How to buy a good painting

OKONOMOS

People usually have a fairly good idea of the worth of what they are buying. Sometimes, however, when they buy a painting in the hope that it will become worth a fortune one day, it doesn't always work out that way.

If they are lucky or unnaturally prescient, their investment pays off. But often it doesn't. So should there not be some scientific way of telling whether a painting will appreciate or just become worthless junk? Yes, there should, says David W Galenson of the University of Chicago in a recent paper.*

He has, without claiming either to do so or saying that it will work for everyone, everywhere, or indeed anyone at all, obligingly worked out a method. They key thing, he says, is to know that "the primary source of genuine importance in art is innovation". He defines innovation as new practices that change the practices of successors. By definition, then, "important works of art embody important innovations; the most important works of art are those which introduce these innovations".

Having told you what to look for in a painting, he then poses the next problem: how do you tell exaggerated claims about a painter from genuine innovation, especially since in the "modern era an even greater premium is placed striking innovations"? Galenson says that in the long run the only measure of true innovation is whether or not a painting hangs on the walls of museums. The real test, according to him is "" reasonably enough "" whether it "becomes the subject of study by later generations of artists and scholars of art. Short-run success, in the form of critical acclaim or lucrative sales of an active artist's work, often does not translate into long-run success".

The history of art is replete with stories of this sort. The paintings of Van Gogh and Gauguin, to name just two, came to acquire value only after their deaths. On the other hand, critics regarded the works of their contemporaries William Bouguereau and Ernest Meissonier very well during their lifetime only to slump precipitously subsequently.

According to Galenson, there are fundamentally only two types of artistic innovation. "What distinguishes them is the method by which they are produced. One of these types can be called aesthetically motivated experimentation, the other conceptual execution."

In plain English, this means that the world of artists can be divided into those who "proceed tentatively and incrementally" and those who make a sudden leap. The former adopt a process of trial and error.

"Their innovations emerge gradually: they are not declared in any single work, but, instead, appear piecemeal in a large body of work. Experimental artists typically build their skills and understanding over the course of their careers."

In contrast is the other lot. Their "innovations typically appear suddenly, as a new idea produces a result quite different not only from other artists' work but also from the artist's own previous work".

Understanding this difference can help in determining the relationship between age and artistic innovation, says Galenson. Thus, "the long periods of trial and error often required for important experimental innovations mean that they rarely occur early in an artist's career.

Conceptual innovations, which are made more quickly, can occur at any age." In other words, "conceptual innovators should generally produce their most important work earlier in their careers than experimental innovators." This is the litmus test.

It leads Galenson to "the prediction that conceptual innovators should produce their most valuable work earlier in their careers than their experimental counterparts." He then goes on to test this hypothesis.

His paper shows that Jackson Pollock, Mark Rothko, and the other leading Abstract Expressionists, who were experimental innovators, produced their best work considerably later in their careers than did Jasper Johns, Andy Warhol, and the other leading conceptual innovators of the generation that followed them.

He considered over 1,600 paintings and water colours by ten artists that were sold at auctions during 1970-97. The relationship between the auction value and age at the date of the work's execution was estimated by multiple regression analysis and each painter's implied age at peak value was then estimated.

The econometric gyrations soon reveal that "the conceptual artists reduced their highest priced work at younger ages than did the experimental artists." But could it be that auction prices only reveal the tastes of wealthy but unsophisticated collectors, instead of reflecting the judgements of sophisticated art scholars?

This, says Galenson, can be tested and proceeds to do so. He finds that "the critical evaluation of these artists' careers agrees quite closely with the evaluation of the market in a majority of cases, and does not strongly disagree even in the cases for which the two differ".

Thus, art scholars and collectors generally agree on when an artist produced his best work. The Life Cycles of Modern Artists, NBER Working Paper No 8779.