Copyright (c) 2004 The Board of Trustees of Leland Stanford Junior University

Stanford Law Review

November, 2004

57 Stan. L. Rev. 485, pp. 545-68

ARTICLE: Reform(aliz)ing Copyright

NAME: Christopher Sprigman*

III. Reformalizing Copyright

A. Defining "Interoperable" Formalities in the Berne Convention

Berne's prohibition of formalities dates from 1908, a time in which requiring authors to comply with formalities in the many countries in which a work may be published - i.e., requiring an author (or publisher) to inform himself about the requirements of the law in countries with which he has no [*546] familiarity, and then to obtain and fill out forms in a variety of languages - would be difficult, expensive, and often result in unintentional noncompliance and the loss of valuable rights. Article 5(2) of the current Paris Text of the Berne Convention was promulgated in 1971, but the nature of the problem had not changed in the intervening sixty-three years: copyright systems remained substantively and procedurally diverse, the mechanisms of compliance in many countries remained balky, and the costs of informing oneself about requirements in different countries, and then complying with them, remained high.

Since 1971, however, there has been a series of technological changes that could make compliance with a redesigned set of formalities quick and easy. Those changes involve, of course, computers and the Internet. Now it is possible for an author publishing a work internationally to comply with formalities in his or her home country, or in the country of a work's first publication, and to have the data generated by that compliance formatted and transmitted reliably and nearly costlessly to other jurisdictions. But changes in technology alone are not enough - changes to the law are also required. To make compliance cheap, the law must ensure that data generated in one jurisdiction will be sufficient to permit compliance in any jurisdiction that chooses to reintroduce formalities into its domestic copyright laws.

The simplest way to take advantage of what technology now allows would be to propose a new Berne text that removes the prohibition in Article 5(2) of the current Paris Act and replaces it with a provision allowing member countries to impose formalities, provided that they adhere to a set of standards that make formalities "interoperable" across jurisdictions. n227 What would the changes to the Berne Convention look like?

1. The reciprocity principle

The most direct approach would install a "reciprocity principle" alongside the existing national treatment and minimum standards principles that now drive Berne. The reciprocity principle would require that all Berne jurisdictions [*547] that impose formalities permit foreign authors to comply with formalities in their national laws by complying with formalities either in their home country or in the work's country of first publication or registration. The reciprocity principle would not require any particular Berne nation to impose formalities - i.e., it would not modify Berne's current minimum standards requirements. It would, rather, require only that Berne nations that choose to reformalize their domestic copyright laws do so according to standards set out in Berne.

Some nations may, of course, choose not to reintroduce formalities into their domestic law. But if some Berne Union countries have formalities, and others do not, the possibility arises that the home country of an author, or the nation of first publication of his work, will not have a registration requirement to which other Berne countries with formalities can grant reciprocity. To accommodate authors in this category who wish to comply with formalities across Berne jurisdictions, the Berne nations should also establish a centralized WIPO registry, subject to the same standards agreed upon by Berne members and subject also to the condition that all Berne nations will grant reciprocity.

To make the reciprocity principle practically workable, Berne signatories would enter into a side agreement that would standardize across jurisdictions the data required to register a copyright and standardize formatting of that data so that registration information - authors' names and addresses, creation and registration dates, etc. - may readily be shared among jurisdictions. Berne signatories could then establish an information-sharing agreement whereby registration data obtained in one country could be made available to other jurisdictions, at the rightsholder's discretion. As the secretariat for the Berne Union, WIPO would be well placed to coordinate the actual transfer of data among Berne members.

Taking this approach, it is not necessary to amend Berne to prescribe a minimum set of formalities. It would suffice, rather, simply to remove the Article 5(2) prohibition, to install the reciprocity principle, to work out a set of standards to ensure interoperability, and then to leave to the member states the decision whether to reinstall formalities or not.

2. The reciprocity principle and neighboring rights agreements

For countries that, unlike the United States, do not include protection for performances, sound recordings (also referred to as "phonograms"), and broadcasts in their copyright laws, but locate them instead in separate "neighboring rights" statutes, reformalization of domestic law must include changes to the law governing both types of rights. Similarly, the same reciprocity principle that would be installed into the Berne Convention must also be introduced into the applicable international agreements governing neighboring rights - an issue of some complexity.

