Economic theory of law and the public domain -
When is Piracy Economically Desirable?
INTRODUCTION AND THESIS STATEMENT
Wealth creation today is driven by technological innovation. New technologies, not only in media but also in biotechnology, aerospace and electronics all present new challenges for legal regimes. In order to protect and encourage the creation of innovation a variety of legal protections have been created. Copyright law seeks to maximize the production and distribution of intellectual property in order to achieve equilibrium of supply and demand.
However, at least in the case of electronic media, these protections are overbroad. While they certainly do protect authors, at least in theory, these rights are often ill respected in practice. Further, if the extensive rights created were actually respected and enforced, societal wealth would actually be reduced because fewer ideas would be diffused.
This paper will try simply to illustrate using common tools of economic analysis equilibrium of supply and demand for cultural goods and how an overprotection of those goods can lead to a shortage of them, while an underprotection can also lead to a glut of poor quality. The paper will also illustrate that data searching and linking are as important if not more so than data protection.
The paper limits it s study to property based economic regimes. It does not adress whether proprietarian or communal systems of property diffusion are better at reaching equilibria of supply and demand.
In order to understand current US intellectual property law we must
first understand the Roman roots from which it ultimately springs. While
English common law is clearly the least faithful to Roman law tradition
it is also clear that many (though not all) elements of Roman law - including
the definition of custom, which is the source of authority of judicial
precedent - are in fact traceable to Roman laws or easily parelelled to
Civil Law systems (the regime of contracts notably). While other regimes
(torts and crimes) are rather clearly sui generis, fundamental notions
such as property, legal personality, presumptions, and even juries, can
be found in Roman or Civil law. This does not change the fact that common
law methodology is exactly backwards from that of Civil Law. Common law
judges reason from specific facts to general principles which are expressed
in binding judicial precedent is exactly backwards from civil law systems
where statutory laws, embodying general principles, are then interpreted
in specific cases which are determinative only of the case at bar. Still
so many legal elements are similar - and in some cases are the same - that
our analysis of contemporary US law of "public domain" must first begin
in an analysis of the Roman law of property.
I: Roman Law
A. Res Nullius
Res nullius are those physical things which " "have not or have never
had" an owner. Res nullius is a category of "things."(1) Those things have
not been reduced to "property" because they are not, or more accurately
cannot, be appropriated by individuals. Light, for example, is res nullius.
B. Res Communis
The property status of such a thing while it remains in a wild, unappropriated, state is "res communes," or a "thing common to all."(2) Unlike a res nullius which cannot be owned, res communes can be owned, and are owned, by the state, though a state may permit anyone to appropriate, and the Roman state did permit anyone to appropriate things which were Res Communis.
While modern law does tend to confuse the two, equating them, there
was a difference at Roman law. Res Communis was property of the state.
Res nullius was the property of no one. This helps us understand the later
confusion between public domain refering to federal lands, and public domain
refering to unowned information.
Customary law in the Roman, Civil, and Common law is nearly identical: Custom is a usage sufficiently adopted over time and which is believed to be and thus eventually becomes, obligatory. Custom is defined most succinctly as: "A usage which had acquired the force of law."(3)
Under common law custom however is required to have been historically dated from "time immemorial". So whether customary law could develop to govern new technologies is at least in the common law, uncertain.
We need to understand res nullius and res communis so that we may understand
what is meant exactly by the term "public domain". We need also to understand
what is meant by custom, in order to understand whether or not information
could enter into the public domain under customary law, either national
II: The Common law copyright regime
A. General Rule
Customarily, the common law provided limited protection for authors. The copyright was a monopoly, limited in time, granted to protect authors in the exclusive exploitation of their creative works. Thus at common law the general rule was no legal protection, with an exception for works which were both published and listed a copyright notice. Even then the work had to be creative and original and the duration of the copyright was limited.
By statute in the United States this presumption is reversed. Notice
of copyright and publication are no longer required. That is there is a
presumption of protection. Further the duration of protection has been
extended in time to the life of the author plus fifty years.(4)
1. Public Domain
Although the general rule of publication today is one of protection for creative writings under copyright, there are exceptions.
