Tracking the Rare Book and Manuscript market

This past week I attended the annual conference of the Rare Book and Manuscript Section of the Association of College and Research Libraries. It's fantastic to be around so many wonderful book people and hear their take on the state of the field. As part of the program, RBMS hosted panel on "the market" with Nina Musinsky and other members of the trade and library world. Seeing the plenary and Musinsky's talk reminded me that I'd started several months ago to make sense of some data on 2013 book and manuscript auction sales but never finished.

On January 1st this year, the collector services site Americana Exchange (AE) posted a list of the "top 500" auction results by price for books and manuscripts for the previous year based on their valuable in-house data. I thought I'd clean up and parse this data a bit and try to make some sense from it. The AE's table makes it easy to see the list by value, capped off by the Bay Psalm Book which sold for $14 million at Sotheby's. I wanted to get a sense though of the field as a whole. First off, while I was unsurprised that Sotheby's and Christie's dominated the field in terms of auction houses selling top lots, I was impressed by the fact that 48 different auction houses were represented over all 500 lots!

Top 10 Auction houses in 2013 by number of the top 500 lots sold.

I then thought I'd look a bit at the age of the items being sold in the market - was the 20th century the hottest? The 16th? After a bit of cleanup I assigned dates to 497 of the 500 items and plotted them out.

Number of items in top 500 by century

There are no huge surprises in the above table with the 19th and 20th centuries responsible for the majority of the top value book and manuscript auction sales with the 17th century the poor relative in the printed-book era. The list is of course worth looking at carefully in comparison with the numbers, you'll see, for example, that a sale of comic books at Heritage Auctions really boosted the number of items from the 20th century.

Total value (US$) of items in top 500 by century

The twentieth century also fares well in looking at the total value of all lots by century, but you'll see the 17th century recovers thanks largely to the Bay Psalm Book whose high price compensates for lower total sales from the period.

Average price (US$) earned by items in top 500 by century

In looking at averages, medieval manuscripts, though numerically fewer on the list, shine through thanks to their higher per-item value. You'll see of course that the Bay Psalm Book is responsible for that inflated 17th century average.

The AE data also includes information on auction house estimates which provides an interesting window on which items blew away expectations (or which had artificially low estimates). I've divided the final sales price by the low estimate to get a kind of 'estimate factor' by which lots overperformed.

Top 10 auction lots of 2013 by how much they exceeded their low estimate

The two top expectation-beaters are illustratively quite different. The top lot, a 1555 first edition of the works of Louise Charly Labé was given an estimate of $3,000-$5,000 at Sotheby's (New York) on 11 June 2013, instead it sold for an astounding $485,000. Where the Labé volume is a beautiful 16th c. letterpress book with a fantastic binding, the second highest performer, sold just eight days later, couldn't be more different. The ugly little pamphlet on the right dates from 1937 and contains a printed offprint of "On Computable Numbers," one of Alan Turing's seminal articles. It was offered at Bonham's in London June 19th for the modest estimate of £3,000-5,000 and instead achieved a whopping £205,250 ($349,591). One of maybe fewer than 100 offprints of the article, this one is inscribed by Turing to a Cambridge philosopher and is clear evidence of the enormous appetite for items related to the history of computing.

Works of Louise Charly Labé
(Lyon, 1555) [USTC 1135] 

Turing's "On Computable Numbers,
with an Application to the Entscheidungsproblem"
Proceedings of the London
Mathematical Society (1937)

It's nice to see this juxtaposition which shows two of the many sides of the collecting market. Early printed books like the 1555 Labé continue to do well both for their physical beauty as well as their historical importance (one of the most important early printed compilations of a female poet) while the Turing offprint demonstrates the power and interest of a cohort of collectors attracted by the recent history of science and computing. Both are historically significant material and intellectual objects and I think pretty compelling evidence for why it's a great time to be working in the Rare Book and Manuscript field.  

My data set, based on that of AE but with my addition of dates and the 'estimate factor' can be found here.