The stultifying confusion over library ebook lending would benefit from a more empirical approach, and I would like to propose, at least for the short term, that HarperCollins’s 26-loan cap model, while far from perfect, should receive more careful analysis.

In a September 25 statement, the Association of American Publishers (AAP) said, “Publishers support the concept of e-lending but must solve a breadth of complex technological, operational, financial and other challenges to make it a reality.” HarperCollins proffered a viable publisher solution 20 months ago.

In a September 24 open letter to recalcitrant Big Six publishers, Maureen Sullivan, president of the American Library Association (ALA), wrote, “Access to [ebooks] must not be denied.” Under HarperCollins’s model, access has been continuous (except for those libraries that boycotted).

But how carefully have publishers and ALA analyzed the accumulated evidence from HarperCollins’s model?

The AAP said, “And while the 9000-plus library systems’ non-profit status permits them to convene, debate and reach consensus on these issues, commercial publishers cannot likewise come together due to antitrust restrictions.”

Leaving aside the dubious assertion of a library consensus, nothing prevents commercial publishers from talking to libraries on their own and exploring the loan cap model that one of their trade brethren has already shaped for them. A careful measure of the model has yet to be taken.

For example, Penguin had a falling out with OverDrive over Kindle compatibility, and it now has a pilot, with time lags in availability, going on with one platform (3M) and two libraries.

According to the Fall 2012 issue of LJ’s Patron Profiles, lags in availability (windowing) are palatable if the lag is not too long: 54 percent of patrons surveyed said a one-month window after print publication is reasonable; 40 percent found a three-month lag reasonable. A six-month lag, as Penguin proposes, is likely considered even less reasonable. But only 20 percent found a model like HarperCollins’s reasonable.

However, as of October 4, only eight licensed titles (seven are Agatha Christie novels) at New York Public Library (NYPL) had hit HarperCollins’s loan cap introduced in February 2011. NYPL plans to relicense them, and HarperCollins offers a discount on second copies. Under the Penguin model (one-year license), all of a library’s Penguin titles would have had to have been relicensed eight months ago, not to mention that they are, for now, restricted to one platform. What if Macmillan, which is considering its own pilot, decides to avoid OverDrive and 3M and make its titles available through yet another platform exclusively? How many platforms can libraries manage?

So what is more reasonable? More sustainable? The debate would benefit from a deeper look at what experience libraries are reporting with HarperCollins’s model.

If adopted (and adapted) by other Big Six publishers, the HarperCollins model removes the nettlesome problem of denied access (thus taking a sharp edge from the debate), it stems the Balkanization of platforms, and it avoids the difficult pricing of Hachette and Random House. But Random House does not object if a library wants to move its collection from one authorized wholesaler platform to another. It says libraries own Random House ebooks, but Peter Brantley of the Internet Archive has very effectively dissected what this means (a must read). Nevertheless, does a promise of perpetual access eventually make Random House's pricing more economical than repeated licensings? We need a better sense of how much relicensing occurs in order to answer that.

Laura Irmscher, chief of collections strategy for the Boston Public Library, said at LJ's online ebook summit last week that she had come to prefer HarperCollins’ 26-loan cap over Random House pricing, primarily because new content is so important for driving circulation. “If 600 people are waiting for a book, I want it to be cheaper…I don’t need 100 copies forever,” she said. “I might need five copies forever.” And Noel Rutherford, the manager of collection development and acquisitions at the Nashville Public Library, agreed, describing the HarperCollins's model as offering an “acceptable” cost per circ at “basically a dollar per download.” She added that the Penguin embargo period effectively eliminates for libraries the benefits of any positive early buzz about the book. “I want something when it’s hot,” she said. “When it’s being marketed and sold…That’s what the public expects.”

There are librarian tenets, such as the duty to preserve our cultural memory, that require ownership or, at a minimum, a carefully worded perpetual license. There are significant publishers such as Gale that not only have developed digital preservation strategies with Portico but also allow ownership (although I hesitate to use the word!) of proprietary titles:

"Because the library owns these titles in-perpetuity if the library wants, they can host them on their own platform. Access to any Gale Virtual Reference Library title is unlimited, any time by any number of people," Cynthia Sanner, the publisher, general reference and international at Gale, told me recently.

Yet the HarperCollins model has been a logical, good faith effort, particularly for those all-important frontlist titles. Perhaps the loan cap can be raised from 26. Perhaps the better meter is time rather than circulation, as Penguin posits, but more than a year: let’s remember the Harry Potter deal grants libraries five years of access before a license renewal is due.

But given present realities and keeping in mind the short-term goal of getting all Big Six digital content into libraries, HarperCollins’s model deserves some credit and more detailed thought.