Re: copyright expiration as a spur to creativity

On 7/31/98, Dan L. Burk <burkdanl[_at_]shu.edu> wrote:
>
> On 7/30/98, Michael Bradley <michael[_at_]vision-soft.com> wrote:
> >
> > Dan L. Burk <burkdanl[_at_]shu.edu> wrote:
> > >
> > > Consequently, consumers of the good should only have to pay for it
> > > to the extent necessary to induce the creator to create it in the
> > > first place.
> >
> > So you would also require manufacturers, say, to reduce the price
> > of their products once the development costs earn out, landlords to
> > reduce their rents once they've paid the mortgage, and lawyers to
> > reduce their rates once they're repaid their student loans? My god,
> > you're a Pinko!
>
> Oh, quite the contrary -- remember that if you go far enough to the
> right, you meet the left 'round the back way.
>
> If (I say *if*) the market is functioning properly, manufacturers,
> landlords, lawyers, and the rest should be naturally pushed by
> competitive pressure to lower their prices to marginal cost. But that's
> because they deal in congestible goods and services. Intangible goods
> are different because we have intervened in the first place -- by
> creating copyrights, patents, and so on -- to artificially inflate the
> price above marginal cost. If we're going to do that, we shouldn't do
> it any more than we have to -- and we generally don't have to past the
> point necessary to induce creation of the work.

   There's a publishing company that's been around for a long time that actually comes as close to Dan Burk's ideal publisher as anyone will likely see: the publisher does not pay royalties to its authors and does not have to factor in authoring costs when setting its prices, the publications are public domain, prices are set at nearly marginal cost, and the publisher has in place efficient mechanisms (including an Internet site and its own bookstores located at strategic spots around the country) for dealing directly with its customers, including individual buyers.

   The publisher is the U.S. Government Printing Office, which has a statutorily mandated pricing system. The pricing of GPO publications is governed by 44 U.S.C. sec. 1708 (reprinted here):

         The price at which additional copies of 
      Government publications are offered for sale 
      to the public by the Superintendent of 
      Documents shall be based on the cost as 
      determined by the Public Printer plus 50 
      percent. A discount of not to exceed 25 
      percent may be allowed to book dealers and 
      quantity purchasers, but the printing may not 
      interfere with prompt execution of work for 
      the Government.
         The Superintendent of Documents may 
      prescribe terms and conditions under which 
      he authorizes the resale of Government 
      publications by book dealers, and he may 
      designate any Government officer his agent 
      for the sale of Government publications under 
      regulations agreed upon by the Superintendent 
      of Documents and the head of the respective 
      department or establishment of the Government.

   The key to understanding the public pricing of a GPO document (such as the U.S. Code) lies in understanding the phrase "additional copies." About four years ago, when I inquired about GPO pricing, the GPO's public affairs office referred me to the GPO administrative and finance office, where I received the following explanation.

   When a government agency directs the GPO to publish a document, that agency bears through its budget all costs associated with preparing and printing that document up through the first copy printed or otherwise published. All copies after the first copy fall into the statutory category of "additional copies," the cost of which is based only on the actual additional cost to produce each copy; in effect, the cost of an additional copy is closely linked to the actual cost of materials (e.g., ink, paper, blank CD-ROM disc) required to produce that copy, to which the statute requires the Superintendent to add a 50% markup (mainly to allow as much as a 25% discount on sales to book dealers).

   So, for example, suppose an authorizing agency incurs costs of $100,000 to get the first copy of its document off the GPO presses, and the Public Printer figures the additional costs for subsequent copies at, say, $2.00 per copy. The authorizing agency "buys" the first copy for $100,000, and the public gets to buy any "additional copies" for the statutorily specified price of $3.00 ($2.00 plus the 50% markup of $1.00) -- except book dealers, for whom the statute authorizes a price as low as $2.25.

   In short, the price the public pays for a GPO publication reflects, essentially, only the incremental (i.e., marginal) cost for the publication, not the pro rata share of all costs of publication, and none of it is burdened by artificial restrictions imposed by copyright protection for the authors or publisher. Moreover, anyone can come along and reprint the GPO documents (which are public domain) without any payment to GPO for the privilege. That, it strikes me, is exactly the model that Professor Burk advocates for an efficient market in intellectual property.

   Despite the long-standing example of the GPO model, not a single private-sector publisher -- large or small -- has, to my knowledge, opted for this well-tested business model as its own preferred approach for conducting a successful publishing business -- at least not without the kind of heavy subsidy the GPO gets from Congress and other government agencies through their absorption of "first copy" costs. The repudiation of the GPO business model by the publishing world might not say a lot to others, but it tells me a great deal.

   In any event, the cost of books and other mass-produced intellectual-property items already reflects careful consideration of marginal cost. Publishers typically base the prices of their products on average cost, which, by definition, involves averaging the marginal cost of each unit and causes average cost to constantly move toward marginal cost (e.g., $5,000 for the first unit [which absorbs all the costs incurred to that point], $3 for the second unit, $2.50 for the third unit, $2.00 for the fourth unit, and $1.50 for the fifth and subsequent units, yielding an average cost of $1,001.80 per unit for a five-unit run, $6.50 for a 1,000-unit run, $4.00 for a 2,000-unit run, $2.00 for a 10,000-unit run, and so on).

   What intrigues me in this discussion, however, is not the sleight-of-invisible-hand diversion regarding marginal cost as some indication of the health of an intellectual-property regime, but the underlying assumption that the current system is somehow extremely inefficient and that a new, improved, fat-free, ultra-efficient mechanism will produce some intellectual-property nirvana. I think that's just dead wrong. The kind of efficiency envisioned in the ether of the academy will, I believe, produce devastating results if implemented here on earth, and yield an intellectual-property world that looks more like "Blade Runner" or "Mad Max" than anything else. Experiments run by IBM that simulate efficient commercial regimes similar to those posited on this list for intellectual property produced unanticipated destruction or preclusion of desirable economic relationships, with purely efficient economic actors, untempered by common sense or other human sensibilities, wreaking havoc as they went their unrelenting ways. (For a brief report on these simulations, see Kevin Jones, "Researchers Explore 'Bot' Downside," Inter[_at_]active Week, July 27, 1998, p. 32 [available online at <http://www.zdnet.com/intweek/print/980727/340974.html>]).

   In his book COMPLEXITY, Mitchell Waldrop quotes economist Brian Arthur as observing that "[a]ll too often, the apparent objectivity of cost-benefit analyses is the result of slapping arbitrary numbers on subjective judgments, and then assigning the value of zero to the things that nobody knows how to evaluate." I think Arthur's right on that score, and I think we're seeing similar dynamic her, with a lot of "[copyright-related] things that no one knows how to evaluate" assumed away as having zero value, and the remaining subjective judgments decorated in "the apparent objectivity" of theoretical economic analysis. We can certainly predicate intellectual-property policy on that kind of analysis, but I think the resulting intellectual-property regime will prove far from desirable, and much less welcome than anything we have now.

Chris Wren
Rice Lake, Wisconsin
cgwren[_at_]wisconsinlaw.com Received on Fri Aug 07 1998 - 04:38:27 GMT