The redundancy of originality Part 1
This article first appeared in 1994, in the International Review of Intellectual Property and Competition Law (published by the Max Planck Institute) and then appeared (in a slightly modified form) in the Entertainment Law Review in 1995.
Its basic thesis is that copyright is built on false foundations. It looks in particular at the Feist Decision of the US Supreme court, which decided that collections of facts lack originality and so are not protected by copyright.
The article might be old, but it is still very relevant. Arguably, more relevant as technology continues to put copyright under strain.
Because of its length, I have split the article into two parts. The footnotes are in part 2.
*IIC 658 Introduction
What should be done with facts? And what is the proper basis of originality? These two questions are inextricably linked because if, as predicated by the personality view of copyright, originality protects only created subject matter, then facts are not copyrightable. If, on the other hand, the correct basis of originality is the labour view then, generally, facts contained in works are protectable by copyright, at least so long as sufficient skill, investment and labour have been put into their collection.1
Facts are the battleground on which the opposing copyright armies most clearly reveal themselves, and it is a battle which the personality basis seems to be winning: in 1991 the US Supreme Court signalled its conversion to the personality view in Feist,2 and in 1992 the EC Commission put forward a draft directive on databases (DDD)3 in which it, unsurprisingly, espoused the personality view. Facts, according to the DDD, should only be protected by a right to prevent unauthorised extraction and re-utilisation.
What, if any, property rights should be granted in facts raises basic questions about the role and function of property in society, and more particularly about the role and function of originality. Originality is a test of propertisation:4 pass the test and there is property in X; fail the test and *IIC 659 there is no property (or less property) in X. But why have a test at all? We have only to look at property in tangibles to see a property system which functions adequately without originality or any equivalent test. Could it be that in the heat of the labour-versus-personality debate, the most central question - whether originality has any useful role - has gone unasked? This article asks just that question.
Part 1, "The Fruit Seller", discusses the role of property in making pre-existing materials available, and sets the scene for the remaining discussion. Parts 2, 3, and 4, examine three reasons commonly used to justify originality and withhold copyright protection from facts in works.
Part 2, "Facts, Copyright and Monopoly", looks at possible approaches to controlling monopolies in facts, and concludes that the right approach is to ask whether copyright in facts is likely to lead to monopolies that cannot be effectively controlled by antitrust rules. Since monopolies in facts, if and when they occur, can be controlled by antitrust rules, this justification of originality does not stand up.
Part 3, "Dissemination of Facts", asks whether the non-propertisation of facts (Feist), or limited property in facts (DDD), do indeed lead to greater dissemination, and concludes that they do not. The non-rival consumption argument is examined and rejected; production, dissemination, and positive-sum propertisation are looked at in turn.
Part 4, "Redundancy of Originality", looks at the notion of reward and incentive in relation to copyright, and concludes that copyright is neither reward nor incentive. It then examines three forms of originality: creativity, creation, and labour, and argues that, in relation to the first two, reward and incentive merely serve as a pretext to propertise on grounds of value. However, since it is the function of the market to allocate value, this pre-empts the market and inevitably distorts it. Nor is any solace to be found in the labour view of originality since, if its logic were ever applied, it would result in an unworkable economic system.
Part 5, "Conclusions on Originality", briefly reviews the preceding discussion, and concludes that originality, whether labour or personality, is redundant; worse still, it is an anti-competitive distortion of market structure. Originality distorts the market so that, compared to physical resources, the exploitation of intellectual resources is underdeveloped. Some empirical data is put forward to support this conclusion.
Having concluded that originality is best abandoned, Part 6, "Life Without Originality", sketches how copyright without originality would function, and points out some of its advantages, including a wealthier econ *IIC 660 omy, a unified theory of property, and a more coherent approach to the antitrust regulation of copyright.
Given the wide-ranging subject matter of this article, it is worth making the parameters of the discussion clear from the outset. Firstly, while both personality and labour bases of originality will generally protect the selection and arrangement of facts in a work, and thereby provide some indirect protection against misappropriation of facts, this article is concerned only with copyright in facts (or information) per se, irrespective of selection and arrangement. Secondly, common law copyright conflates the notion of originality with the notion of what constitutes a work. However, this article does not propose to discuss what is or should be a work (except to the extent that facts can constitute a work), nor how much protected subject matter is required to constitute a work, and reference to facts should be taken as referring to facts where they constitute or are contained in a work. Such a work might be a compilation, a historical work, a scientific work, or any other form of literary work containing facts (for example, a mainly fictional work which has some factual elements). Thirdly, this article is not intended to have any impact on what should or should not constitute substantial appropriation or fair use, except to the extent that it re-defines what constitutes protected subject matter. Fourthly, although this article takes facts as its main subject matter, the reasoning also applies to unoriginal works and unoriginal elements of works. Lastly, while a number of right-based arguments can be put forward in favour of collectors of facts, this article deals exclusively with economic arguments.
