'Deconstructing' Hostility to Economic Ideas

Paul Krugman, in a commentary written ten years ago, discusses the division between the humanist/literary and mathematical/scientific visions of the world, a division that "is a hidden but powerful force confusing and garbling the public discussion of many issues" in economics:

The implausible pundits, Paul Krugman, May 1996: Late in the summer of 1992 I received the page proofs of an article I had written for a liberal intellectual American magazine. I phoned in a few corrections; only one of them was really important. Editors always change your title - and the new title was snappier than my own pedestrian proposal. But the editor had added a subtitle that horrified me. It was, I thought, utterly inappropriate for a plainly written article that was distinctly populist in its implications. I pleaded with him to drop the subtitle, but when the article came out there it was: 'The rich, the right, and the facts: deconstructing the income distribution controversy'.

'Deconstructing'? Why would an editor (who, incidentally, regarded himself as a champion of ordinary working Americans) insist on using a buzzword that means nothing to most people and alienates most of the rest? Was he not gratuitously playing into the hands of his (and my) political opponents? I was baffled. But I now think I understand his motives - and I believe that this seemingly trivial incident offers an insight into a hidden conflict that lies behind many of the confusions that bedevil the public discussion of economics. Every economist who ventures beyond the confines of the academic world gets used to facing a certain amount of hostility. Some of that hostility comes from people who hold strong political views and do not want them challenged; some of it comes from people who want more from economists than they can deliver - easy answers to hard problems, accurate predictions of the inherently unpredictable. What I have gradually come to realise, however, is that there is an extra reason why certain people, namely literary intellectuals, are hostile to economists. Their hostility is not so much political as cultural.

My point is a familiar one: that our society remains divided between C. P. Snow's two cultures; or perhaps to put it differently, there is a continuing struggle between two ideas of what it means to be an intellectual. One culture is humanist and literary; the other mathematical and scientific. The editor of my article, I now believe, was perhaps unconsciously using the language of critical theory as a way of declaring his allegiance in that struggle; he was saying: 'I may write about quantitative stuff like GDP and real wages, but my sensibility is literary.'

In this particular case, the harm done by the culture war was minor: the audience for my article was probably reduced only marginally. But I have become convinced that the divide between two kinds of intellectuals - those who feel comfortable with a more or less mathematical approach to the world and those who do not - is a hidden but powerful force confusing and garbling the public discussion of many issues.

In a way, economics can be regarded as one of the humanities. Like history or sociology, it is concerned with human beings and what they do. And it is therefore a subject in which many humanist intellectuals are interested. But, as a discipline, economics is firmly on the mathematical side of the great divide. It is, indeed, a field in which ideas are mainly expressed in the form of mathematical models. This means that humanist intellectuals, even when they are deeply interested in economic affairs, generally find what mainstream economists have to say repellent if not incomprehensible. And because such intellectuals are uncomfortable with the way economists think, they systematically favour economic thinkers and ideas that most economists, with good reason, regard as unworthy of serious attention.

The result is that the wider public debate about economics, a debate that is largely filtered through publications edited by people who are or would like to be literary intellectuals, is deeply distorted: facts and concepts that research has established beyond a reasonable doubt are rejected or ignored, while views that are flatly wrong but that appeal to a literary imagination remain stubbornly in circulation. And these misguided views directly shape debates over real economic policy.

Consider The Atlantic Monthly - one of America's most influential intellectual magazines, and one with more coverage of economic affairs than most of its rivals. I recently did a quick review of The Atlantic's articles from mid-1993 to date. Over that period, there were by my count 16 articles dealing mainly with economics, about one every other month. Eight dealt with international trade (a surprisingly high proportion for the United States, where imports are only 13 per cent of national income); every one of those articles, and most of the others, expressed views that no serious researcher would support, and explicitly denigrated the economics profession as a whole. Admittedly, this record is at least partly a reflection of the personal preferences of The Atlantic's editors. Still, a similar hostility to economics may be found in many other publications.

Of course, one might argue that this hostility is a mark, not of the prejudices of intellectuals, but of their perceptiveness. Maybe economics really is a 'failed profession', and literary intellectuals just happen to be clever enough to realise it. But this explanation does not survive a close inspection: whenever one looks at an issue on which the views of mainstream economists and those of the economic thinkers favoured by the literati diverge, one finds that the opposition to economic orthodoxy rests not on a reasoned critique but on a failure to understand basic concepts.

Nowhere is this more apparent than in the discussion of international trade. Free trade is an idea intellectuals love to hate. As I said earlier, The Atlantic has devoted half of its economics articles in recent years to international trade - a far higher proportion than the intrinsic importance of the subject can justify. And without exception these articles conclude that the economists' case for free trade is all wrong. Similar attacks on conventional views of trade have appeared in virtually every intellectual magazine in the English-speaking world.

What is it about mainstream economic views on international trade in particular that inspires such hostility? It cannot be that these views are particularly stupid: the modern theory of international trade is nothing if not subtle, and its distinguished lineage includes some pretty impressive intellectuals (David Hume, John Stuart Mill).

It also cannot be that conventional trade theory is excessively dogmatic: on the contrary, economists have long recognised a variety of qualifications and exceptions to the classical case for free trade. So why not emphasise these qualifications, rather than attack the whole intellectual structure built up over the past two centuries?

The answer, I would argue, is that what literary intellectuals dislike about the conventional economic analysis of trade is that it is necessarily expressed in a language they do not understand or want to understand. For someone who is comfortable with even low-level mathematical thinking, the classical analysis of international trade as John Stuart Mill developed it can seem trivially simple. But to get that analysis across without the maths - to explain the basic story entirely in plain English - is, as seven generations of economists can attest, nearly impossible.

Simple as they are, the basic concepts of conventional international trade theory turn out to be as inaccessible to the anti-mathematical intellectual as those of quantum mechanics. And so the same ideas that seem clear, beautiful and compelling to most economists seem like obscure mumbo-jumbo to many highly intelligent people.

Conversely, ideas that seem evidently absurd to mainstream economists can seem plausible to many other intellectuals. For example, any economist who picked up Sir James Goldsmith's manifesto The Trap, which has received attention in several magazines and newspapers, quickly realised that Goldsmith's ideas about the world economy were not merely open to question, but the very same classic confusions that have been the source of exam questions on a thousand undergraduate midterms.

And now we come to my final point: the peculiar way in which the hostility of humanist intellectuals to mainstream economics has ended up giving aid and comfort to some rightwing politicians. For Goldsmith is not just a billionaire author; he is also a leading voice in a political movement that has recently begun to flex its muscles on both sides of the Atlantic. There is a new kind of rightwinger abroad in the land, one who mixes harsh social conservatism with protectionism and nativism. Where does this ideology come from? A few months ago, when the initial electoral successes of Patrick Buchanan had stunned America's political commentators, The Wall Street Journal asked the candidate about the sources of his ideas. Given the belligerent populism of his campaign, one might have expected Buchanan to disclaim any intellectual influence, to insist that he was simply preaching native common sense. But instead he offered an extensive intellectual pedigree for his ideas, citing not only modern authors like Goldsmith but 19th-century predecessors. In particular, he cited the protectionist doctrines of Friedrich List.

How did a late 20th-century American populist become a disciple of a confused early 19th-century German economist? Well, Buchanan did not discover List on his own: he was led there by the books and articles of James Fallows, the Washington editor of, yes, The Atlantic, who has in recent years attempted to promote List as an iconic figure, a sort of anti-Adam Smith. In short, here is a case in which the hostility of a literary intellectual to conventional economics has helped to provide an intellectual gloss to ideas that every educated citizen should instantly recognise as nonsense.

It would be silly to suppose that we can bring the war between the two cultures to an end. After all, that conflict has raged for at least two centuries. But it is possible for the educated individual to become aware of what is going on; to recognise that the negative things intellectuals say about economics may have less to do with content than with culture and style.

Comments

[Richard]

I'm willing to give Krugman his point - that many do not understand some of the subtlety of economics - without granting him the larger argument, to wit:

"But, as a discipline, economics is firmly on the mathematical side of the great divide."

None of classic economics - Ricardo or Adam Smith - used econometrics, because it was not a tool that was available. They used philosophical argument, like the rest of the social sciences.

