The Dealer and the Critic

For most of the nineteenth century, a painting career in France went through one door. The Salon was the official exhibition, juried by the Academy, and it was enormous: hundreds of thousands of visitors, thousands of works, reviews in every paper. Acceptance brought commissions, teaching posts and state purchases. Rejection meant you had not happened.

The pressure on the jury was therefore intense, and in 1863 it rejected so many submissions that the resulting complaints reached the emperor. Napoleon III ordered the rejected works shown separately, and the Salon des Refusés opened with Manet's Déjeuner sur l'herbe in it, to laughter. The lasting significance is structural rather than scandalous: it demonstrated that an audience existed for work the official system would not show.

What replaced the system was the dealer.

Paul Durand-Ruel's innovation was to buy an artist's output outright rather than take work on consignment. He bought hundreds of Monets, Pissarros, Sisleys and Degas, held them as stock, and mounted solo exhibitions, a format he effectively invented, to build the reputations of the painters whose canvases were in his warehouse.

Look at what that does to the incentives. When a dealer holds stock, the dealer's capital is tied to the artist's reputation rising. Promotion is no longer a favour; it is asset management. Friendly critics, catalogues, foreign exhibitions, careful placement of works in collections that confer status: all of it follows from who is carrying the risk.

This is the origin of the modern arrangement in which the value of art is substantially a story told about an artist, and it is not a conspiracy. It is what happens when someone buys inventory whose worth depends on how it is understood.

Durand-Ruel nearly went under twice. He was saved in 1886 by an exhibition in New York that sold, which is the direct reason so many of the best Impressionist pictures are in American museums. The canon everyone now treats as self-evident was a leveraged bet that came close to failing, and the alternative history in which it did is entirely plausible.

Two features of the market that grew from this are worth understanding, because they explain most of what reads as strangeness from outside.

First, the primary and secondary markets are different animals. A primary sale is the first one, from the artist through their gallery; a secondary sale is a resale between owners. Auction results are overwhelmingly secondary, and in most countries the artist receives nothing from them. European artists have a small resale royalty, the droit de suite; American artists do not. So the headline price for a living artist's work may have no connection to that artist's income.

Second, galleries choose their buyers. A work flipped at auction below its gallery price damages an artist's market publicly and permanently, so dealers place important pieces with museums and with collectors who will hold them, sometimes with contractual rights of first refusal. Managing the price history is part of the job. That is why the art market looks curated in a way no other market does: because it is.

And the prices themselves are more structured than they appear. Auction houses lend against consignments, guarantee sellers a minimum, and lay off that risk on third parties who commit to bid irrevocably in exchange for a fee. A lot can therefore be effectively sold before the room opens. The published figure includes a buyer's premium of a quarter or more on top of the hammer price.

A record result is a fact about a transaction, and the transaction is a financial instrument with a painting attached.