[*548] The principal international agreement defining protection of neighboring rights is the 1961 International Convention for the Protection of Performers, Producers of Phonograms and Broadcasting Organizations, also known as the Rome Convention. n228 This instrument extends Berne-style national treatment and minimum rights principles to neighboring rights, although the minimum terms established are shorter. n229 Unlike Berne Article 5(2), Article 11 of the Rome Convention does not prohibit signatories from conditioning protection of neighboring rights on formalities. Article 11 provides, however, that any signatory that conditions protection for performers on, or producers of, phonograms on compliance with formalities must permit its requirements to be met by affixing a prescribed notice to the recording or its container. Adoption of the reciprocity principle, therefore, would require replacement of the rule allowing blanket compliance through notice with formalities pertaining to protection of phonograms.

The 1973 Convention for the Protection of Producers of Phonograms Against Unauthorized Distribution of Their Phonograms, referred to as the Geneva Phonograms Convention, which is aimed at cross-border record piracy, requires signatories to protect qualifying phonogram producers "against the making of duplicates without the consent of the producer and against the importation of such duplicates, provided that any such making or importation is for the purpose of distribution to the public, and against the distribution of such duplicates to the public." n230 Like the Rome Convention, the Geneva Phonograms Convention allows signatories to impose formalities as a condition of protection for phonograms, but provides that affixation of notice must suffice to comply with all mandatory formalities. n231 Thus, the same changes that would be required to the Rome Convention must also be applied to the Geneva Phonograms Convention.

Now that we have seen what changes to international law are necessary to shift the treatment of formalities from hostility (Berne), or at best, grudging acceptance (Rome and Geneva), to acceptance with reciprocity, let us briefly [*549] examine how such a system would work with respect to each of the familiar types of formalities.

3. The reciprocity principle in practice

Registration. The application of the reciprocity principle to the registration formality is comparatively straightforward. Once a work is registered in one jurisdiction (or with the centralized WIPO registry), it would be registered in all Berne Union jurisdictions that have reinstalled a registration formality in their domestic law.

Recordation of transfers. The same scheme established for registration should also apply to recordation of transfers - a transfer that is successfully recorded in one jurisdiction (or with the WIPO registry) should suffice to record that transfer in all jurisdictions in which the work previously has been registered.

Notice. Berne Union nations would be free, under the reciprocity principle, to require that notice be given for some or all works. Of course, if a Berne nation creates an effective, easily accessible copyright registry, there is little to be gained by also requiring notice: the registry should provide enough information to make the tracing of copyright ownership simple and cheap. n232 In the instance, however, that some Berne signatories choose to include a notice requirement in their reformalized domestic law, the signatories should agree to standardize the form of notice for different types of works to ensure (1) that no more information is required to be elicited to comply with notice requirements than was supplied to complete registration, and (2) that the same form of notice that suffices in one jurisdiction for any particular type of work will also be accepted throughout the Berne Union. These rules would prevent differing standards for notice that might cause unintentional loss of rights. They would also encourage publication with the standardized form of notice even in those jurisdictions that do not require it as a condition of protection.

Renewal. Creating a renewal formality that is interoperable across jurisdictions raises a number of somewhat more complex problems, but should be achievable with an increased level of coordination among Berne members. The first problem is whether, in order to permit Berne nations to reinstall the renewal formality, a revised Berne Convention would have to remove the provision in the current version of Berne requiring all signatories to grant a [*550] minimum copyright term of life of the author plus fifty years. n233 The answer is likely no. Berne's prohibition of formalities is contained in an article separate from its minimum term requirements. Remove the current ban on formalities, and nothing in the Convention specifies that the term, if offered equally to every author, must be enjoyed in full by every author without condition.

The second problem is a more practical one: how to coordinate renewal across jurisdictions when different Berne member states may impose renewal requirements at different points in the copyright term. A simple application of the reciprocity principle threatens to create substantial confusion. If a rightsholder who complies with the renewal formality in the jurisdiction in which a work was first registered is deemed to have complied with renewal in any jurisdiction in which a renewal is required, then absent detailed knowledge of the point at which renewal may be required in a potentially large number of Berne jurisdictions, a would-be user will find it difficult to determine whether a work has been timely renewed.

