Actually, I'm just saying that because technically, I can. What he actually did was make the underlying point of my posting of Thursday for me; if I say this kind of thing, it looks like I have an irrational prejudice against unpublished authors (given that I represent a fair number of authors who have not achieved commercial publication, that's a bit odd).
CE Petit makes publishing on demand sound like a sucker's bet. But what about the nonpecuniary psychic benefit to a frustrated author of actually seeing a book in print?
Actually, this is precisely the point. There is a substantial—and perhaps controlling—desire to "see the book in print." Even that, though, should be tempered by the number of people who will see it in print.
Time for some more math. Other data from XLibris indicates that only about 350 of its titles have sold more than 500 copies, and only about 1400 of its titles have sold more than 200 copies (sources leave unclear whether this includes those 350). This is much flatter a sales curve than one finds in commercial publishing—but it exposes the equal probability fallacy of using means bounded on one side. Since one can't sell less than zero copies, sales of zero is a bound to the data. On the other hand, the theoretical limit of sales is four or five orders of magnitude greater (at least if one believes the various boosters of "POD publishing"). Let's factor the minimum possible effect of those sales in to see what happens:
- Mean XLibris sale per title = 1,300,000/10,300 = 130 copies
- 1,300,000 total - (350 * 501 copies) - ((1,425-350) * 201 copies)
= 1,300,000 - 175,350 - 216,075 = 908,575 copies of books that sold < 200
1 - (1,425/10,300) = 86% of population is in the remaining group
908,575/(10,300-1,425) = 102 copies mean in this population - 1,300,000 total - (350 * 501) - (1,425 * 201)
= 1,300,000 - 175,350 - 286,425 = 838,225 copies of books that sold < 200
1 - ((1,425+350))/10,300 = 83% of population is in the remaining group
838,225/(10,300-(1,425+350)) = 98 copies mean in this population
Of course, every copy over 501 depresses things further for the remaining population, as does every copy between 201 and 501.
This is exactly what a sucker's bet is: A bet that the "nonpecuniary benefit" will make up for the financial cost. In this case, the "nonpecuniary benefit" is "seeing the book in print." However, "seeing the book in print" had better mean that 100 or fewer others will also see it in print, because that is the expectation value for the vast majority of participants in this market. I can think of few clearer demonstrations of irrational economic behavior than vanity publishing. The main "attraction" of POD vanity presses is their alleged chance for economic reward; one can readily infer such from the emphasis that they place on financial aspects of "publishing" with them, as opposed to more than lip service to even the "satisfaction of seeing your book in print!" (direct quotation; source withheld to protect the guilty).
POD is a printing technology. It changes the price points of the vanity publishing model from investment of from $5,000 to $15,000 (or even more) resulting in a breakeven point in excess of 2,000 copies of a typical 2,500 copy print run—an astounding, and almost never achieved in the history of publishing, sell-through of 80%—to a breakeven point of between 150 and 200 copies (assuming that the author will buy a few copies, for no royalty, as gifts for friends, for reviews, and for marketing). What does not change is that the money flow still points away from the author; the distinction in degree is not nearly significant enough to make it is distinction in kind. This is still vanity publishing.
So, then, the Perfesser is right. His question goes to the core of why so many economic models don't work in the real world: Once one admits that valuation cannot be perfectly converted to utils (or any other dimensionless unit, such as "dollars"), both the individual actors and the market being studies have substantial "irrational" behavior. In the vanity publishing market, "irrational" would be a very generous description indeed.