All 4 books by Edward Tufte now in
paperback editions, $100 for all 4
Visual Display of Quantitative Information
Envisioning Information
Visual Explanations
Beautiful Evidence
Paper/printing = original clothbound books.
Only available through ET's Graphics Press:
catalog + shopping cart
All 4 clothbound books, autographed by the author $150
catalog + shopping cart
Edward Tufte e-books
Immediate download to any computer:
Visual and Statistical Thinking $2
The Cognitive Style of Powerpoint $2
Seeing Around + Feynman Diagrams $2
Data Analysis for Politics and Policy $2
catalog + shopping cart
Edward Tufte one-day course,
Presenting Data and Information
Arlington, June 5, 6
Bethesda, June 8
Seattle, July 11, 12
Portland, July 14
The economisting of art

Here is brief excerpt from a draft chapter for Beautiful Evidence. The complete chapter is posted in the thread on "Corrupt Techniques in Evidence Presentations" in the New Answers threads.

-- Edward Tufte


Response to Punning for Relevance: A Case Study of Economisting

With regard to Las Meninas: there are a multiplicity of readings of the painting, as is the case for many great pieces. Picasso painted 45 paintings based on the Velázquez.

 Pablo Picasso, Las Meninas, after Velázquez (August 17, 1957)                                                               Laurent de Brunhoff, Babar's Museum of Art (2003), p.17

-- Edward Tufte


Response to Punning for Relevance: A Case Study of Economisting

The Barr comment was with a wink, thinking that chart might end up looking like the Reinhardt.

Of interest might be the course of auction prices as measured from the year the work was created, rather than against the age of the artist when created. This could be tracked along with points of interest (establishment or dissolution of significant collections, retrospective exhibitions, monographs, incidents of public notoriety, economic situations ...). What this would really tell you, of course, is what interested the auction-going public at any given time. While the ebb and flow of popular taste (and thus buying habits) is worthy of examination, I can't see auction prices ever having an effect on the working practices of an artist, they lag too far behind. Gallery prices, now that's another story...

A few points related to rhetorical practice:

Galenson confounds issues of aesthetic merit with market prices by using the term "most valuable" to describe his measurement of worth; in this context he's describing cost but to most readers this means excellence.

He also mixes "units of measure," so to speak, by writing of the valuable work of Cezanne "late in [his] career" while giving the same measure of worth to Gaugin and van Gogh "at an early age." Since all three (even Cezanne at 67) died young, their most expensive work could all be said to have been done late in their career.

An example of the broad brush: that which paints the "late in their career" production of the very individual contributions of the Abstract Expressionists: Pollock, having also died young and thus with a serious career lasting 15 years or so, could be said to have made his "most valuable" paintings in mid-career, Rothko likewise, deKooning, given the great length of his working life, produced his crowd pleasers early-on.

The "pre-planned escape": adding "almost invariably" to "late in their careers."

At least the Modern is reopening and we can go see a lot of this stuff again.

-- Steve Sprague (email)


Response to Punning for Relevance: A Case Study of Economisting

A small suggestion - perhaps instead of economist-ing, you could alter it to ekonomist-ing. This allows for the noun form to indicate ekonomists (specific or general) whom engage in ekonomisting, as opposed to all economists.* It further plays on the "k" substitution already familiar (to much of Amerika) indicating sarcasm.

There seems to be no mention of auction theory (winners curse, private versus public value). Also, since a single piece of art changes hands rarely, I wonder if his sample was large enough to do the unbalanced panel series he would need. (that is a very specific modeling question).

Have you read the original paper this is based on http://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=147435 ? I am off-network, so I cannot reach it myself.

*Full disclosure: I am an economist who fully enjoys your work, but feel somewhat attacked as your analysis seems inclusive of all economists. Is your intention is constrained to loose use of terms in economics that are loaded in conventional senses (value, price, freedom, equity), then I wholeheartedly support it — we need it.

If your intention _is_ to attack economics on the whole, the statement about the "k" is superfluous. If you are at the prelude to refuting economics on the whole, a good place to start is Dasgupta's discussion of what economists analyze, and why:

"What Do Economists Analyse: Values or Facts?", paper presented at the conference on The Place of Value in a World of Facts, London School of Economics, October 2003. Mimeo. Faculty of Economics, University of Cambridge, 2003.

http://www.econ.cam.ac.uk/faculty/dasgupta/factvalue.pdf

Looking forward to the book,

Cheers,

-- Dan Egan (email)


Response to Punning for Relevance: A Case Study of Economisting

Dan Egan's comment is very thoughtful and goes to several important issues.

