Krugman is all excited about Dixit-Stiglitz, but its basis is the idea that the source of agglomeration economies is consumers loving variety, the "bright lights" theory of big cities. This is a factor, but most people discussing this think that the more important sources of agglomeration economies are on the production side, something Krugman himself has recognized in some of his writings. Of course it is not just Krugman who does not cite any of these people, but Glaeser as well.

[dale]

In Oct. 2003, on the eve of Paul Krugman's visit to my church in Portland (where he was so speak on his then recently published book, The Great Unraveling), Professor Brad DeLong passed on the rumor that Paul was on the short list for that year's Nobel. He didn't win that year of course- but he went on to speak to over 650 folks at that event. All eager to hear the man who was one of the few sane lights in the mainstream media at that time.

It's almost hard to remember those early years of the Bush administration- when everything was going so drastically wrong (as they still are) with little mainstream media criticism. Paul had one of the nation's most valuable piece of media real estate and he used it well. His columns and his speaking appearances were a reality check against the irreality of what seemed to be the Spirit of those dark days.

So UO 07- here's to a little hero worship and gratitude to Professor Krugman. I don't give a fig about Mao- but I don't think the major figures of our nation's liberal history would begrudge our admiration for Paul Krugman. I second Paul Samuelson's remarks that it is a bad on the Pulitzer committee that PK has not won their award as well. For it's his popular writing for the NYTimes that has been his greatest contribution from my POV.

[reason]

BJF
Paul Krugman writes opinion pieces for the NYT. He is angry, because there is lots to be angry about. Go to Princeton if you want him to be a professional economist.

[hari]

Berkeley Rosser is claiming he knows better than Nobel Committe on the *raison detre* for their (single) selection of Laureate this year. There is always jealousy and whatnot in academic world, as I know it well from my days as one, and it will always be reflected in Nobel Price award.

[Joe]

I fully agree with Barkley Rosser's views. I believe that there are more deserving candidates for Nobel than Krugman--but apparently they are not as 'smart' as Krugman is.

I was an early addict of Krugman's academic and non-academic writings. However, I have come to the realization that Krugman's desire to be a public intellectual has forced him to be the expert on anything that is making waves in the public space, resulting in his making seriously wrong opinions. His dismissal of the East Asian economic growth 'miracle' as all about factor intensification is a notable example. A Singaporean friend once told me that policy makers there took Krugman's view so seriously that even routine greetings at shops were cut down in order to improve efficiency.

[anne]

"His dismissal of the East Asian economic growth 'miracle' as all about factor intensification is a notable example."

Quite the lying, for the truth is entirely opposite, but go right on lying.

[Joe]

It appears that there is another person who uses the moniker Joe in these threads. Henceforth I will use the name Joe P.

As to the above comments by the other Joe, I don't have the time to pull apart his assertions other then to say that Krugman has striven mightily to keep discourse on the right path.

There are people, given their wherewithall, may have sincere disagreements with Krugman's assertions. I think in a vast majority of the cases these people are just sincerely wrong.

Other, less sincere, have objections whose root origin stem from venality.

Regardless, we need to dissect the issues not the motives for no one knows another's heart.

More or less I need to be remindful of my own admonition, for more or less I follow it.

[hari]

BR must be stupid to go on other blogs and register his opinion on the choice of Nobel Committee. I wander who gives him the license; or is he such a fool to think his views are of any real value on Paul's academic achievements.

With or without a white beard, intellectual stuff is not part of the herd or tribal mentality. It belongs to our civilization.

For those of us in trade policy, Paul is indeed an authority and has been for a long time. I even mentioned here sometime ago he is worthy of Nobel Price because of his academic work.

I've checked other sites and there are few and far between who dare express an opinion contrary to Swedish Nobel Committee - Paul could have been on their list now for atleast five or more years!

[anne]

"However, I have come to the realization that Krugman's desire to be a public intellectual has forced him to be the expert on anything that is making waves in the public space, resulting in his making seriously wrong opinions. His dismissal of the East Asian economic growth 'miracle' as all about factor intensification is a notable example.:

Among the singular accomplishments of Paul Krugman, for which he has been continually praised in Asia, was recognizing inherent development weaknesses that characterized what in the early 1990s came to be considered a development miracle, only to be found quite wanting later in the decade. Criticism, that where properly considered is allowing for fuller development currently.

