First, though, one needs to eliminate one purported problem from the discussion. I've heard rumblings — not quite silverback-gorilla chestbeating, but close — from several different perspectives on this issue to the effect that "commercial publishers will refuse to deal with authors/literary agents who are really competitors and not content providers." Hogwash. Leaving aside the antitrust problems that would/might create, an author/literary agent who has a literary property available to a publisher that appears likely to the publisher to be highly profitable is still going to get a commercial publishing offer. In marginal cases, it is perhaps possible that a particular commercial publisher would look askance at the author/literary agent "competitor"... but that can be solved through the contracting process itself, and if it can't that just might be a hint that the property was a poor fit for that publisher in the first place.
Second, there are essential preconditions on the relationship between the author and the literary-agent-acting-to-facilitate-e-books. To expand upon my comments at Dean Wesley Smith's blog, the minimum arrangement must include all of:
- No commingling of funds at all. E-book revenues are completely separated from representation revenues, and preferably in a separate business structure to avoid liability spillover from one to the other. The obvious corollary is that the e-book is made available to the public under a separate written agreement, distinct from the author/literary agent agreement. This is protection for both the author and the literary agency, since e-books may end up in markets for which they were not originally intended, and thereby (under some legal systems) create liability for, say, blasphemy.
- The agent’s compensation is no greater for the e-book than the agent would get from any other license transaction. This is essential to avoid both the fact and appearance of a conflict of interest. If the agent is taking, say, half the revenues from e-book sales, that may be a better deal than commercial publishers presently offer... but it also creates a disincentive to actively market the rights package as a whole, for which the agent's compensation would be only 15%.
- The electronic publication concerns an already-printed work that has already reverted/never licensed electronic rights, especially on behalf of an estate, and for which the agent and author have agreed that existing (and actual) commercial offers have been inadequate. This may, under some circumstances, be a collection of an author's previously published shorter works, especially fiction. There are possible circumstances in which a related not-previously-published work could be published as an e-book — say, an orphaned-by-demise-of-magazine novelette that provides linking material between two commercially published novels — but those circumstances are very, very rare and require involving a sophisticated literary-rights attorney early on to ensure that the proposed publication is, indeed, appropriate.
- The deal is for a limited term with opt-in renewal periodically — certainly terms not more than five years, and preferably shorter than that. This is an essential ethical and economic contrast with commercial publishing practices and attempts to grab "life of the copyright" licenses... not to mention deal appropriately with the "out of print" problem.
- There is an appropriate way to terminate the agent’s e-book license for a particular work if the author gets a better, non-self-interested offer elsewhere (such as, but not only, a package deal with a print publisher that pays more and/or provides other revenue streams in the package). On the other hand, if the author dies, and the author's heir's niece's boyfriend says he can do everything cheaper than the agent does (which is unlikely if point 2 has been satisfied), that's not a non-self-interested offer, and is not sufficient grounds for termination.
Of course, these are the bare minimums to make a deal possible, not everything that should be there. No deal that violates any of those conditions passes even the laugh test (“Your Honor, our position is that {breaks down into laughter}” = EPIC FAIL), let alone is a good deal for an author. My point is only that under certain seemingly narrow conditions (that are disturbingly common given the abject idiocy and incompetence at the commercial publishers), it is possible to create an appropriate electronic publishing relationship between a copyright holder and an agent acting to facilitate electronic publishing.
Not every literary agency has the necessary technical and/or legal knowledge to facilitate placing electronic versions of its clients' works on the market. There are some significant tax issues, and legal issues, and technical issues, and vendor-compliance issues, that will not get simpler in the foreseeable future — rather the opposite — that will make many (and perhaps even most) literary agencies inappropriate partners for authors, and especially heirs of authors, seeking to bring out electronic versions of a backlist. Conversely, not every author's backlist is appropriate for this kind of structure; for example, coauthored works in which one or more authors is deceased or has a short life expectancy are completely inappropriate, if only due to the insanity of and delays imposed by probate proceedings and taxes (and you do have a current will designating how to handle your literary estate, right?).
On the other hand, the electronic publishing rights grabs being imposed adhesively by commercial publishers are inexcusable and — in any just world — prohibited by unfair trade practices law. This not being an exceptionally just world, I can only argue that publishers' present practices implicate unfair trade practices regulations, and jump up and down about it. (I'm smart enough not to hold my breath while doing so.) That is, an appropriate author/literary agency electronic publishing deal can be a win-win for the author and literary agency... and a win for the reader; any disadvantage to the media conglomerates is just schadenfreude.