You read the header right: Amazon has access to roughly twice as many new texts from the nineteenth century as it does from the twentieth. This is thanks to copyright laws that frankly haven’t kept pace with the promise of the printing press, much less electronic publishing.
The real grunt work in preparing the data came from University of Illinois law professor Paul Heald, who described his effort this way:
We broke these out by decade. … You would expect that if you can crawl through Amazon looking at only new books and only books sold by Amazon — so these are not used books, these are not sold by Amazon associates, this is what’s in Amazon’s warehouses — of course, the biggest number of books is from the decade 2000-2010. That’s what you’d expect; they’re more recent, more popular. Drops off really quickly for books in the 1990s, 1980s, 1970s, ’60, 1950, 1940, 1930 — here’s the point in time where books start falling in the public domain. Suddenly it goes up and up and up. There’s as many books [that] Amazon is selling brand new right now from the 1900s to 1910 as from the 2000s to 2010. You go all the way back to 1850 — there’s twice as many books from the 1850s being sold on Amazon right now as the 1950s. So this sort of confirms the notion that there’s some sort of positive public-domain effect …
Read the whole article here.