A couple of months ago I downloaded and read 2 of the NBER working papers (only $5 each) to find out what the models actually were (the test against the null hyothesis was that all 5 regression coefficients on t, t-squared, up to t to the fifth were simultaneously equal to zero, the ridiculousness of which inevitably leads to statistical significance in the age/price multiple regressions). In the 2 papers and in the book, I saw nothing about auction theory. And how can auction theory justify the punning or the poor practices in presenting statistical evidence?

As in the case of cladogram in the thead on causal arrows, what I am concerned about are the external, universal standards of statistical evidence not the internal standards of a field itself, except insofar as those internal standards (as manifested in the work of notable professors and notable publishers) are more lax than external standards. Within-discipline appeals don' t make it.

I once wrote a book on politics and macroeconomics, Political Control of the Economy (Princeton, 1978). Back then my first thoughts about economisting arose when reading an AER article by George Stigler in the mid-1970s about macroeconomic conditions and election outcomes. In that paper, it seemed to me that Stigler was claiming that the economy didn't affect election outcomes because economic theory said it couldn't. Some perhaps tendentious empirical tests then followed in that article. I was later a minor consultant on the plaintiff side of an anti-trust case in Chicago, with George Stigler and his consulting firm on the other. There was the same problem with their approach: the plaintiff just couldn't be right because economic theory said so. This attitude, combined with a certain loosenss with regard to empirical evidence and the view that evidence didn't matter all that much if you had theory and a Nobel-prize winner on your side, failed to impress the jury and produced what was at the time one of the largest anti-trust awards to a plaintiff ever.

I probably first heard the word "economisting" as a graduate student, when the word might have been used by Robert Dahl in his seminar at Yale in the 1960s. I later wrote about the economisting style, not using the word, in the concluding chapter of Political Control of the Economy.

The spelling of economisting is fine. The preferred pronunciation (accents on the "con" and the "mist") indicates the spirit of the word.

To put the matter way too broadly, I think the rise of behavioral economics (where evidence trumps theory) is a revolt against economisting (where theory trumps evidence). Or, Galileo not the Church.

-- Edward Tufte


Response to Punning for Relevance: A Case Study of Economisting

Very interesting - I will write up a full reply when I have a few more minutes, but a couple of quick responses:

1) Regarding the polynomial terms - that was a trick that I learned in my multivariate calculus class - how to lie with regressions 101. I would be extremely surprised if that got by a good reviewer.

2) Auction theory doesn't support him, it undermines him further - google winners curse, or auction theory to see why. The first hit for me googling Winners Curse is a book by Richard Thaler, who is one of the pillars of behavioral economics - and one of my personal favorite writers.

3) Regarding ekonomisting/eConoMisting: I just realized this, but I hope I am not the only one who sees the irony of the title: Punning for Relevance

4) I agree with you heartily about behavioral economics, which is why it is what I research. Economics has quite a few self-fulfilling prophecies, most stemming from the rationality assumption. A good teacher of mine once commented that economic theory only returns consistent positive evidence when the sample is economists.

Cheers,

-- Dan Egan (email)


Response to Punning for Relevance: A Case Study of Economisting

On behavioral finance, mentioned earlier, see the superb collection of materials at http://introduction.behaviouralfinance.net/

-- Edward Tufte


Response to Punning for Relevance: The Economisting of Art and Art Auctions

It would be helpful to find other punning examples (in addition to the obvious one of "significance tests").

At a land-use hearing I once encountered the phrase "the highest and best use of the land," which I naively thought meant leaving the land alone and using the land for open space. What could be highest and best, after all? It quickly turned out that the phrase in fact meant the most intense development that the land and land-use regulations would tolerate. Another pun. Silly me.

-- Edward Tufte


Response to The Economisting of Art

A small point, perhaps, but should you identify the Professor Levine in the Weinberg quote? Or perhaps sympathetically excise the reference from the quotation?

And as a matter of fairness to the individual: a refutation is made to this person's thesis without really showing the argument in the first place. I agree with the point Weinberg and you are making, but it seems only fair, if a person is identified as authoring a contrary idea, to give him or her a fuller reading.