[Barkley Rosser]

Joe and Hari,

For starters, let me repeat the opening of the first sentence of my second paragraph: "I happen to think Krugman deserves the prize." I did not say so, but I even think he deserves getting it alone. He is brilliant, and this is clearly a "career" award, for his doing many innovative things, including the unmentioned but worthy work on foreign exchange rate models.

Also, I did not put words in the mouth of the committee, other than pointing out specifically that "the variant of Chamberlin's monopolistic competition model" as Glaeser so inelegantly put it, was the Dixit-Stiglitz model. It was the committee that clearly emphasized his work on both trade and location theory, which uses this model.

I was not going to put a comment on this here, as I had put several elsewhere and Mark Thoma was clearly very enthusiastic and putting up all these posts about it. Frankly, it was more annoyance with Glaeser and his inane and overweening inaccuracies that led me to do it, not annoyance with Krugman so much, whose work I admire.

Now, hari is all bent out that somehow I should not dare to comment on all this, presumably because I am an ignorant upstart, or whatever, compared to the "Nobel Price [sic] committee." Sorry, hari, but as seems to be often the case you do not know what you are talking about. The reason I felt I had to go onto the blogs and make clear my position is because I am the author of a book review of Krugman's _Development, Geography, and Economic Theory_, which is about his "new economic geography" work, my review being published in the Journal of Economic Behavior and Organization, Dec. 1996, vol. 31, pp. 45-454.

This review made some of the points I made above about Krugman not citing this large literature from the 1980s, only far more harshly than I have here or on any other blog. That review has been read by many people, and I knew some would be quoting some of its more pungent lines, which was indeed the case yesterday on some blogs. So, I had to clarify where I stood without repeating any of that harsher stuff on a day in which Paul Krugman was being honored. Just to add to this, I sent him a copy of the review prior to publication, and he never even acknowledged receipt. However, everyone I have ever spoken to about the review has praised it, and no one has ever found anything wrong or incorrect in it, with most enjoying the strong language in it, some of which was quoted in the blogosphere yesterday. Furthermore, its rate of downloads has been steadily rising.

So, what is really the issue here. The issue is whether or not Krugman ever read any of this literature that he ignored, and even denied the existence of by claiming to be the first to mathematically model agglomeration economies in an, as Glaeser sycophantically put it, "internally consistent mathematical framework." I made it clear in certain posts yesterday that I do not know if he did read any of this literature (or if Glaeser has), which would take them both off the hook from some of my sharper remarks in my book review if they have not (although it leaves them somewhat more ignorant than I am if that is indeed the case).

So, if you want to read a summary of this literature, check out Chapters 9-11 of my 1991 book from Kluwer, _From Catastrophe to Chaos: A General Theory of Economic Discontinuities_. My book review lists some of the more important items that Krugman (and Glaeser) ignored.

I list two here, just so folks know what we are talking about, one of which was in a prominent economics journal.

Y.Y. Papageorgiou and T.R. Smith, 1983. "Agglomeration as Local Instability of Spatially Uniform Steady States." Econometrica, vol. 51, pp. 1109-1119.

And, more importantly for reasons I shall not explain further here:

Wolfgang Weidlich and Gunter Haag, 1987. "A Dynamic Phase Transition Model for Spatial Agglomeration Processes." Journal of Regional Science, vol. 27, pp. 529-569.

[Barkley Rosser]

Just to be clear on those last references. None of them use the Dixit-Stiglitz model, which is not the only possible game in town for having an "internally consistent mathematical framework" for modeling agglomeration economies in spatial models.

[Barkley Rosser]

Oh, and for the record, the paper by Krugman on location theory cited in the Prize was published in 1991, after these papers. Fujita's paper applying Dixit-Stiglitz to location theory appeared in 1988, also after the papers I listed above. Krugman's work on this in trade appeared earlier.