Propertisation of Pre-Existing Materials
At the heart of the personality view of originality is a dichotomy between creation and pre-existence. An original work is a work created by the author out of his own imagination, and by definition it does not pre-exist its propertisation.5 To the extent that the author uses pre-existing materials, he6 has no copyright in them: facts pre-exist and therefore cannot be original. An author who includes facts in his works does not create them, he simply appropriates them and therefore has no *IIC 661 copyright in them. The distinction between the created and the merely appropriated is explicit in Feist:7
facts do not owe their origin to an act of authorship. The distinction is one between creation and discovery: the first person to find and report a particular fact has not created the fact: he has merely discovered its existence.
[Original works] are founded in the creative powers of the mind … they are the fruits of intellectual labour.8
and is equally explicit in the EC Commission's definition of originality - the author's own intellectual creation.9 For both Feist and the DDD then, originality is a propertisation test: pass the test and there is property; fail it, and there is either no property (Feist), or at best a limited form of property (the DDD proposes that facts should be protected by a right against unfair extraction and re-utilisation).
But what is most striking about originality's ban on pre-existing materials is how unusual it is. For tangibles - the bulk of property, after all - there is no requirement of originality. Pre-existence is not a bar to propertisation, and nor, for obvious reasons, could it be. It could be said that tangibles have no need of a propertisation test: since most of the world is already property, the laws of tangible property need only be laws of transfer, not propertisation.10 But still, the property system for tangibles is not a closed loop, endlessly recycling. Goods do both enter it and leave it, and tangible goods that enter the property system necessarily pre-exist their propertisation. Given that tangibles do pre-exist their propertisation, it is worth asking what criteria should govern their propertisation.
*IIC 662 The Fruit Seller
Imagine that 500 miles away there is a valley in which fruit trees grow. Neither the valley, nor the trees, nor the fruit belong to anyone. Everyone is free to turn up and collect as much fruit as they want.11 Let us also suppose that an enterprising person drives down to the valley, loads up his truck with fruit, and brings it back to town. Should this person be allowed to exclude others from the fruit so that he can sell it on the market, or should those others be allowed to help themselves?
If we applied a personality analysis, no property right should arise because the fruit pre-exists. "The distinction is one between creation and discovery: the first person to find and report a particular fact has not created the fact: he or she has merely discovered its existence."12
But for most people the answer would be otherwise. The fruit seller should be given property rights in the fruit because everyone benefits. His property rights make it worth his while to collect the fruit in the first place, and we benefit because it is cheaper for us to pay that person than to drive to the valley ourselves. How can we be sure that these prices will be cheaper? Because, if not, we will collect the fruit ourselves; and as the fruit seller knows that is what we will do, he adjusts his prices accordingly. We are in a no-lose situation.13
This conclusion holds true whether the fruit is initially no-one's property or is initially everyone's property in common. And the more people there are collecting the fruit and selling it on the market, the cheaper the price of fruit will be. Even though the fruit seller had no a priori right to the fruit, propertisation of fruit is in everyone's interest because it gives the fruit entry to the market. The fruit seller does not physically make or create anything, he simply appropriates pre-existing fruit and transfers it from A to B. He makes the fruit more available.14
And, indeed, this is how it works in the real world. The mackerel in the open sea is no-one's property until it is caught, and the oxygen in the air is no-one's property until it is bottled by a gas company. Despite pre-existing in the Feistian sense, all three things pass - unproblematically - from the state of being unpropertised to that of being someone's property. Property, by bringing the market to bear, performs a function of making things available. It performs this role by protecting the partici *IIC 663 pants' investment, not only at the moment of first appropriation, but all along the distribution chain; in fact, without it, neither local nor international distribution and transfer of goods could be sustained.15 Property is the guarantor which makes the movement of goods possible.