As a matter of fairness to others: you risk running all the professors Levine together (the person that came to my mind was Professor Neil Levine, Art and Architectural History at Harvard, though I have some doubt that he's the one being discussed.)

On this point, a similar thing happened to "paradigm" and "paradigm shift" as proffered by Thomas Kuhn in "The Structure of Scientific Revolutions." He penned a strong condemnation of any use of these terms other than in the realm of the philosophy and history of science in the second edition of his book.

-- Steve Sprague (email)


Response to The Economisting of Art

In a preface to his remarks on "Sokal's Hoax", Steven Weinberg notes that he is replying to a letter in the New York Review of Books by Professor George Levine of the Rutgers Center for the Critical Analysis of Contemporary Culture. A Google search reveals Levine's letter appeared in the NYRB, 43 (October 3, 1996).

-- Edward Tufte


Response to The Economisting of Art

I think that a major issue with Galenson's work is that a sound statistical approach is to state beforehand what it is that you are attempting to prove and how you think the relationship should be broadly specified. In the instances you cite, it appears to be that instead he has run a number of different regressions and then attempted to infer relationships based upon what he has found. This relates to your point on the amount of information contained by the parameters of a quintic polynomial. Galenson is attempting to argue different phases of an artist's work, their age, their productivity, etc. have an impact on the price of their paintings but in that case, these should be explicit variables rather than inferred from the results of a quintic polynomial fit.

-- Will Oswald (email)


Response to The Economisting of Art

Self-reported self-surprise at one's own results, as we saw above, has calibration problems. "Compared with what?" the skeptic might well ask.

Here is a variant: non-surprise as evidence. From the recent London Review of Books (20 Janurary 2005), p. 9, Colin Burrow's dismantling of Stephen Greenblatt's new book on Shakespeare:

"There is absolutely no evidence that Shakespeare met Campion, though Greenblat hints this might itself be a kind of evidence that it happened: '"Not surprisingly, Shakespeare never referred openly to Campion." "

Burrow's verb "hints" is just right. Note Greenblatt's suspicious adverbs, "suprisingly" and "openly". "Openly" hints, without evidence, that there may have been private references by Shakespeare.

The review is posted at

http://www.lrb.co.uk/v27/n02/burr01_.html The London Review of Books is a notable source for intelligent and often beautifully written reviews.

-- Edward Tufte


Response to The Economisting of Art

What I hope is the final draft is now posted at the top of this thread.

-- Edward Tufte


Response to The economisting of art

Robert Hughes published an excellent essay "Art and Money" in The New York Review of Books (1984). Here is part of Hughes' insightful account relevant to the economisting of art:

"The art market we have today did not pop up overnight. It was created by the great liquidity of late-twentieth-century wealth. Sell a block of shares, shift the money elsewhere. But liquids do not flow where you want them to unless you dig channels, and this patient hydraulic effort has been, since 1960 at least, one of the wonders of cultural engineering. The big project of the art market over the last twenty-five years has been to convince everyone that works of art, although they don't bear interest, offer such dramatic and consistent capital gains along with the intangible pleasures of ownership--what Berenson might have called "untactile values"--that they are worth investing large sums of money in.

"This creation of confidence, I sometimes think, is the cultural artifact of the last half of the twentieth century, far more striking than any given painting or sculpture. Its origins lie in the mid-1960s, and although it is hard to assign a single starting point to a cultural movement so diffuse and international in scope, I think of it as beginning with a curious enterprise called the Times-Sotheby Art Indexes, which created much interest in London and afterward in New York around 1966. These indexes were the brainchild of a public-relations man who had been hired by Peter Wilson, the chairman of Sotheby's, to spruce up the somewhat fuddy-duddy image of his house; and what they purported to give was reliable statistics on the price movements of all manner of works of art--seicento Bolognese drawings, netsuke, old master prints, nineteenth-century animalier bronzes, Chinese porcelain--showing, in an extremely generalized way, how everything was going up by 25 to 200 percent per year. They were short, undetailed, memorable, and embellished with graphs.

