The perpetuity model allows for a clear illustration of what the possible return on a constant stream of book sales might be. In reality, for virtually all works, the actual story is that you'll see a few sales in the first few years, and virtually none afterwards. A very small number of works hit the long-term publication lottery.

Which means that you're looking at a future payment stream that's both decreasing nominally AND is being discounted to an ever increasing amount.

But in order to support the increasingly perpetual duration of copyright, for which only a handful of works benefit, the public is enjoined from having access to _any_ works published in the past century, unless the copyright holder can be found, identified, and persuaded to publish the work.

This has a number of profound effects.

There's a hole in the sale of new editions from Amazon (not "freely available downloadable books", mind you, but sales) from 1920 to 2000. There are more books from the 1840s available than from the 1930s.

Germany's industrial expansion from a series of disjoint states to the powerhouse of Europe in the 19th century is thought to have been greatly promoted by weak copyright.

I explore both here: http://www.reddit.com/r/dredmorbius/comments/23xrkd/copyrigh... (with some bonus entertainment thrown in).

Freakonomics has looked into the question of whether or not copyright makes books disappear: http://www.reddit.com/r/dredmorbius/comments/1wp3o8/freakono...

And noted copyright jurist and author Pamela Samuelson, of UC Berkeley, has stated: "Copyright has become the single most serious impediment to access to knowledge" http://www.reddit.com/r/dredmorbius/comments/1wssas/copyrigh...

You're not just doing math, you're making a moral argument: that if the length of a copyright term is lesser due to the lifespan of the author, that this is somehow 'discriminatory'.

I'm actually willing to accept the argument that copyright term should be independent of the life expectancy of the author -- that does affect the perceived present-value of the work. It's also how copyright was initially construed: a 14 year term, extensible once. Later extended to 28 years, with one renewal (by the author or heirs), to a maximum of 54 years.

What I don't support is for the term to be either indeterminate (as the repeated extensions of the past century have established), or perpetual, or even particularly extended. There are only a small number of cases where a 50-60 year term wouldn't be sufficient, and the concept of an extension means that unvalued works would enter into the public domain quickly.

Then we are in agreement on the first part; and I have no opinion re. the latter.

Re: the latter. Look at the Constitutional basis for copyright and patent, at least in the US.

To achieve the ends, that is, the progress of science and useful arts, the means is provided: exclusive rights for limited times.