In the eyes of the bourgeois world, the chief fault of feudal property is not in its origin (seizure and violence) but in its immobility - in the fact that it is incapable of becoming an object of mutual guarantees as it is passed from hand to hand in acts of alienation and acquisition.16
Are things any different for pre-existing intellectual subject matter? According to the Supreme Court in Feist, there is no point in copyright in facts because, "facts are part of the public domain available to every person".17
Except, of course, facts are not necessarily available. The Supreme Court equates the legal classification of public domain with the practical notion of availability: the two are clearly not the same. It is quite possible for a fact to be in the public domain, as defined by the Supreme Court, and totally unknown. Yet, according to the Supreme Court's Platonic view of the universe, the last 200 years of scientific development have been simply a matter of reporting what was already "in the public domain and available to every person." It makes you wonder why scientists bother with research.18
Since being in the public domain is not the same as being available, it seems that the parallel between fruit and facts holds true. Both pre-exist, both are in the public domain, but in both cases it might be cheaper to pay someone to collect fruit or facts than to collect them yourself; in which case, the first collector and any subsequent distributors will need exclusive rights in order to be able to secure their investment. Allowing some form of propertisation of the pre-existing serves the common good by allowing the market to perform the function of making available.
*IIC 664 We can now conclude that, given the way business works in a market economy, some form of protection for facts is required. And if appropriation (without any requirement of creativity) is sufficient to found property in tangible matter, why should appropriation not also be sufficient to found property in non-created intellectual property, such as facts? If there is no requirement of originality in physical property, why have it for copyright?
Before going on to consider the role and function of originality, this article looks at two reasons which, in relation to facts, are typically put forward in support of a requirement of originality. The reasons are, firstly, the prevention of monopoly; secondly, that facts are best disseminated by refusing, or reducing, property.
Copyright and Monopoly
There are two basic strategies for controlling monopolies.19 The first, and the simplest, is not to grant property at all. If there is no property in the thing, then the chances of monopoly are substantially reduced. This is the approach taken by Feist in relation to facts. The second strategy is to grant property but to regulate monopoly by a set of antitrust rules triggered by actual abuse of the monopoly position. This approach has two advantages: firstly, it allows the economy to reap the benefits of propertisation. Secondly, by emphasising abuse, it recognises that a monopoly is not a wrong in itself. Since we want firms to compete for the largest market share, it would be perverse to penalise firms which, by their skill, industry and foresight, have managed to build up sufficient market share to constitute a monopoly.
This suggests that, when considering whether or not to grant property to a particular class of goods (or elements of those goods), the correct question is not whether propertisation of X might result in a monopoly, but whether propertisation of X is likely to result in a monopoly which cannot be efficiently regulated by the normal rules on antitrust. It is only when the chances of unregulatable monopolies occurring are high that the no-property approach becomes a sensible alternative.
How should this be applied in relation to copyright? That copyright is often thought of as a monopoly right results from frequent, though erroneous, confusion between the notions of property and monopoly; the two notions are, however, quite distinct. Property entails a right to exclude others from the object of your property: monopoly, though gen *IIC 665 erally based on a right of property, goes much further than this; it requires market power. It is not because someone has the exclusive right to reproduce a particular work that he or she acquires market power, any more than an apple tree owner's ability to reproduce apples gives a monopoly over apples. That conclusion can only be reached by going through the usual antitrust analysis, in other words by defining the product market, the geographic market, and evaluating the availability of substitutes, actual or potential. Once the confusion between property and monopoly has been removed, it becomes clear that copyright is not inherently a monopoly right, nor even a monopolistic one, not least because a right, such as copyright, which allows independent generation of an identical product20 will not, generally, provide much protection against substitutes. This does not mean that copyright cannot give rise to a monopoly, because in certain circumstances it can, but it does mean that, since copyright is no more likely to form the basis of a monopoly than physical property, it should be subject to the same rules of propertisation. In other words, the optimal approach is to grant copyright, and then to subject copyright to regulation by antitrust.
Is the situation any different with copyright in facts? Or, to put it another way, is there something so inherently or inevitably monopolistic about copyright in facts in works that we should refuse them copyright? In Feist, the Supreme Court stated that "to accord copyright protection [to facts] distorts basic copyright principles in that it creates a monopoly in public domain materials"21 (italics added).
But is this plausible? There are two reasons, one minor, one major, for disregarding the reasoning in Feist. The minor reason is that substitutability cannot be determined a priori; some facts can act as a substitute for other facts, in which case there is no monopoly. An example will suffice. Suppose a firm wants to carry out research on the 50 largest cities in the UK, but plans to use a sample of only ten cities. If each of the 50 cities is in a position to supply the relevant facts, then no single city is a monopolist in relation to that particular market. This merely illustrates that the intended use of the facts, which cannot be determined at propertisation, is part of the product market definition.
The major reason for disregarding the reasoning in Feist is this. It is the nature of most facts to exist in a source which cannot be exclusively appropriated: viz. the world about us. Mere collection and compilation of such facts will rarely establish a monopoly because, where the facts are propertised by copyright, it is always open to others to go direct to *IIC 666 the original sources.22 However many compilations are produced then, so long as the real world source remains publicly accessible, there is, to use Locke's phrase, always enough left over. In more formal terms, what the analysis in Feist ignores is the distinction between source and version.23 A compilation of facts is a version: the facts themselves still exist in the source. London itself is the source for a list of all the companies based in London, while the list itself is a version. Including facts in a compilation ("copying" and "reporting" in Feistian terms) does not and cannot lead to the sort of monopoly envisaged in Feist, since the source of the facts continues to pre-exist and remains as available to second-comers as it was to first-comers. What is source for the goose is also source for the gander.24
The approach by version and source also suggests that copyright in facts is most likely to lead to some form of "monopolisation" where the source of the facts is not equally accessible to all. This will occur in two situations. The first is where the facts, though once publicly accessible, are no longer so; for example, the number of companies in London in 1900 is no longer available from the source (i.e. London in 1900), but only from versions. But even here there is no good reason to refuse copyright to facts; since the source was once available to all, refusing copyright because of a possible monopoly (rather than waiting for abuse of monopoly) serves only to penalise those who by their skill, industry, and foresight, have taken the trouble to preserve a version, and discourages the making of future versions.
The second situation where facts are not equally accessible to all is where the facts in question are created by an entity which, as a result of its own creation, enjoys privileged access25 to those facts. Typical exam *IIC 667 ples are railway timetables, telephone directories, and TV schedules. But facts or (information)26 which an undertaking supplies about its goods or services will only constitute a sufficiently distinct market to form the basis of a downstream monopoly where there is already a monopoly upstream. For, if there are one hundred TV broadcasters, it is hard to imagine a situation in which the supply of a TV schedule would constitute a distinct market. Or even if it did, the broadcaster would want to restrict dissemination of what is essentially marketing information. Since, when dealing with monopolies of creator-specific facts, the undertaking, as an upstream monopolist, will already be subject to the monopoly rules, it is not clear that there is anything to be gained by applying a different (non-propertising) approach to its downstream activities. This is particularly so because:
(a) Even if a personality version of originality is applied, the monopolist may still have copyright. In Magill,27 for example, the programme schedules were created by the BBC themselves, and it can hardly be claimed that the BBC appropriated pre-existing facts in any Feistian sense. It would also seem that the BBC showed sufficient creativity to have copyright in the schedules as a selection and arrangement.28
(b) The monopoly will exist independently of copyright, and so refusing copyright will not always be either a sufficient or a successful solution to monopoly abuse. In Magill (again), the TV companies would have been able to exert a monopoly over the weekly magazine publishers, irrespective of copyright, simply by choosing to release the programme schedule after the publishers' production deadline.
*IIC 668 Given the preceding discussion, let us return to the test set out at the beginning of this section. Are monopolies in facts likely? Since the vast majority of facts are facts to which there is equal access, the answer must be no.29 In the rare instances when monopolies do occur, can they be controlled by the normal rules on antitrust? Given that the answer is yes30 (and that, as outlined in a) and b) above, originality is not sufficient to control monopolies), using originality to restrain monopolies of fact simply throws the baby out with the bath water.
Free Flow of Information31
This section looks at the role of non-propertisation in dissemination. According to the Supreme Court, the dissemination of facts and factual works is best achieved by refusing copyright to facts.
The primary objective of copyright is "to promote the Progress of Science and useful Arts." Throughout history, copyright law has "recognize[d] a greater need to disseminate factual works than works of fiction or fantasy. [That] raw facts may be copied at will … is neither unfair nor unfortunate. It is the means by which copyright advances the progress of science and art."32
It is worth noting that the Court is putting forward an argument that is not primarily legal, but primarily economic. Given the goal (the progress of science, etc.), the Supreme Court's intention is to create a legal and economic structure which will lead to the greatest dissemination of facts.
But is refusing property likely to bring about greater dissemination? It is worth looking, initially at least, at how the no-property argument would apply to tangibles. Let us assume that overnight, property in manufactured goods has been abolished. Naturally, we all go down to the shops and help ourselves to manufactured goods without making any payment for them. In the short term, because we can all take these goods without fear of legal retribution, there is a dissemination of manufactured goods. The problem arises in the longer term, because the producers and resellers of manufactured goods have a reduced incentive to produce or to hold stock. If we do not recognise their property in the things *IIC 669 they make, then how can they take them to market? And if they cannot take them to market, then where is the incentive for them to produce? Ex post, after the manufactured good has been produced, reducing protection increases dissemination.33 However, seen ex ante, before the good has been made, abolishing property reduces the producer's reason to produce, and because production necessarily precedes dissemination, future dissemination is reduced.
Since abolishing property in tangibles is clearly not a successful economic strategy, it must be that the Supreme Court felt that, in economic terms, there are significant differences between tangible goods and intangible goods such as facts. This difference, presumably, is that intellectual goods are infinitely reproducible. Unlike tangible goods that have to be allocated between competing users, two people can simultaneously "consume" the same intellectual property without diminishing it (i.e. intellectual property is non-rival consumptive).34 By reducing the level of protection so that copying is allowed, the argument goes, everyone has more and no-one has less. Therefore non-propertisation is not only justified, but increases the amount of facts available to society. To understand why this conclusion is misleading,35 the question is best approached in two stages: production, and then distribution.
Production
If the only reason anybody desires an intellectual good is so that it can be used or consumed, then intellectual property is non-rival consumptive. But there is a class of people who are not primarily interested in intellectual goods in order to consume them, but because of the value they can obtain for the good. Typically, these people are producers, resellers, and investors. Because they invest in intellectual property as trading stock, dissemination is not for them a matter of economic indifference; since value is heavily dependent on scarcity, uncontrolled dissemination of the intellectual property literally devalues the good. For producers and resellers therefore, there is little difference between non- *IIC 670 propertisation of tangibles goods leading to a loss of physical stock, discussed above, and non-propertisation of intangibles goods leading to free dissemination. In both cases they lose the only thing that is relevant to them, and that is the market value of the good.36
As regards production then, there is little difference between non-propertisation of physical and intellectual goods. Ex post, after the good has been produced, reducing protection increases dissemination. Seen ex ante, before the good has been made, abolishing property reduces the producer's reason to produce and future production is reduced. The refusal of property in facts diminishes the producer's ability to gather revenue on the market, and will result in fewer new facts being collected and fewer new databases produced. On the other hand, although refusing property will mean less new facts and databases coming onto the market, the facts that already are on the market, or subsequently come on to the market, will be more freely accessible to all.37 Unlike the refusal of property in tangible goods, which results in a transfer of stock without any corresponding change in overall levels of stock, the refusal of property in intangibles, because infinitely reproducible, means that everyone can have more of that particular intangible.
Which, then, is better? To have more, new, facts coming onto the market, but for which you have to pay, or to have less facts but, to an undefined extent, for free? This - a more complex variant of the traditional jam today or jam tomorrow question - is, in economic terms at least, one of the two central questions of propertisation in facts.38 Yet it is one which both the Supreme Court in Feist and the EC Commission in DDD studiously avoid. Nor, in theory, is there an easy answer, because it all depends on the amount of facts sacrificed, on the price of facts on the market, and on the actual cost of access to free facts. Yet it is possible to reach some sort of conclusion. If, as the Supreme Court argues, the reason for non-propertisation of facts is to advance the progress of science (i.e. knowledge of the world around us), then to choose a system which gives priority to the recycling of existing facts rather than the discovery and collection of new ones is strangely perverse. And although assessments of future market behaviour are speculative, there are precedents *IIC 671 for assessing how effectively a reduction in the exclusivity of an intellectual property interest will increase dissemination so as to benefit society generally. In the former USSR, an inventor had no right to exclude others from his invention, and any state-owned firm could use, without a licence or payment, any other firm's invention. In other words, inventors in the former USSR received about as much protection as collectors and suppliers of facts do under Feist. Given the economic track record of the Soviet model, the Supreme Court's approach is unlikely to be successful.39
What about the right to prevent unauthorised extraction and reutilisation, put forward by the European Comission in the DDD? The right to prevent unauthorised extraction and re-utilisation is essentially copyright but with one major difference - the term of protection is only fifteen years, as opposed to copyright's usual life plus fifty.40 The DDD's approach represents a compromise. On one hand, it acknowledges commercial realities by providing for property in facts; on the other hand, by putting forward a reduced form of property, it pays lip service to the notion of no property in facts. But the arguments put forward against Feist hold equally, pari passu, for the DDD. Like Feist, the DDD's reduced protection sacrifices future production of more for a (hoped for) wider consumption of less. Given that the Commission believes that: "information plays a role of fundamental importance in the development of trade and industry, giving strength and coherence to the European economy as a whole, as well as being an essential component of both the cultural identity of the Community and the fabric of a modern society41 [italics added]" it is hard to see how it came to choose a policy which sacrifices long-term production. Indeed, its reasoning is opaque. In the Explanatory Memorandum to the DDD, the EC Commission justifies an unfair competition provision over a no-property approach by stating that "data … is similar to a raw material….In other industries it would be considered … an act of unfair competition for the raw material procured for processing at one company's expense to be freely appropriated by another company to make a similar product or service."42 But this proves too much. If one company appropriates another's raw material then, generally speaking, it is not unfair competition, it is theft.