For example, suppose that a work is first registered in country A, which imposes a renewal requirement at thirty years. Twenty-five years into the work's term, a would-be user in country B inquires whether the work is in the public domain. Country B imposes a renewal formality at fifteen years. The user sees that the work was registered twenty-five years ago; under country B's law, the work would have passed into the public domain when the rightsholder failed to timely renew. But under country A's laws, the work is still in its initial term; renewal will not be required for another five years. Accordingly, under a simple application of the reciprocity principle, unless the user understands (1) where the work was first registered, (2) when the renewal requirement occurs in that jurisdiction, and (3) that country A's renewal requirement is the relevant one, the user will not easily be able to determine whether the work is in the public domain.

These information problems can be mitigated, of course, even if they cannot be eliminated. The standardized registration and notice format should include information on the nation of the work's first registration, and that information should be made available in all online registry sites maintained in the various Berne nations. In addition, Berne signatories should be encouraged to disseminate information about the rules governing renewal, and how to determine which renewal term applies to a particular work. Taken together, these measures might mean that the benefit, in terms of the simplicity of a straightforward application of the reciprocity principle to renewal, outweighs the cost in terms of the increased complexity of determining the status of rights.

An alternative, which would require a greater degree of coordination among Berne nations, would be to standardize renewal terms for all jurisdictions that reinstall a renewal formality in their domestic law. The [*551] reciprocity principle would then apply to grant automatic compliance with all Berne nations' renewal requirements based on timely compliance with the requirement in the country of first registration, or by renewing with WIPO, if original registration was made with the WIPO registry. Based on the depreciation calculations made by Landes and Posner, the Berne nations could impose more than one renewal obligation during the copyright term. A first renewal obligation set at ten years would move approximately 50% of registered works into the public domain. A second renewal requirement set in the vicinity of forty-three years would result in only 10% of the number of originally registered works remaining under copyright. A third renewal requirement set at sixty-five years would move all but 1% of the number of originally registered works into the public domain. The works left under copyright after sixty-five years would be those of truly enduring commercial value for which the full term of copyright would be likely to provide significant continuing benefits.

B. Defining "New-Style" Berne-Compliant Formalities

If changing Berne to explicitly permit formalities is not possible, is there still a way to reformalize U.S. domestic law? There are several alternatives of varying merit. The next few pages will first briefly discuss two long-shot possibilities: the reintroduction of formalities for U.S. (but not foreign) authors and U.S. withdrawal from Berne (which, for reasons that will quickly become obvious, this Article does not recommend). I then focus on an alternative that seems much more sensible: the reintroduction into U.S. law of "new-style" formalities that provide the benefits of traditional formalities, but that do not run afoul of Berne's proscription of conditions that interfere with the "exercise and enjoyment" of copyright.

1. Reintroducing old-style formalities for U.S. authors

Because Berne does not prevent signatories from imposing formalities on the works of domestic authors or authors from non-Berne signatory nations, the United States could have retained a full set of traditional formalities for those works. In fact, an advisory group established by the Department of State to assess what changes to U.S. law would be necessary for Berne accession advocated this position, as part of a more broadly minimalist approach to the implementation of Berne that sought to alter only those portions of U.S. law that the group deemed clearly irreconcilable with the Convention. n234

[*552] The minimalist approach of restricting unconditional copyright to foreign authors is, for reasons that are not difficult to imagine (e.g., the antipathy to granting foreign authors more rights than U.S. authors), not the approach that Congress took. Whether the politics of copyright are likely ever to shift in a way that would make the minimalist approach to unconditional copyright viable is a question beyond the scope of this Article. It is worth noting, however, that restricting unconditional copyright to foreign works would represent a significant improvement on the status quo without creating any risk of noncompliance with Berne.