And this is what economics has in common with psychology, sociology and other social sciences: there is a mathematical tool set that is powerful, but the fundamental ideas are conceptual and not in need of fancy mathematics. Psychology gave us factor analysis and item analysis for reviewing psychological inventories, but the mathematics simply provide tools: the technical aspects are not fundamentally needed to understand the arguments.

Marx had perhaps one of the best insights when he said "capitalism converts all qualities to quantities". I have a strong statistics background and I like nothing better then a well-rendered technical paper, but the technical aspects are just a tool, and the meat of the argument relies upon the technical aspects only for support to the argument: it is never the argument in and of itself.

[Jim Harrison]

A lot of us are skeptical about economics, not because we can't fathom the math but because it often seems inapplicable to real issues. I used to edit applied math books and was frequently told that economists were much fussier about mathematical niceties than physicists. The physicists could afford to be cavalier because they could rely on experiments and observations to correct their errors whereas the theorems of the economists were always several fudge factors away from empirical confirmation and had to rely on their formal elegance to convince anybody.

Lots of us have read Ricardo, for example; but what he postulates as free trade is rather different than anything that happens in the real world. The problem isn't with the reasoning, but with finding any situation that really satisfies the hypotheses of the theorems.

By the way, I can suggest another reason not to use "deconstruction" in a title. Even in the humanities, the word, which once had an exquisitely specific meaning in the early works of Derrida, has long since ceased to have an identifiable signification. It simply functions as a sort of lexical Masonic handshake that advertises membership in a club. In other words, it has the same function as the largely ornamental differential equations one often encounters in economics papers.

[calmo]

There are tons of engineering types who blog here Jim.
Brave post.
"largely ornamental differential equations" is so eloquent.
And provocative.
I'd switch names now if I were you and hide while they group and storm. Pray for some disarray while they fume about the details and other familiar (only to them) connotations of "deconstruction". You can count on my silent support.
Then, late in the thread, with only a couple of survivors, swat them again with a 2nd lexical Masonic reference.
The handshake took me out early, but then I didn't get past Kreyszig.

[reason]

Good comments here and to the point.

I'm not sure why Krugman thinks that economics leaves people cold because it is too mathematical.

What I object to is the partial analysis, especially the tendency to comparitive statics. I would like to see proper dynamic adjustment models, not general equilibrium theory. With financial and real properly linked, externalities properly valued and a realistic view of the value of unpaid work and leisure. And quantitively, it is particularly difficult, because the theory can change behaviour (rational expectations anyone?).

Did Krugman never hear of the can opener joke?

[reason]

In fairness, I will say that some people (especially those who extrapolate from a single firm to an economy) are even more partial in their analysis than most economists. But to my way of thinking, it is not so much what language is used that is the issue, but the content of the message.

[Tom Schofield]

Krugman's essay is an illustration of why those of us who lack a mathematical facility are suspicious of economics: the fuzziest concepts and the most dubious ideas, once reduced to a mathematical formulation, are regarded as immutable "facts". It is reminiscent of the Church using Latin as a means of insuring that only the priests could truly know and interpret the Gospel. Compare also econometricians to the hedge fund managers who advertise themselves as "quantitative analysts". Remember Long Term Capital and be grateful that economists don't usually get to play with real money.

[Richard]

An additional note: a social science like psychology is arguably more of a science than economics. Correlations are frequently stronger, and unlike economics you can generally do experiments. But there is little handwringing that psychology is misunderstood because readers have an inadequate mathematical knowledge; on the contrary, there is the healthy realization that results in psychology are not just the outcome of innate biology but also of any number of social influences. The human being is an enormously flexible creature, and any experiments must also relect the uniqueness of time and place and history - as is true, also, of economics.

[Richard]

If one truly wants a deconstructionist economics, look to Bataille: he conjectures not infinite wants and limited means, but excessive productivity. His claim is that societies are overly productive, creating excess wealth and an "accursed share" which must somehow be destroyed. Hence the history of accumulation and destruction, hence the excessive feasts of some tribes.

Without the "motor" of infinite wants and limited means, much of economic theory breaks down. But economics as a study of distribution within and through a society remains.

[kharris]

Perhaps what Krugman should have said to his editor was that his commonplace misuse of "deconstruct" was objectionable. Most of our middle-brow elites have adopted "deconstruct" to mean analyze in the same way than everything that hasn't stopped happening has to have "ongoing" slapped next to it. Derrida may have had something undefinable in mind when he got us using the word "deconstruct", but he was clear that he did not mean merely analysis or critique. Fine for the newly-minted BA to grab whatever word is making the rounds, but you'd hope an editor for the Atlantic would have a greater devotion to precision and accuracy.

Tom,

Care to show us how Krugman's essay demonstrates what you claim it demostrates? Saying that (modern) economics is in the mathematical camp is not a demonstration that "the fuzziest concepts and the most dubious ideas, once reduced to a mathematical formulation, are regarded as immutable "facts"." We got us a nonsequitor here. And while I will grant you that shallow-minded disputants from any discipline are prone to wave a few of their discipline's tools around and then claim to have won the argument, I don't think economists are prone to let their colleagues get away with such behavior for long.

[Bruce Wilder]

A couple of weeks ago, Bill Maher had on Gloria Steinem, who observed, there are two kinds people in the world, those who try to divide the world between two kinds of people, and those, who don't.

Me, I like economics and I think Derrida's ideas about deconstruction as a method of criticism are powerful and interesting. The title of this post made me think of "Zen and the Art of Motorcycle Maintenance", a book I thoroughly enjoyed.

Krugman makes an unlikely left-wing polemicist, just as Derrida is not usually thought of a UC-Irvine professor. (Orange County is not so far from the Sorbonne, after all.) So, Buchanan took up a reading list from James Fallows, former Carter speechwriter? Will wonders never cease?

[anne]

http://www.nytimes.com/2004/10/14/opinion/14taylor.html?ex=1255492800&en=6f805b298f0aa5e7&ei=5090&partner=rssuserland

October 14, 2004

What Derrida Really Meant
By MARK C. TAYLOR

[Jacques] Derrida's name is most closely associated with the often cited but rarely understood term "deconstruction." Initially formulated to define a strategy for interpreting sophisticated written and visual works, deconstruction has entered everyday language. When responsibly understood, the implications of deconstruction are quite different from the misleading clichés often used to describe a process of dismantling or taking things apart. The guiding insight of deconstruction is that every structure - be it literary, psychological, social, economic, political or religious - that organizes our experience is constituted and maintained through acts of exclusion. In the process of creating something, something else inevitably gets left out.

These exclusive structures can become repressive - and that repression comes with consequences. In a manner reminiscent of Freud, Mr. Derrida insists that what is repressed does not disappear but always returns to unsettle every construction, no matter how secure it seems....

[anne]

Actually, use of the term deconstruction for the essay in question, which I just read, strikes me as completely appropriate. John Kenneth Galbraith repeatedly, and almost alone, tried to deconstruct economic arguments for general readers, which is what Paul Krugman has come to increasingly do and what deconstruction is all about.

[anne]

Jacques Derrida, who for reasons that have never made sense to me was a purposely obscure writer. Deconstruction is meant to be an unpretentious method of analysis, and there was no reason for Derrida to be in any way an obscure writer having formed what is meant to be a general social tool of broad based, inclusive, analysis. Paul Krugman's columns in the New York Times are completely accessible deconstructions, whichy is why they are found so ominous by conservatives who would prefer what Krugman is about never be understood.

[JRossi]

Part of the problem is hostility to math, but I think a bigger problem is politics and world-views. Taking international trade as an example, economists generally think that it's a good thing overall. And it is--unless you dislike foreigners or don't care about them, or unless you care more about those who lose their jobs here than those here who benefit from lower prices and who will gain from new jobs. And then there's the question of the non-human component--natural environments, non-human species, etc. These are complicated issues. And I seem to recall reading a few years ago that Krugman himself was puzzling his way through how economies affect the environment and had not yet come up with firm conclusions.

[anne]

J Rossi

Actually this gives us a useful sense of what deconstruction means in terms of looking to a text from multi-dimensional perspectives. John Maynard Keynes was especially effective in taking such perspectives, but professional analysis can be strikinhgly narrow in any field. Paul Krugman is happily becoming increasingly broad.

