I was thinking (not again!) about models that might work for staring a new publishing company as a business, as opposed to as a charity or a public service. When you sit down and start looking at the variables that go into launching a new press that isn't based on self publishing, one of the unknowns is the quality of manuscripts you'll be able to attract. It would be nice to imagine that there's some direct relationship between the advance and royalty terms you'll offer and the quality of the manuscripts you'll receive, but unfortunately, offering better terms doesn't discourage low quality submissions. Provided you have a mechanism to present your offer in front of authors, the better the terms you offer, the greater the quantity of manuscripts you'll get to weed through.
So lets say you have a business model you'd like to test and a way to tell authors about it. When is ethical to start accepting manuscripts? It doesn't strike me as fair play to solicit a large pile of manuscripts just to skim through them and see if something of indisputable commercial value shows up. There are human beings behind those SASE's who actually go to bed the night they put the manuscript in the mail thinking that they may have taken the first step on a path to a new life. There's also the query letter and proposal process that you're putting the author through, as they torture their synopsis into something that they hope matches your criteria. It may be more humane to just accept whole manuscripts, sans cover letter, by e-mail.
Accepting manuscripts represents one half of a bargain, where the other half is that you'll actually be publishing some of them. But getting back to the new publisher's standpoint, committing to publish Y manuscripts out of a stack of X submissions amounts to buying a pig in the poke. What if none of them are commercially viable? If you put a limit on the amount of time the publisher will remain in a state of accepting submissions without publishing any, getting stuck with non-commercial titles becomes a very real risk.
Further clouding the issue is that the commercial viability of the manuscript is in part dependent on the resources and overhead of the publisher. On the one extreme, a new publisher with no experience in marketing or book production and no budget for either is highly unlikely to be able to turn the best manuscript of the year into a commercial success. It's one of the main reasons trade authors usually prefer working for the larger publishers. They know the book will at least get its day in the sun, even if it fails to make money. On the other extreme, large publishers have large overhead, and the number of books they need to sell to break even in their overall accounting can vary from 5,000 to 20,000 copies. A small publisher paying a small advance and running a very tight ship might break even on as few as 500 copies.
Let's make an unsupported leap of faith and say that there are as many manuscripts capable of selling 500 copies from a publisher with modest resources as there are manuscripts capable of selling 5,000 copies from a large publisher with extensive resources. That implies that the odds of a new publisher attracting commercial manuscripts are decent, as long as the publisher models the business around low sales. Of course, low sales means a low advance and low royalties for the author. One thing the publisher can do to try to compensate the author for the risk is to pay a high percentage of the net in royalties than the large trades. That's not difficult to do, considering the relatively low royalties on net, 5% to 15%, most trade publishers offer on trade paperbacks.
Just thinking out loud.