*IIC 672 Distribution
Quite apart from the loss of production, it is often assumed, inter alia by the Supreme Court in Feist, that non-propertisation actually increases dissemination. This might seem intuitive, but is it really correct? There are a number of reasons why, compared to non-propertisation, property may actually increase the dissemination of intellectual property. Firstly, where there is property, resellers have an active interest in stimulating demand and increasing dissemination. Where there is no property, on the other hand, no-one has a legally protectable interest in doing so. This is simply a restatement of the preceding discussion: both producers and resellers have an equivalent economic motivation and are subject to the same economic rules. Moreover, in relation to distribution, the debate as to the appropriate level of exclusivity (and property is simply another form of exclusivity) does not take place at the level of property, where the benefits of exclusivity are taken for granted, but at the level of the territory. However, the role of territorial exclusivity in protecting against free riders and thereby stimulating competition is generally recognised. Witness the EC Exclusive Distribution Regulation which states that:
exclusive distribution agreements facilitate the promotion of sales of a product and lead to intensive marketing … they stimulate competition between products of different manufacturers … the appointment of an exclusive distributor who will take over sales promotion, customer services … is often the most effective way, and sometimes indeed the only way, for the manufacture to enter the market43
Secondly, the free dissemination argument assumes that, for the recipient, the costs of acquiring facts will be either nil or less than under a system with property. But this is not necessarily the case. Although intellectual property is cheaper to copy than to make, it is not necessarily cheaper to copy than to buy. Copying, even if illegal, is likely to be cheaper than buying when there is a plentiful and easily accessible supply of the product. However, for the reason given above, that is more likely to be the case where the relevant products are protected by rights of property. An example will illustrate the point.
Assume you want a copy of a CD. Under a system with copyright, you go to your local market (travel and time cost, say £2) and buy a copy for £10 (total cost 10+2=£12). Alternatively, you can buy or make a pirate copy, but this will be illegal - though cheaper. Under a system without copyright, you have the same choice - buying or copying - except that *IIC 673 both options are now legal. If you buy, you may well pay more than under a system with copyright. Producers and resellers will still attempt to protect their investment but will run up greater costs doing so.44 Or you can copy the CD. But where are you going to find the CD from which to copy? It may be that your search and travel costs make copying the CD more expensive than the £12 it would have cost you to buy the CD under a system with intellectual property.45
Whether it is cheaper to buy under one system than to copy under another is an empirical question, the answer to which will depend on a number of circumstances including, inter alia, the level of demand and the cost of non-property forms of protection. The important point, however, is that it is not safe to assume that non-propertisation leads to greater dissemination and, indeed, the empirical evidence (such as it is), suggests otherwise.46 If this is the case, then exclusivity, and its corollary, paying for access, are as inimical to the free flow of facts as they are to the free flow of tangible goods. And if the free flow of facts is important, then surely it is worth paying for - literally.47
Positive-Sum Propertisation
It seems clear that no property, or reduced property, in facts (and other unoriginal works), will result in reduced production. It also seems likely it will also result in reduced dissemination. But what also seems to underlie both Feist and the DDD is an argument which is not based on *IIC 674 economics, namely that facts are too important to be propertised in the normal way. But this begs the question - more important than what? Most people find food and housing to be more important than facts (certainly the sort of facts that feature in most databases), yet we are content to have exclusivity in those. Objecting to exclusion from facts on non-economic grounds, while simultaneously accepting exclusion from more important things such as food and housing, smacks of "let them eat cake". It also ignores something which we find self-evident in relation to other forms of property, and that the whole point of exclusion is not that people should be excluded, but to provide a bargaining counter which can be surrendered in return for the appropriate price.48
However, there is yet another argument in favour of the propertisation of facts. One of the basic assumptions of a market economy is that property is generally pro-competitive. Although, ex post, all property excludes, ex ante, it encourages production and distribution, with the result that, in the long term, more is available for everyone.49 And though property in tangibles is pro-competitive, property in tangibles is still a zero-sum game: because the global stock of tangibles is finite, every tangible that is propertised involves a depletion of a finite stock of public domain.50 Copyright in facts, on the other hand - where the source remains public and accessible to all on equal terms - is a positivesum game, simply because it is the nature of intangibles to be infinitely multipliable. A form of property which allows independent generation does not and cannot deplete the original stock.51 Not only will copyright *IIC 675 in the version not exclude from the public domain source, its effect is quite the reverse. Since every version of the source provides another route by which public domain facts can be accessed, the more versions of the source exist, the more available they are. And given a competitive market, the cheaper the public domain becomes. The logic of the situation is not to refuse copyright to facts in works, but to do exactly the opposite: to grant every compiler copyright in the versions that he makes. In a market where the participants are motivated by profit, granting copyright encourages a multiplicity of collectors and a multiplicity of collections.52 Copyright increases availability.53
Reward and Incentive
The preceding sections have discussed two reasons, prevention of monopoly and dissemination, for refusing originality to facts contained in works. This section looks at originality itself, at its role and function in the regulation of propertisation.
At the heart of copyright are the notions of reward and incentive:54 "The decision should not be construed as demeaning Rural's efforts in compiling its directory, but rather as making it clear that copyright rewards originality, not effort55 [italics added]".
The notions of reward and incentive are buried so deep in copyright that they are rarely questioned. And after all, they appear to make good sense. Where the author has clearly contributed something to society, he or she should be rewarded. On the other hand, where the contribution is minor, then the author should receive a lesser form of property, if any.
*IIC 676 But what happens if we apply the notions of reward and incentive to tangibles? For example, a baker bakes loaves of bread. If I told him that, as a reward, he could have property in his loaves, he would find me rather strange. If I then told him, that as an incentive to carry on baking, he could have property in any future loaves, he would find me stranger still.
Applying the notions of reward and incentive to tangibles produces odd results, yet it is an oddness which we tolerate quite happily when it comes to copyright. What it illustrates is that the function of property is neither incentive nor reward. Both are there, of course, occuring not at the level of propertisation, but at one remove, at the level of the market.56 And there they take the usual form - money. Having spent all day working, the baker does not finish with a contented sigh, thinking, "I have worked hard all day, and have reaped my reward in the property rights now vested in my bread"; he goes out and sells the loaves on the market.
Is the situation any different where the work is the result of an act of creation rather than appropriation? How can it be? Property is property, after all. An author may write a book in the expectation that his work will be copyright, but he certainly does not hope that the copyright will be the only reward. The primary function of copyright, like any other form of property, is not to provide a reward, but to provide a mechanism by which the reward can be appropriated. It "commodifies" the good and thereby makes it tradeable on the market. On its own, property has no value. If there is no demand for your work then, however solid your property rights, you are not going to get a reward. On the other hand, if your good does have market value, property makes it easier to appropriate that value, and will usually increase the amount of value that can be appropriated. It may seem that the distinction between property being a reward, and property allowing a reward to be appropriated, is small (after all, how much can depend on such a minor distinction?), but it is not. Because, once you assume that property is a reward, then the next step is to use property to distinguish between those works which you feel deserve to be rewarded, and those which you feel do not.
And this is exactly what originality does. Implicit in the notion of originality is a test. Some works - the original ones - pass the test: they will *IIC 677 be copyright. Other works - the unoriginal ones - will fail the test and will not be copyright. As the Supreme Court in Feist points out, if there is going to be a test, then it is necessary that some works should fail. "Given that some works must fail, we cannot imagine a more likely candidate. Indeed, were we to hold that Rural's white pages pass muster, it is hard to believe that any collection of facts could fail57 [italics added]."
But why should works fail? Why should there be one type of property in an original work and another in an unoriginal one? We do not find it necessary to distinguish between tangibles in this way. No-one would say, for example, that this fish is original and therefore of property type X, whereas this other fish is not so original and therefore of property type Y. Implicit in copyright's idea of reward are two things: firstly, the belief that some works (or some types of work) have more value than others; secondly, that this difference in value is relevant to propertisation. Originality, therefore, establishes a relationship between property and the value of the thing propertised. In so doing, it fundamentally misconceives the role of property in a market economy.
The next three sections look at different versions (or aspects) of originality; originality as creativity (primarily Inkasso (Collection Program)), originality as creation (primarily the author's own intellectual creation), and originality as skill, labour and investment.58
Originality as Creativity
The equation of copyright with value is seen most clearly in those systems that require creativity. For example, in Feist the Supreme Court stated that "copyright protects only those constituent elements of a work that possess more than a de minimis quantum of creativity."59 In Germany, at least where computer programs were concerned, the requirement of creativity was set much higher. In the Inkasso case, the Federal Supreme Court held:
The know-how of the average programmer, the mere craftsmanship, the mechanical/technical linking and assembly of the material, do not fall within the subject-matter of copyright. The minimum requirements of copyright are met only at a somewhat higher level: they presuppose a significant amount of creativity with respect to the selection, arrangement and organisation, as compared to general, average ability [italics added].60
*IIC 678 But if creative works have more value, then what sort of value is being referred to? Broadly speaking, there are two sorts of value, intrinsic value61 and market value. The usual argument against a creativity requirement is that it makes the law too uncertain. Given the difficulties we all have in agreeing on the value of something, how can judges be expected to determine intrinsic value, let alone apply it in a consistent way?62 But there is a much more fundamental objection to such a requirement; and that is directed not to the accuracy or even feasibility of measuring creativity, but to its relevance to propertisation.
In a market economy, the fundamental assumption is that the market is the most efficient allocator and generator of wealth. The market works by allocating value according to the price mechanism. Only one sort of value is generally speaking relevant, and that is market value which, in turn, can only be determined in one place: the market. This is the difference between a command and a market economy. In a command economy, the worth of goods, and their consequent production and allocation, is decided by a central authority: versions of originality which propertise on grounds of merit belong to a command system. In a market economy, on the other hand, the working assumption is that the optimal allocation and production is brought about by everyone acting in their own self-interest.63
Every individual endeavours to employ his capital so that its produce may be of greatest value. He generally intends neither to promote the public interest, nor knows how much he is promoting it. He intends only his own security, only his own gain. And he is led by an invisible hand to promote an end that was no part of his intention. By pursuing his own interest he *IIC 679 frequently promotes that of society more effectually than when he really intends to promote it.64
Propertisation on grounds of value is redundant because it is an attempt to determine, a priori, something that can only, and should only, be determined on the market. This entails that however obvious, uninteresting and generally worthless a work is, then - so long as the product is a work - the lack of value of a work should never be a ground for refusing propertisation because property is not the reward, it is the means whereby the reward is evaluated and appropriated. And the market is not only a much more efficient mechanism for evaluating value, it is also a self-regulating one. Not only will no-one bother to copy a worthless work,65 those who produce worthless works do not stay in business very long.
Market Value and Propertisation
Does this mean that only market value is relevant to propertisation? Not at all. Property is a conclusory term, and it is not because something is capable of being propertised that it should be propertised. Even in a market economy, we accept that it is within the legitimate ambit of a legal system to refuse property or restrict the market for reasons that have nothing to do with the optimal allocation or production of resources. As a generalisation, this is where the thing or activity is proscribed for reasons that may be called reasons of public order.66 But three things are implicit in a market economy: firstly, that these restrictions will be the exception rather than the rule; secondly, that unless something is excluded from the market system, it should be presumed to be included, since it is only by becoming property that something can gain effective access to the market, and thereby, ex hypothesi, be efficiently produced and allocated. It is worth noting that the combination of *IIC 680 these two conditions results in the existence of a vast category of goods that are tradeable.
The third condition is that, as regards that vast category of tradeable goods, the market should be undistorted. Producers and resellers should have equal access to the market, otherwise the market is distorted and allocation and production is sub-optimal. However, since unoriginal works and unoriginal elements are not proscribed or controlled for reasons of public order (but for reasons of value), unoriginal works belong just as much to the vast category of tradeable goods as do original works. This means that originality is not only redundant in that it attempts to perform a function that properly belongs to the market, it is also destructive. It distorts the market at its most fundamental level. Instead of the free play of market forces determining what is produced, by whom, for whom, originality ensures that producers of unoriginal works find it harder to appropriate the value of their goods on the market. A system of property, such as copyright, which distinguishes, on grounds of value, between members of the class of tradeable goods, usurps the function of the market and inevitably distorts it. The playing field is no longer level.67