"Perhaps it was the graphs that did it. They gave these tendentious little essays the trustworthy look of the Times financial page. They objectified the hitherto dicey idea of art investment. They made it seem hardheaded and realistic to own art. From this modest beginning the idea ramified, and for the next ten years it was rare to open an airline magazine without finding yet another excited piece of hackwork pushing the idea of art investment. By 1980 the idea had become so familiar that it was no longer necessary to stress it, and the collector-as-investor dropped out of favor as a journalistic hero; even the dealers felt that such people should not be paraded too much, partly because it seemed a bit vulgar and partly, I would guess, because prices had already gone so high, and confidence in their continued ascent was so well implanted, that it was time to talk about eternal spiritual values again.

"This confidence feeds and is fed by a huge and complicated root system in scholarship, criticism, journalism, PR, tax deductions and museum policy. And it cannot be allowed to falter or lapse, because of the inherently irrational nature of art as a commodity. Art prices are determined by the meeting of real or induced scarcity with pure, irrational desire, and nothing is more manipulable than desire.

"The market is always converting works of art into passive fictions of eternity and immutability, of transcendent value for which no price may necessarily be too high. When the word priceless crops up, the haggling has only just begun. Hence the battered state of the word masterpiece, which used to mean a work that proved an artist's graduation into full professional skill, but now means an object whose aura and accumulated myth strike people blind temporarily and render their judgment timid. It refers more to myths of status than processes of comparison, and that kind of mythmaking is the seed of what New York dealer Ben Heller, in one of the great Freudian slips of recent art history, was heard to call "creative pricing."

"It is the element of fantasy in the art market, the sense that art prices are so weakly tied to more mundane kinds of economic activity, and that there is something neurotic about them, that gives them their odd lability. The art market can be set pitching and rolling by a single act, which is why it is so notoriously vulnerable to manipulation. A ring of three or four promoters can bid up the price of a dubious young star painter at auction, and although the New York art world may know what's going on, the collectors in Akron, Ohio, are not so likely to--all they see is the price that was, after all, publicly bid and duly paid, and is henceforth true."

-- Edward Tufte


Response to The economisting of art

A recent RAND Corporation report describes changes in the art market:

"At the same time that prices have reached headline-grabbing heights, the arts market has become increasingly like other asset markets. The value of an artist's work is determined not, as was traditionally the case, by the consensus of experts, but increasingly by a small number of affluent buyers who are drawn to purchase works for their potential investment value."

http://www.pnnonline.org/article.php?sid=6160&mode=thread&order=0&thold=0

-- Edward Tufte


This interesting Los Angeles Times account by Christopher Reynolds describes the intensely multivariate causes of art prices.

-- Edward Tufte


Dr. Tufte will be thrilled to hear that Malcolm Gladwell of the New Yorker has become infatuated with Dr. Galenson's theory and is currently making the Picasso vs. Cezanne analogy the subject of many lectures. Gladwell has explained to math teachers that fast learners are like Picasso while slow learners are like Cezanne; to his own fans that the Eagles were like Picasso, while Fleetwood Mac was like Cezanne; that American "Health-care system needs Cezanne, not Picasso or Michael Moore"); and that corporate R&D can be like either ("Is Your Company a Cezanne or a Picasso?")

-- Steve Sailer (email)


One of the ways in which the art market is different than other asset classes is that the seller of the work often has different reasons to sell rather than just price. For many galleries and artists, getting the work into the right collection is just as if not more, important than getting a particular price. Another way that the market is different is that the price of the work can change dramatically based on what the owner does with the work. When purchasing a share of stock the owner is mostly a passive participant. An owner of an artwork can do things that actively increase or decrease the value of the work they own, and the other works the artist produces.

In addition, like all markets the art market is subject to fraud and misinterpretation.

It seems that some artists in China are learning about the joys of the art market and are not happy about what they are learning.

http://www.nytimes.com/2008/05/07/arts/design/07coll.html?ref=design

-- Michael Joseph (email)


Art auction prices

http://online.wsj.com/article/SB122670620372529693.html

Comparisons between sold-prices and pre-sale estimate prices are distorted by the inclusion of the auction house commission in sold-prices.

-- Edward Tufte


A fun book review by Evgeny Morozov

"This is a world in which pundits are increasingly using the word "increasingly" whenever they feel too lazy to look up the actual statistics, "

http://www.tnr.com/article/books-and-arts/magazine/105703/the-naked-and-the-ted-khanna

-- Niels Olson (email)




Threads relevant to art history, especially cubism:


Threads relevant to evidence reasoning: