We do not therefore conclude that the basic economic relationship in a given entertainment industry is determined by the bit players and the wannabes. It isn't. Rather, and exactly the opposite from industries dominated by wage labor as the principal form of labor, the basic relationship is developed from the top down. It is, like it or not, the superstars and the most valuable players who primarily determine the level of compensation and the principles of that compensation.
Why? For the good and simple reason that, push comes to shove, they are the ones who draw the customers. Not their "bosses," who aren't bosses to begin with.
People came and paid for tickets to see Michael Jordan play basketball or Reggie Jackson or Nolan Ryan play baseball. They did not come to see the owners of those teams sitting in their sky boxes.
People pay for movie tickets or buy DVDs to see Will Smith or Tom Hanks or Meryl Streep or Julia Roberts perform. Some of them are also drawn by the name of the movie's director. But I can assure you that the number of people sitting in a movie theater who know or care who the producers of that movie were can be counted on the fingers of one hand—and the number that comes up will usually be zero. The same is true, even more so, for people sitting at a musical performance.
The same is true, as a rule, for people buying a book. Almost every reader can tell you the name of the author of the book they're reading. Not more than one in a hundred are likely to notice the name of the publishing house. There are some exceptions, of course, but not many. In science fiction, over the decades, one or another publishing house would develop such a distinctive identity that at least some readers were drawn by the identity of the house itself, and an even larger percentage recognized the identity. But, even there, the main reason for the distinctiveness was that a publishing house or magazine developed a very popular set of stable lead writers.
And that's why it's ridiculous to try to make an analogy between independent workers in an entertainment industry and wage laborers. The identity of any individual wage laborer vanishes in the process of the work. Indeed, as a rule, it has to vanish or the end product is deficient. When I worked as a machinist, for instance, any shaft I made on a lathe or housing I machined on a boring mill had to be indistinguishable from any other in that line. Why? Because what the customer was paying for was a product all of whose proportions fell within a very narrow range of tolerances. However creative may have been the work of figuring out how to make that product, the individual worker's creativity became invisible in the product itself.
The opposite is true in entertainment. Much as they might like to—in some cases, fervently wish to—it's just a cold fact of life that a movie producer can't use actors interchangeably. He can't produce a "Tom Hanks movie" without Tom Hanks. He can't tell the real Tom Hanks that if he doesn't agree to work for peanuts he'll go out and hire another Tom Hanks.
Granted, a producer—or a publisher, or the owner of a sports team—can decide that he'll make as much money if he contracts for the services of a different actor or writer or athlete. But then he still has to deal with that actor or writer or athlete, and he'll still have to deal with them on whatever economic basis is determined by that individual's performance in the market place.
So, over time, certain basic contractual terms become established for everybody—or most people, at least—in a specific entertainment industry. To use commercial fiction publishing as an example, the standard contract for any writer (except writers on a work-for-hire basis) is as follows, in terms of royalties paid:
For a hardcover title, the author will get either 10% or 12.5% or 15% of the cover price of the book for each and every copy sold. The difference is determined by the level of sales: an author gets 10% for the first 5000 copies sold, 12.5% for the next 5000, and 15% for any copies sold past ten thousand. For a mass market paperback, the percentages are 8% for all copies sold up to one hundred and fifty thousand, and 10% thereafter. For trade paperbacks, there's a slightly wider fluctuation in terms—some houses pay the same rates as for hardcovers—but the general rule of thumb is that an author will get 7.5% of the cover price for every copy sold.
Those established percentages are not and were not set by little-published authors. They were established, over time, between publishing houses and their bestselling authors and the agents for those authors. And having been established, became generalized throughout the industry for all authors. (It's true that new authors often have to sell a first book on a slightly sub-standard basis—for instance, 6% and 8% for paperback royalties instead of the standard 8% and 10%. But that's a passing phase, one way or another. If they can't sell any more books, it's a moot point. And if they do, they'll soon start getting the standard rates.)
Okay, so much for the basic economics involved. Now let's go back and look at the question posed directly.
Given these specific economic relationships, what is the net effect on the income of authors as a whole if some of the authors start handing out some of their works for free in electronic format?
Well, the first and most obvious effect is that the authors who do so will become better known to the reading public at large. And, as a result of their greater visibility, will generally see an increase in their sales.
I need to stress that this is not guesswork on my part. It's something that has been proven in practice over a period of years—by me, among other authors.
I can demonstrate this concretely, by using two of my novels as my examples.
The first book I put up for free online—in fact, it was the first volume in the Baen Free Library—was my first novel, Mother of Demons.
Mother of Demons was published in September of 1997, and it was only published in a mass market paperback edition, as was the standard practice at the time for first novels. (Most science fiction novels, actually.)
The initial sales were decent—good enough for the publisher to contract for more work from me—but hardly spectacular. Within two years, by which time the vast majority of books will have had most of their sales, the book had sold a little over ten thousand copies, with a 55% sell-through. (To remind you, "sell-through" refers to the percentage of books shipped which are actually sold.) That was slightly above average for a first novel in science fiction at the time, but only slightly.
At that point—partly because, being frank, I knew I had nothing much to lose—I put Mother of Demons up for free online in the Baen Free Library. Jim Baen and I, after discussing the matter, both thought it would be an interesting experiment. Jim was convinced that an author would lose nothing from doing so, and might very well gain as a result, and I was willing to be the guinea pig.
What happened next, over the many years that have followed, is both fascinating and instructive. Here are my royalty figures for Mother of Demons, for each successive royalty period. For those of you who don't know, there are two royalty periods a year in publishing, at six month intervals. As a rule, though not always, the periods end on June 30 and December 31.
You probably have to be an author to grasp immediately just how anomalous these results are. Here are the basic rules of thumb in commercial publishing when it comes to sales:
Rule Number One. 80% of all the sales of a title will happen in the first three months after it is published.
Rule Number Two. Thereafter, sales will decline steeply.
Rule Number Three. Within a few years—almost always less than five, and usually only two or three—sales are so low that it's no longer worth it for a publisher to pay the warehousing costs to keep the title in print.
Rule Number Four. Sell-through rarely improves, and usually declines over time.
But with Mother of Demons, none of those rules applies.
As of the end of the last royalty period, almost nine years after the novel was first published, more than half of the sales came after the initial sales period. Almost 60%, in fact.
Although no other period, of course, came close to that first period in total sales, there was no notic
eable decline thereafter until June 2005. For the first six years, though the end of 2004, the book sold an average of 873 copies each period. Then, even the clear drop that took place after six years simply seems to have set a new plateau for regular sales. In the two years that have elapsed since then, average sales dropped to 515 copies. But that "average" has seen hardly any fluctuation at all, and certainly no decline.
As of the day I'm writing this essay, August 29 of 2007, I am only three days away from the tenth anniversary of the publication of Mother of Demons. My first novel, mind you, and a novel which has been available for free in an electronic edition for the last seven years—and which has never gone out of print.
Do you have any idea how unusual it is for an author's first novel to still be in print ten years after it came out? The exact percentages are unknown, but they are undoubtedly less than five percent, and probably less than one percent.
Before much longer—I figure, about a year from now—the existing print run for Mother of Demons will have been sold out. Normally, even for a novel with "long legs," especially a paperback edition, that's the point at which a publisher decides to let it go out of print. Even if sales are steady, they simply aren't big enough to justify another print run, given the economics of mass market printing. It's not worth doing a print run smaller than five thousand copies, and publishers prefer at least ten thousand. Even with the solid existing sales of Mother of Demons, it would take too many years to sell that many copies.
As it happens, however, Mother of Demons is not going out of print. Toni Weisskopf, who replaced Jim Baen as Baen Books' publisher after Jim's death last year, told me a couple of months ago that she was planning to do a limited hardcover edition of the novel. It'll be a small print run, but that's economically viable for a book which has small but solid sales.
So. Almost eight years ago, I put up my first novel for free online—as a result of which it got most of its sales since then, and is still selling well enough that even after the mass market edition finally runs out, the publisher is going to keep it in print in a hardcover edition.
Nobody knows exactly what percentage of first novels never go out of print for ten years and then get reissued in a hardcover edition. But the percentage is probably somewhere in the top one-tenth of one percent.
We'll see the same pattern, in a moment, with another novel of mine. But, before I move on, let me just make a note. The next novel whose sales I'll analyze, 1632, is the first novel in what has become a long-running and popular series, with (to date, as of this October) six novels and six anthologies in print by January 2008. Some people who argue with me on this subject will claim that the ongoing sales of 1632 are due to that, not to the fact the book has been made available for free online for many years.
Baloney—and I know it's baloney because Mothers of Demons is not part of any series. It's a stand alone novel, which has no sequel that might keep stimulating its sales as the first book in a series.
Here are the comparable royalty figures for 1632, which was first published in February of 2000 in a hardcover edition. That edition was followed by a mass market paperback issue that came out exactly a year later, in February of 2001. I've forgotten the exact date that I put the novel up for free in the Baen Free Library, but it was sometime in that same year.
One note of explanation. I'm only using the paperback royalty figures because, for our purposes here, it's the paperback figures that are critical. A book that comes out first in a hardcover edition, followed by a paperback reissue, sees an almost complete stop to hardcover sales once the paperback appears. So you can't use hardcover royalty figures to gauge a book's longevity.
Except for the first sales period, ending December 31 of 2001, all of these sales took place after I put the novel up for free online.
What we discover is exactly the same pattern. As before, no period enjoyed the same level of sales as the very first period. But that's a given with almost any book, certainly a novel. (The rules for textbooks mandated by schools are different, of course. But very few novelists—and practically no living ones—have a phalanx of teachers and college professors to keep maintaining their sales by making their books assigned texts in classes that the students have to buy, whether they want to or not.)
But—this is completely out of whack with normal commercial publishing experience—most of the sales took place after that initial period. And the result is even more striking than it was with Mother of Demons. As of the last royalty period, 1632 has sold almost ninety-five thousand copies in paperback. Fully two-thirds of those sales happened after that first big sales period—and after the book was made available for free to the public in an electronic edition posted in a well-known and well-advertised web site.
There hasn't been the same steady increase in sell-through that we saw with Mother of Demons. But that's because 1632 started at 85% and soon reached 90%. If you ask any experienced publisher or publishing sales representative, they will tell you that a ninety percent sell-through rate is just about as high as any sell-through can ever get, in the real world. A "perfect" sell-through of one hundred percent is only theoretically possible—and it wouldn't actually be "perfect" anyway. Truth be told, I've never been that happy that 1632 has maintained such a phenomenal sell-through for all these years.
Yeah, yeah, fine, it's great for bragging rights. But the fact is that a sell-through that high is partly reflecting a bottleneck caused by the fact that the miserable misbegotten distributors—do the jerks pay any attention to their own figures?—aren't ordering enough copies to begin with. I'd far rather see the sell-through drop to eighty or even seventy percent, if that reflected greater shipping orders and therefore a better circulation of the book.
Still, that grumpiness aside, the fact that 1632 has maintained almost a ninety percent sell-through for many years, despite being available for free in an electronic edition, is actually enough in itself to make a mockery of the argument that putting up works for free hurts an author's sales.
Notice, furthermore, that the sales have never dropped at all, once we get past that initial period. Since the period ending December of 2002—almost five years, now—1632 keeps selling, period after period, with no discernible decline in sales. There are fluctuations from one period to the next, naturally, but there is no overall decline at all. The average sales for any given six-month period during that five year stretch have been 6,645 copies—and the sales in the last period were higher than that average.
Part of the explanation, of course, is that 1632 is the first novel in a long series. One of the other rules of thumb in publishing is that the first novel in a series will always outsell all the other, later volumes—for the good and simple reason that readers who come into a series midway and enjoy whatever book they happened to stumble across, will typically start the series at the beginning by going out and finding the first book. To a lesser degree, the same phenomenon applies to all later books in the series.
So, if you step back and look at sales of a series over a long stretch of time, what you will almost always discover is that the sales fall on a sliding continuum, with the earlier books selling more copies than the later ones. Mind you, this is a relative relationship of sales, not an absolute limit. With a successful series, all of the books in it will show increasing sales. But the earlier volumes, especially the first one, will always sell more copies than later volumes.
Fine. But I think it's patently absurd to claim that the fact the novel has been available for free in electronic format during that entire period has nothing to do with it—much less that it somehow hurt the sales of the book.
I will now make two very blunt statements:
One. It is preposterous—downright lunacy—to accuse me of being a "scab" and lowering the pay rates for all authors, when I make far more money than ninety-nine percent of them. For the past several years, my annual gross income as a writer has been around $150,000. In fact, it's been slowly climbing. That certainly doesn't p
ut me in the league of really top-selling authors like John Grisham or Robert Jordan or Nora Roberts—much less J.K. Rowling—but it does put me in the top one percentile of all commercial fiction authors, including science fiction and fantasy authors.
To put it another way, you might just as well—and just as idiotically—accuse major league baseball players of lowering the pay scales for ball players in the minor leagues. Or accuse established character actors who get regular work in the movies of lowering the pay scales for bit part actors and extras. (That's what an author like me is, by the way, if we use the movies as an economic analogy. I'm not one of the top stars, but I get lots and lots of work and make an income that's at least twice the median income for Americans today.)
Two. It is just as preposterous—downright lunacy, again—to claim that my policy of putting up all of my works for free online after an initial period is hurting my income, and therefore the income of other authors. In fact, it does the opposite.
Jim Baen's Universe-Vol 2 Num 3 Page 32