First, and most importantly, information which is in "the public domain" is not subject to copyright. Whether the public domain is statutorily defined as all that which is not subject to copyright - that is a definition in the negative(5) -oor is an emanation of the common law (and ultimately Roman law) conception of "res nullius" appears to be unanswered. For a monist it would appear that in fact public domain is nothing other than res nullius - though this issue is complicated by the fact that publicly owned lands in the United States are also called "public domain". Yet public lands in the United States clearly are not res nullius, and are in fact owned by the government.
When a copyright on a writing expires, it is said to enter into public domain. The question as to whether an author can cede his work into the public domain is also unanswered - one theory argues that the copyright is maintained though the author has given an unlimited license to the work. This question is unlikely to be litigated, since there is no evident contention, unless attempts to create a customary law of public domain force this issue to be litigated.
Another notable exception to the legal presumption of copyright is the surprising fact that US government publications are considered in the public domain and may be freely reproduced witout cost or permission by any person.
The granting of monopoly to authors can of course be criticized as contrary to free market liberalism.(6)
Thus arguments that allowing the appropriation of the expression of
ideas by individuals is economically desirable is in fact ill founded,
as this paper will show.
2. The Fair Use Exception to Copyright
Another major exception to the presumption of copyright is the "fair use" doctrine. This legal rule could be summarised briefly though somewhat inaccurately, as holding that "any author's writing may be used for criticism or academic scholarship". Fair use could also be expresed in terms of equilibrium analysis: when the benefit to society of breach of the authors monopoly outweight the benefit to the society of the author's monopoly plus the benefit to the author of such monopoly, the information may be used despite the fact that it would otherwise be the monopoly of the author. Both these definitions are provided only to sketch the concept, and also lay the groundwork for our economic analysis of copyright. Happily the court provides a more complicated but also more explicit definition.
According to the court there are four conditions which will determine whether or not a use is "fair".
"Factor #1: Purpose and character of use. The
courts are most likely to find fair use where the
use is for noncommercial purposes, such as a book
Factor #2: Nature of the copyrighted work. The
courts are most likely to find fair use where the
copied work is a factual work rather than a
Factor #3: Amount and substantiality of the
portion used. The courts are most likely to find
fair use where what is used is a tiny amount of the
protected work. If what is used is small in amount
but substantial in terms of importance - the heart of
the copied work - a finding of fair use is unlikely.
Factor #4: Effect on the potential market for or
value of the protected work. The courts are most
likely to find fair use where the new work is not a
substitute for the copyrighted work. "(7)
In the common internet case of copying, images or sounds are downloaded
and uploaded, sometimes edited, sometimes commented and sometimes linked
back to their source. Internet authors who "borrow" images without permission
but then "link" the image back to the source would be more likely to be
considered as "fair users" as would editors or comentators about images.
These guidelines are of course flexible and indeterminate. This permits
the court to decide whether a use is "fair" or not depending on the facts
before it. This sort of balancing test can be criticized as potentially
arbitrary, it is however all that is available for the court to develop
it s decisions when the court rejects categorical analyses which are currently,
and incorrectly, percieved as "mere formalism".
3. Economic valuation is problematic as to accuracy
The first point which must be made when making an economic analysis
of law is that economic evaluations of the worth of any good are always
approximations which determine the mean or median between an extreme low
and extreme high value. They are rough guides, though they are not as precise
as one might think. Healthy scepticism as to the validity of any statistic
is probably prudent. Though statistics are a useful analytical tool they
can, intentionally or not, be misused or abused.
III: The New Media
1. Economic Analysis
Our first economical analysis will expose the rationnale of the copyright law. Essentially copyright is predicated on the presumption that information production and diffusion are antithetical. Granting information monopolies will encourage the production of information - discovery or creation - but will discouragge the reproduction of that information - copying. On the other hand, allowing free reproduction of information - encouraging diffusion, would also discourage the creation through discovery or invention.
