I am an economist, but I study creativity, particularly in the arts. I began doing this research more than a decade ago, almost by accident. I was buying a small drawing by Sol LeWitt, when a friend who worked for LeWitt told me it was overpriced: they were selling new ones the same size for less. But this one was ten years old; shouldn't it be worth more -- or less -- than a new one? After conversations with the dealer and my friend, it became clear to me that neither had a conception of an artist's age-price profile -- how the price of a painting is related to the artist's age when the work was made. As a quantitative economic historian, I have spent much of my career estimating the relationship between age and productivity for large populations of workers in the past, from indentured servants and slaves in the American colonies to immigrants in 19th-century cities. I was intrigued by the question of what this would look like for artists: when in their lives do painters produce their most valuable work?
I began estimating the age-price relationship for modern painters, using large samples of observations drawn from auction records. The results were among the most surprising I have ever seen. Economists have found that workers' productivity generally rises with age, throughout most or all the life cycle: experience makes workers better at their jobs. For many great modern artists, I found just this result. So for example Cézanne made the paintings that today are most highly priced at the end of his life, at the age of 67; Kandinsky made his most expensive paintings at 52, and Rothko his at 54. These artists fitted my expectations. The surprising thing, however, was that for many great modern painters the results were radically different: their most valuable work was produced when they were very young, and their age-price profiles declined for most of their lives. Remarkably, Picasso made his most valuable work when he was just 26 years old; Jasper Johns made his at 27, and Frank Stella his at 24.
These results were startling, but what did they mean? Did these ages at peak value tell us something significant about these artists' careers, or were they merely curiosities produced by a luxury market many art critics dismiss as irrational? I did what economic historians are trained to do -- to look for other sources, independent of the one I had already used, that would allow me to measure the same relationship. I created two data sets that effectively surveyed the judgments of art experts on this question. I made one by tabulating all the illustrations of the same artists' work that were reproduced in all available textbooks of art history, and the second by tabulating all the paintings by each of these artists that had been included in retrospective exhibitions of their work.
The two new sources agreed closely with the age-price profiles. Thus textbook reproductions heavily favored late Cézannes, Kandinskys, and Rothkos over their early works, just as they heavily favored early Picassos, Johns, and Stellas. The retrospectives did the same, including larger numbers of late Cézannes, Kandinskys, and Rothkos, and of early Picassos, Johns, and Stellas. And the degree of agreement was startling: so for example both American and French textbooks were most likely to reproduce paintings Cézanne executed at 67, and Picasso at 26; and the most recent Cézanne retrospective included more paintings he made at 67 than at any other age, while the latest Picasso exhibition had the greatest number of his works from the age of 26.
The close agreement of these very different sources convinced me that the puzzle was real: why have some great artists produced their most important work late in their lives, whereas others have been greatest early in their lives? To be continued: I will explain my answer to this in my next post.