New Book Stats Describe the Flat Trade In New Ways--Including $838 Million in 2010 eBooks

After a couple years of no industry-wide statistics and many years before that of poor, illusory numbers, the publishing industry is back on the road to measuring itself with the new BookStats, prepared jointly by the AAP and BISG, released in highlight form to trade media on Tuesday.
All of the new numbers as you will read about them are modeled estimates rather than actual data--but they draw on data supplied by just under 2,000 publishers, and the methodology for extrapolating the estimates was far more careful and rational than the process used by the BISG alone for its annual Book Trends numbers, and has introduced fewer errors of gross exaggeration-by-multiplication. (More on this later....) Book Trends was abandoned after publishing data for 2008, and the BISG did not compile any statistics for 2009. The AAP has published annual estimates though they were considered less complete. So the new data effort covers 2010 but also compiles industry numbers for 2009 and 2008 using the same new method.
In the headline numbers, BookStats estimates trade publishers' 2010 sales at $12.59 billion. That's a slight decline from $12.71 billion in 2009, when sales were 7.4 percent higher than the year before, which registered $11.83 billion in 2008.
(Here's our first point of divergence with some other accounts: we calculate trade sales as adult fiction, adult nonfiction and juvenile sales. BookStats has put the "religion" breakout inside trade in their general press release. There is some overlap, but with $1.35 billion in wholesale sales, we're more comfortable leaving "religion" outside of the trade basket.
(Here's our second point of divergence: the official BookStats press release celebrates all of the increases in publishing, some of which can only be derived by looking at changes over a two-year period, instead of the traditional year-over-year view. And 2008 was when the economy plunged in the final quarter, so it may make for a skewed baseline, particularly on a two-year comparison. We are breaking out percentage changes on year-by-year comps, and our focus is on changes in the past year.)
*From Publisher's Lunch