[Lee A. Arnold]

A real problem for economics is that many of the systems that it impacts, lie strictly outside its comprehension. In fact it doesn't do all that well with some systemic problems WITHIN its purview: the prices clearing for general equilibrium are beyond computation, for example, and the syllogisms rely upon an existence theorem. For outer systems, systems bigger than the economy -- social systems, cultural systems, ecosystems, climate systems, e.g. -- economics is inapplicable, and relies upon revealed preference and utility. The trouble is that the progressive division of labor results in an atomization of the intellect, as Adam Smith pointed out. This is the flipside of the fact that more knowledge is being generated, as a whole. So you don't have time to know it all, in fact you can't even prove that you know all you really need to! So then Hayekian idea, that markets will reintegrate all the information from its actors into a functioning whole, is nonsense, if they don't know enough to begin with. Yet this is precisely the state of people's knowledge. The only alternative is institutions, and institutional economics is unmathematicized. In fact we revert to politics, changeably dealing with the vagaries of complex systems, as it should be. This is very difficult for 20-yr-old college students, who want final answers, or final methods to find answers.

[James Killus]

Speaking as a mathematical modeler (photochemical kinetics in one incarnation), I'll observe that the greatest problem with modeling in general isn't so much the math part, but rather the linkage of the model to the world. A perfect model of a ship will not be much use if you're flying an airplane, in other words.

This, it seems to me, is the flip side of the argument: the beauty of a mathematical model can blind someone to the defects of application.

I ran across a very interesting essay a little while back that spoke to the free trade issue:

http://thelookingglass.blogspot.com/2005/01/and-now-another-few-stupid-questions.html

Essentially, the writer notes that the doctrine of comparative advantage is a stationary analysis. It assumes that all cost matrices stay the same, so each country's specialization will also remain fixed. Thus, a country that is primarily agrarian will remain agrarian forever, as there is no way for it to alter its relative costs and efficiencies within the comparative advantage framework.

Of course there _are_ ways of changing those factors, but many of them violate the doctrine of free trade. What then? The math is correct, but the policy prescription is pernicious.

There are numerous examples of such leaps from an incomplete analysis (and all analyses are incomplete) to a fixed doctrine. One such example is the minimum wage debate that frequently crops up here, with a vanilla analysis showing that an increase in minimum wage would be expected to slightly increase unemployment but nevertheless increase income to minimum wage employees as a whole. The last feature is frequently overlooked (or deliberately unmentioned) by economists writing about the situatiion.

Just incidentally, the phenomenon of (for example) conservative economists decrying the increase in unemployment from minimum wage hikes, without admitting that it benefits the class of minimum wage workers, is a phenomenon that is precisely contained within the rubric of "deconstruction," which attempts to decode hidden agendas.

That said, James Fallows is nevertheless a twit.

[2slugbaits]

kharris said:

>>>>
Care to show us how Krugman's essay demonstrates what you claim it demostrates? Saying that (modern) economics is in the mathematical camp is not a demonstration that "the fuzziest concepts and the most dubious ideas, once reduced to a mathematical formulation, are regarded as immutable "facts"."
>>>>

While I'm generally sympathetic to the gist of Krugman's argument, I do think that there is a tension between modern mathematical economics and modern econometrics. Just to throw out a couple of examples:

1. A lot economic data that traditional textbook economics treats as trend stationary have been shown to be difference stationary. That raises some pretty serious challenges to textbook economics; e.g., the long run effect of stochastic supply shocks.

2. About 25 years ago Sims caused some panic in academia by outlining what was an atheoretical economics. Today's structural VAR models are a compromise between the purely atheoretical VAR models and traditional structural OLS models.

3. Try coming up with an economically intuitive explanation of cointegration; but yet finding a new cointegration test is the surest way to get published in Econometrica.

[reason]

James Killus...
What a brilliant post. My hat off (if I had one on that is)!

[reason]

May I suggest a way forward here.

1. Stop trying to pretend that models represent reality. Don't stop making them, or making them more complex, but don't explicitly say they represent reality. They are engineering experiments.
2. Recognise that time series econometrics is probably useless, because the world keeps changing
3. Do start building more complex models and using computer simulations to test whether the model performs something like the real world.
4. Iterate 3 to get the models better.

i.e. Economics should try to be more like science.