Bookstore Sales and Publisher Dreams

The US Census Bureau recently released the first six months of 2005 unadjusted data for bookstores. Sales shown in millions (M) of dollars:

January - $1,969 M
February - $1,027 M
March - $1,038 M
April - $940 M
May - $1,055 M
June $1,084 M
Year to Date $7,113 M
Same period last year $7,395 M

Sales in 2004 are actually running behind 2005 by about 4%, despite higher prices. This means that overall number of books sold dropped more than 4%. The census isn’t perfect (these numbers are for brick-and-mortar bookstores only) but I wouldn’t ignore them.
Industry organizations, on the other hand, can produce truly bizarre statistics. The PMA (Publishers Marketing Association) released a study titled "Rest of Us 2003" which concluded that small and independent publishers were selling between $29 and $34 Billion dollars. My own calculation indicated that this total was off by an order of magnitude, about ten times higher than reality. How could they get it so wrong?

The basic methodology was to passively survey PMA members (active publishers willing to pay dues to an industry organization) and extrapolate the whole small publisher universe from those who responded to the survey. Could it be that a self-selecting group of successful publishers responded? Their method of extrapolation from "active" Bowker titles was terribly unsound. Besides, what percentage of small publishers EVER takes their titles out of print, rendering the records inactive?

It would not surprise me a bit if 90% or more of the "active" publishers reported by Bowker have such low sales that they don't even report taxes as publishers. The IRS, by the way, estimated book publisher revenue in 2001 (including the big trades) at around $23 billion. The BISG put all publisher sales at $26 billion in 2002. Are tens of thousands of small publishers each selling tens of thousands of books all cheating on their taxes? That’s what the math would require.

The extrapolation they made is similar to interviewing frequent fliers (the publishing parallel is PMA members) about their travel habits, than trying to extrapolate the number of international miles Americans fly each year by multiplying the survey average by the number of passports issued by the government in the past 10 years. There is simply no rational reason using the total number of "active" ISBN block holders as the basis for any estimate.

Just because most small publishers aren’t successful selling their books through traditional outlets doesn't mean that non-traditional outlets are capable of generating tens of billions in sales; if they did, they would be called traditional. I corresponded with the author of the survey, who acknowledged the methodology could be flawed, but felt it was worth keeping in order to compare apples-with-apples from a previous survey. Dreams are cheap, yet while it's easier for a publisher to have dreams than have sales, it's not a business model I'd want to be stuck with.