« Io v. Veoh: Harmful to YouTube? | Main | Why don't feminists like federalism? »

Friday, August 29, 2008

Why course packets ought to be fair use

Yesterday I received a set of "scholarly perspectives" from UCLA Law School -- a handy little reprint containing digested versions of their faculty's scholarship. Doug Lichtman's contribution contained the following statement about publishers' desire to facilitate the creation of course packets -- an assertion that is, from my experience, so wildly implausible that it makes me suspect that Lichtman has never actually spoken with an employee of a publisher's "permissions department" in person.

Here is the statement: "Without fair use, copyright holders would for the most part license [course packet excerpts from their copyrighted works], anxious to earn the additional royalties associated with classroom adoption and cognizant of the fact that a faculty member can always assign other reading if a given copyright holder asks for an unreasonable price or imposes unreasonable terms."

As someone who has done a bit of casual field research in this area, this statement seems to me an especially egregious example of economoid tendencies to mistake a model for reality. In fact, in my several interactions with the permissions departments of various publishers, I have never detected the slightest hint of anxiety to negotiate a plausible price. The reason, as I'll explain after the jump, is that assembling material for a course packet is like assembling parcels for a New York City skyscraper: The transaction costs destroy the possibility of a voluntary assembly even when the assembly value obviously exceeds the value of the disassembled pieces. This is why I'd argue that course packets ought generally to be "fair use."

When I've asked publishers for permission to use a chapter -- always less than 10% -- of a historical work, the clerk at the end of the 'phone never budges from a price that is wildly disproportionate to the value of the property that I am trying to purchase. They simply do not care that (for instance) charging my students $10-15 each for the one-time use of a 20-30 page chapter of an out-of-print history of political parties is preposterous. Inevitably, I walk. Inevitably, the clerk does not give a damn. Inevitably, the publisher loses the opportunity to earn a few bucks off of a book that now gathers dust in libraries.

The clerk knows that I can easily choose another book on political parties: Morton Keller, Richard McCormick, Joel Silbey, John Aldrich, and many others allegedly vie for my business. Much more significant, the clerk knows that I can simply write up my own summary of the relevant history using their copyrighted material, which is what I generally end up doing. (Publishers cannot copyright history itself, after all). Indeed, I tell the clerk all of this in response to the absurd price quote before I thank him or her and hang up.

So why does the clerk not negotiate "anxious[ly]" as Lichtman predicts? The reason is probably transaction costs: I suspect that it simply is not worth publishers' money to hire someone with the knowledge and training to exercise discretion in the permissions department to negotiate over the penny-ante sums available in the course packet business. Instead, the clerk is given a price sheet -- say, a nickel per page -- from he or she can no more budge than a salesclerk at CVS can haggle over the price of toothpaste. Course packets are custom-tailored anthologies, typically used for specialized seminars with small numbers of students. The material is assembled from a lot of different books and magazines, so no single copyright holder will reap a significant reward. Sure, the Copyright Clearance Center will handle the 'phone calls for the prof -- but they do not haggle over the price: If they did, they'd get the same response that I got. And the CCC has no more capacity than the permissions department gnome to negotiate intelligently over the relative value of, say, Silbey's chapter on the "Shrine of Party" versus my time in writing up a summary of the same.

The transaction costs are the result of what my former colleague and co-author Michael Heller would call a "tragedy of the anti-commons." (His recent and extraordinarily readable book, Gridlock Economy , provides illustrations in areas ranging from real estate to intellectual property). Lots of owners have small entitlements necessary for a more valuable assembled item, such that assembling the bits into a single valuable whole becomes a major headache. Like a land assembly, a course packet assembly can falter simply because the publishers cannot pool their resources to hire an agent with authority to bargain intelligently over a realistic price for their bits of prose. (Unlike land assemblies, the particular transaction cost is not strategic behavior but simple administrative costs of collecting and evaluating information).

Keep in mind that this price ought to be very, very low, if the publisher is economically rational. The publishers, after all, are absolved from the need to pay production, shipping, marketing, or administrative overhead: The prof handles all of that. The books' being out of print, any money that the publisher makes is pure profit, and the course packet itself is free advertising for the press and author. Academic presses are happy to move even 2,000 copies of a book that is expensive to produce and usually badly marketed. The notion that they would lose revenue from charging a low price -- say, a cent per page -- for chapters of out-of-print academic titles is goofy.

Sadly, the sort of frictionless world of anxious sellers and cagey buyers that Lichtman assumes in his scholarship has become the foundation for really bad "fair use" precedent in Princeton University Press v. Michigan Document Services, 99 F.3d 1381 (6th Cir. 1996)(en banc) (a case that Lichtman cites approvingly). Moreover, I doubt that the Congress will amend the statute any time soon: The classic Olsonian political economy of course packets is that publishers are well-organized and student-consumers are not, leading to Congress' refusing to touch preposterous results like MDS.

The rational solution would be to treat as "fair use" any course packet that used a fraction of a book so small that it would be implausible to expect the students to buy the whole book. Lichtman's response would be, I'd guess that such a doctrine might undermine some incentive for publishers to increase production to sell to a market that they are manifestly ignoring. When he produces some evidence that publishers have any interest in serving this market, I'll re-think my position that Lichtman is an apostle for pure deadweight loss.

But I've got no time to investigate the question right now: I've got to write up a summary for a course packet.

Posted by Rick Hills on August 29, 2008 at 10:47 AM in Property | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Why course packets ought to be fair use:


I agree with you wholeheartedly. At the very least, I'd like to see fair use standards relaxed for materials that are out of print. How can a publisher lose a sale if no one can buy the book/video/program in the first place?

Posted by: Justinian Lane | Aug 29, 2008 12:04:23 PM

It's my understanding, as well, that you could put the book on reserve and have the students each make their own copy and that this would be unquestionably fair use. It would just be much more of a pain for the students. Given this, and the facts above, it does seem quite dumb of book companies not to think more about this.

Posted by: matt | Aug 29, 2008 11:48:39 AM

Post a comment