The relation between information quantity and quality of information as a "trade-off" is illustrated below
Production possibilities frontier for information
The difficulty in analyzing this market is partly due to the fact that
changes in available information have the effect of bulging the production
possibilities frontier. In other words, the information supply curve is
quite dynamically reactive. The effect of the addition of a new information
technology on the information supply curve is shown below:
In fact we are dealing with not one uniform good "information" but a
range of possible goods "high quality restricted information" or "lower
quality open information". If we nevertheless choose to analyze the market
for business reports, movies, television programs, multimedia cd roms,
sound recordings, and books as a single market we would have aggregate
supply and demand curves looking something like this
The problem which we assert is occurring in information supply is in fact shortages of information which do not need to be proprietary. Available information could be distributed more openly but is not in order to protec the creation of new information. However much of this information would be produced anyway. In other words although the technology to shift the production frontier of information to the right exists and has been used to some extent it is nevertheless underutilized.
One difficulty with any economic analysis of supply and demand is of course determining the point - or even points (in the case where two or more pareto optimal situations coexist) - of equilibria. This difficult problem is compounded by the fact that the market consists of millions, even billions of persons and transactions.
Determining the aggregate supply and demand curves is also rendered impossible in practice through the size and scope of the market. Today's information market is global in scope, and instantaneous. Information today can be produced and distributed globally at little or no cost beyond production cost. This fact, interestingly, tends to support the chicago school's efficient capital market hypothesis of perfect information and zero transaction costs. However while information may be instantly available at very low costs, finding and using that information is neither instant nor costless. Further the Chicago theory ignores the existence of false and misleading information, as well as the inevitability of some transaction costs such as legal niceties and translations. The Chicago theory also ignores capital market entry costs - although thin capitalization is admittedly the norm in the English speaking world, this is not the case in civil law jurisdictions. Further even if in theory a thinly capitalized company can enter the market with low cost real time information it is still limited by material capital requirements for production such as machinery, vehicles, and land. So while information costs are dropping all the time, transaction costs are not.
Fortunately information market imperfection is not fatal to our juridical analysis. The problem of economically inefficient information "protection" as a hindrance to efficient information markets, theoretically illustrated at the macro level, can also be practically demonstrated at the micro level.
Our point, though illustrated by economic analysis and represented through
the traditional graphs is in fact demonstrated juridically through a consideration
of the practical realities facing producers and distributors of information.
In order to understand the new media marketplace we need to appreciate certain facts about technology. Computer software can be first studied as either data - raw information such as personal names, adresses bank balances - mere facts, and algorithms used to store, transmit, or manipulate that information. Data - that is, raw information - cannot itself be the object of copyright ssince there is no original creation. Algorithms are more complex however. A computer algorithm is generally written in a "source" language (today most commonly C++ though many others exist). This "source code" (which generally somewhat resembles English or mathematics) is then compiled into "object code". Object code is simply strings of ones and zeros - data - organized in such a way as to send instructions to and from the computer's microprocessor to manipulate data.
At first there was some question as to whether computer programs - algorithms - should be subject to copyright, patent, or both. Today it is possible to either patent or copyright source code, object code, or both. A still unsettled question is whether the "look and feel" -also known as "trade dress" of the resulting program can itself be copyrighted. In general software authors should not rely upon an assertion of "trade dress" in the "look and feel" of their graphical user interface; in the heavily litigated case of Apple Computer Inc. v. Microsoft the court ruled against Apple's claim against Microsoft for copyright infringement on the ground however that an software contract between Apple and Microsoft was sufficiently ambiguous as to be interpreted to authorize Microsofts acts. While this case is therefore less than perfectly unambiguous, it is also clear that if the court wished to protect a software company for the "look and feel" of its graphical user interface it would have done so - and in our opinion to the detriment of the public, since such monopolistic guarantees limit the diffusion of information and do little, if anything, to encourage the creation, invention, or discovery of new information.
Another unclear case - prior to the recent "Millenium" act - was decompilation.
Decompilation consists in taking object code - zeros and ones - and disassembling
it into machine code - mnemonical instructions which while not as readily
understood as uncompiled commented source code can nevertheless be understood
by humans. Decompilation, which was perfectly legal, unless in contravention
of a software's licensing agreement, is now considered a crime, unless
done in pursuit of scientific software research undertaken to further security
and cryptography - a sort of "fair use mini exception" to the criminalization
of what is however generally innocent study. Further even if decompilation
is used for "reverse engineering" the resulting infringment was already
punished - though not as a crime - under the old law. In other words, the
neww law goes to far creating an unneccesary aura of paranoia and insecurity.
The criminal punishment forseen in the Millenium act far outweighs the
"crime" of studying source code.
3. The Internet
The first area of new media which we explored is software - programs used by persons in the privacy of their own home. The second area we will study is the internet.
Essentially the internet "shifts" the information production possibilities frontier dramatically to the right, allowing for the low cost instant global diffusion of information. Further while there is now a vast ocean of information, most of that information is true and accurate, and almost all of the information is rather easily searched. In fact there is so much information freely available that the real task of research is not finding the information so much as it is culling out irrelevant information. Happily the same technology which makes this vast ocean available also allows us to search it. So while we are swamped in data, we do at least have hip waders! However advertising on the internet does increas the amount of "chaff" that one must winnow to get at the "wheat" – that is the relevant data.(8)
The internet can, and should, be seen first as a gigantic planetary public library and only second as a commercial space since most commercial services are proprietary. More simply put, why clog limited bandwidth which should be dedicated to scientific research and discovery with personal information?(9) However logical this argument may be however, the fact is clear that the internet is irrevocably now both a commercial and scientific space - and more the former than the latter.
Well that is the technological side. In terms of the copyright side however, this vast storehouse of information, once downloaded cannot - legally - be uploaded or otherwise redistributed without either the permission of the author or by relying on one of the "fair use" exceptions.
Our position is that this assymetry between restrictive laws, and open
information should be decided in favor of open information freely available
for all persons use and against closed information. Further this argument
is based on sound economic reasoning.
4. Why have presumption of public domain if diffused on internet?
Our argument is that information diffused on the internet - if legally placed there - should be presumed to enter the public domain and be, in law as well as fact, freely redistributable. Our argument for this radical position is essentially economic: allowing the free reproduction of information - where such is already technologically possible - maximizes wealth not only of society but also of individuals in society. Information which is not distributed is as useless as any other undistributed good. Information is different from physical goods in that it cannot be consumed, only used. True and accurate information actually increases in value as it is diffused and used.
Essentially the conflict is between the public interest in the optimum mix of low cost and high quality information and individual creators or appropriators of information who are rent seeking. Favoring the latter can lead to sub optimal situations of semi monopoly. Such rent seeking is best illustrated where each individual injury is de minimis, yet teh sum of these injuries over thousands or millions of individuals translates into millions of dollars. The transaction costs of litigating each individual persons interests are far higher than the potential gains to individuals, whereas the costs to the appropriator to maintain their monopoly is much less than the benefit which accrues to them because of that monopoly. Rent seekers thrive off of suboptimal economic situations, and proprietary information encourages this (for example, domain name squatting, also the "anti" virus business, or even piracy-for-profit are all examples where transaction costs encourage sub optimal outcomes).
Another argument in favor of the free use and copying of information on the internet - is that such copying, even if not falling within the fair use exception - is already widespread in practice. This is essentially the argument that free copying is a practice sufficiently widespread as to constitute a usage, and which is believed legal, and thus becomes a customary law. Juridically however this argument is weak: custom canot be asserted in opposition to a statute to the contrary. Thus while this author does observe the practice of wholesale copying on the internet, more often than not made with the innocent but mistaken belief that such copying is legal such a usage has not yet ripened into customary law because a contrary statute exists. Further even if the existence of the statutory copyright could be avoided through a very broad interpretation of fair use, a claim of customary open source availability of internet information would also face the question as to whether all custom must date from "time immemorial" - the correct position, at least under common law - or whether "new" customs can arise - which is in fact the case of international law. So the argument that information should be legally free in theory because it is in practice while intriguing is not valid - at least not yet.
Scripting a CGI or Applet to guarantee the proprietary nature of information
is no more an undue burden on those who wish to appropriate information
than the cost of building fences, walls, locks, and other methods of protecting
and preserving exclusivity in one's property. While CGI and Applets are
is not that great. Thus when weighing the interest of the public in having
low cost, high quality, accurate information against the interest of private
appropriators in rent seeking to take advantage of suboptimal conditions
created by transaction costs, it seems rather clear that the burden of
elaborating CGIs and Applets (which could then be sold) rightly belongs
on the shoulders of the appropriator - and that the freedom of information
rightly belongs to the public. Information should be free, for only then
can it be useful. If information appropriators wish to seek rents they
can do so quite easily by using closed source methods such as applets and
CGIs. Further protection can be added by creating systems of passwords,
authorizing access only to subscribers - such as AOL, one of the most proprietary,
and succesful, internet services.
This paper has argued that information should be free not for idealistic reasons of human self development or romanticism - though these arguments are superficially aattractive. It has argued that free information makes sense economically and translates into greater productivity and more efficient resource allocation. This paper has shown using economic analysis, both graphic and algebraic, that the cost of guaranteeing the proprietary character of information rightly belongs on the shoulders of the appropriator, and that in any other case it is simply more economical to allow free information distribution.
Whether these economic facts evolve into customary international law - for internet transactions are in fact very often international - is an open question. Since not every statee is a signatory to the Berne convention one can forsee states which will refuse to adopt the US sponsored proprietary model of information. States which offer secure private servers will be able to profit from those providers of information who wish to avoid regulation or even taxation. Consequently the universal character of the Berne convention could become undermined because of the overly strict proprietarian character of current US law, notably the "millenium" act, but also recent the attempts at creation of an international convention on information "crimes" - which is even more draconian than the "millenium" act. These laws are not only overbroad in their definition of property, and overly harsh in the criminalization of thought - they are also uneconomical.
1) Geer v. Connecticut, 161 U.S. 519 (1896)
2) Geer v. Connecticut, 161 U.S. 519 (1896)
5) "In many ways, our current conception of the public domain is that nobody affirmatively owns public domain materials. It is this unowned characterization that is somewhat at odds with a characterization of the public domain of intellectual materials as a commons."
Neocolonialism, Anticommons Property, and Biopiracy in the (Not-So-Brave) New World Order of International Intellectual Property Protection Keith Aoki
6) "First, one should note that there is a deep contradiction between the definition of an 'intellectual property right,' that is, a state-backed monopoly handed out to individuals or firms, and the popular neoliberal vision that valorizes "privatization" and free market economics."
"Neocolonialism, Anticommons Property, and Biopiracy in the (Not-So-Brave) New World Order of International Intellectual Property Protection"
7) "Intellectual Property Law Primer for Multimedia Developers"
(1994) by J. Dianne Brinson and Mark F. Radcliffe
8) Any internet researcher is doubtless familiar with the "Amazon" phenomenon - searches for almost any online papers or books on any academic topic will eventually return bookshops selling commercial information. Which is good if you are modestly funded academic looking for rare or out of print books, but bad if you are limited to free information or information supported by advertising.
9) Available bandwidth is shrinking due to more and larger data files (graphics, are notoriously large, and a 3 minute silent move can take as much as 3 megabytes). On the other hand cable is emerging as a solution, allowing rapid transfer of vast amounts of data. Unfortunately however cable is not free, and is likely to remain a subscription based service. This situation is however tolerable so long as scientific information – which is mostly writing and occasionally graphics – remains freely available. One can make a cogent argument that internet tv, phone, and interactive movies are luxuries and as such should not be free – unlike scientific research which advances the collective well being of humanity, and thus should be free.