Statistics and the Used Book Market by Susan Siegel Book Hunter Press © Book Hunter Press, 2004 #21, 21 June, 2004

Counting anything having to do with the used book market is not easy. I should know. I've been doing it for the past 12 years.

The U.S. used book market is made up of a fragmented community of anywhere from 8,000-10,000 used book dealers with roughly 50% of the dealers operating brick and mortar stores.

As discussed in greater detail in our just released report, A Portrait of the U.S. Used Book Market, other than the ongoing tracking of dealers done by Book Hunter Press, there are no systems or procedures currently in place for gathering statistical data, especially sales data, from the many segments that make up the "used book market."

There is no single, easy to look up, "used book industry" classification code in the North American Industrial Classification System (NAICS) that is used by the U.S. Census Bureau to track businesses. In fact, depending on whether the books are considered secondhand, used, rare, collectible or antique, used book dealers are scattered across several different NAICS codes, including one classification code that also includes thrift shops and used furniture stores. Compounding the problem, mail order and internet businesses under the NAICS classifications do not distinguish between new and used books.

There is no national organization that represents a significant number of used book dealers; the Antiquarian Booksellers Association of America (ABAA) has 475 members and the Independent Online Booksellers Association (IOBA) has 300 members. Only a few states have membership associations.

It gets even more complicated. How does one gather statistics from the thousands of new dealers now selling used books online? (Some industry observers would question whether these new sellers should even be considered legitimate used book dealers and be included in any statistical survey of the used book market.) How does one define a "used book seller"? Does one have to have X number of books? Been in business Y number of years? Have minimum gross sales of Z dollars?

Then there's the issue of what "used" books to gather statistics about: the term "used book" can include a 50 cent mass market paperback and a $7 million first edition - and everything in between.

And what about used textbooks? Or manuscripts, documents and ephemera. Are/should these be included or excluded?

The next hurdle in statistics gathering is deciding what selling venues to include - or exclude. Including sales from stores and online sites are obvious. But what about auctions or library sales?

Then there's the issue of which sales to count. Sales to consumers only? To libraries? What about dealer to dealer sales?

How does one count online sales from multi-dealer sites such as Amazon? A dealer may list a book on Abebooks, and then through Abebooks' associate program the same book is listed on Amazon with "Abebooks" as the seller. When the book sells, who gets credit for and counts the sale: Amazon, Abebooks, the dealer who owns the books - or all three sellers?

And how do the multi-dealer sites distinguish between the sale of new and used books? On Amazon, for example, books listed in the "used" category are described by their sellers as "brand new."

v Just the other day I saw an online posting claiming that eBay has sold a "staggering" 75 million used books. How does eBay know which books on its site are new or used? Has someone read all 75 million descriptions?

Finally, if a methodology could be found that satisfactorily addresses all of the above problems (and a few others not mentioned here), two additional problems remain: record keeping and diversity.

Only a small minority of dealers have their entire inventories online and/or have the software that enables them to track their sales. That leaves a significant number of the dealers who can only provide estimates or guesstimates about their sales.

Then there's diversity. A Portrait of the U.S. Book Market confirms what many used book sellers already know: used book dealers came in all sizes and types. Even within the same dealer category and size grouping there are wide differences in inventory, number of books sold per month, average gross sales, use of the Internet and other selling venues, etc. As the Report points out, when dealing with used book dealers, averages and generalizations must be used with extreme caution - and an understanding of their limitations.

Where does this leave someone like myself trying to gather statistics about the used book market?

Like most researchers and analysts, we do the best we can with what is available. And, if we're honest, we're transparent about our methodology and numbers. Which is what we hope we've been able to accomplish in what we believe is this very first statistical look at the U.S. used book market.