2. Withdrawal from Berne and reliance on the Universal Copyright Convention

Because the United States is a signatory to the UCC, and because before it acceded to Berne the United States negotiated bilateral copyright agreements with several nations that were not UCC signatories, it would be possible for the United States to withdraw from Berne and rely instead on the UCC, which, unlike Berne, allows the imposition of formalities for the works of both domestic and foreign authors. This strategy would, however, impose unacceptable costs, the largest of which would arise from our resulting noncompliance with the TRIPs accord, which incorporates by reference Berne's standards, and with the North American Free Trade Agreement (NAFTA), which replicates Berne's ban on formalities. Although its applicability is subject to considerable debate, it is also possible that a provision of Berne's Appendix Declaration would prevent U.S. authors from claiming the benefits of the UCC in countries that are Berne signatories. n235

A further cost of withdrawal would arise from Berne Article 6(1), which permits Berne nations to restrict the protection accorded to works of authors who are nationals of a non-Berne country that "fails to protect in an adequate manner the works of [Berne nationals]." n236 There is little commentary on this provision, so it is difficult to forecast whether subjecting foreign works to formalities (at least formalities that do not discriminate between domestic and foreign works, and for which compliance is easy and cheap) would rise to the level of a "failure to protect in an adequate manner" the rights of foreign authors.

[*553]

3. Indefinitely renewable copyright

Landes and Posner have proposed a system of indefinitely renewable copyrights - i.e., a perpetual copyright term, conditioned on periodic renewal. n237 Landes and Posner suggest that such a system would result in more works entering the public domain more quickly; their conclusions in this regard are very likely correct. The Landes and Posner proposal is subject, however, to two critiques: the first is significant, and the second, for my purposes, is determinative.

First, a system of indefinitely renewable copyrights would prevent any work of enduring commercial value (many of which would also have important cultural value) from ever entering the public domain. For reasons explained above, extending copyright indefinitely for valuable works raises the cost of transformative use of these works and would give rightsholders a perpetual veto power over uses they don't like. These cultural and First Amendment costs are not balanced by countervailing benefits. Because the current regime of limited but very long copyright terms gives rightsholders virtually the same return (from a net present value perspective) as would be produced under a perpetual term, a shift to perpetual copyright for valuable works would yield no significant enhancement to the incentive to create.

Landes and Posner also discuss a series of limited-term options conditioned on repeated renewal requirements. These avoid the first objection, but they do not avoid the second: because they employ an old-style renewal formality (i.e., one that results in termination of rights for failure to comply), and because none of the proposals would guarantee a minimum term of life plus fifty years, all versions of the Landes and Posner approach would require the United States to withdraw from the Berne Convention. n238 The authors make note of the incompatibility of their proposal with Berne, n239 but their concerns, unlike mine, are focused solely on the economic effects of the proposal, not on its consequences for U.S. participation in the international copyright system. n240

[*554]

4. The Public Domain Enhancement Act

Another possible approach is set out in a bill currently before Congress, the Public Domain Enhancement Act (PDEA). n241 Sponsored by Representative Zoe Lofgren, a California Democrat, the PDEA would give copyright owners of works by U.S. authors unfettered rights for fifty years. At that point the copyright holder would be required to file a notice of continuation and pay a $ 1 fee every ten years to continue the copyright. Because only a small number of works would retain any commercial value at the expiry of the minimum term, most copyright owners would not bother to file a notice of continuation and pay the fee. On September 4, 2003, the PDEA was referred to the House Subcommittee on Courts, the Internet, and Intellectual Property. There it has languished.

Unlike the Posner and Landes proposal, the PDEA is very likely compatible with Berne. The renewal requirement is limited to the works of U.S. authors, thereby avoiding conflict with Berne's rule against formalities. Although the renewal provision may cut off a work's copyright prior to the expiration of Berne's minimum term, that should not, for the reasons given above, cause Berne noncompliance. n242 Nonetheless, the PDEA is vulnerable to the critique that its effect is limited to tinkering around the margins: A large percentage of works are commercially valueless at inception or have an initial value that is quickly depleted. All of these works, however, would continue under the PDEA to be subject to a very long copyright term. While fifty years is certainly better than life plus seventy years, it may reasonably be asked whether the game is worth the candle.

5. New-style formalities

A fifth option, and by far the most attractive, is to formulate and install in U.S. law a set of new-style formalities that apply to both domestic and foreign works. New-style formalities would provide the filtering and information-creation benefits of traditional formalities. However, there is a good argument that, if structured properly, new-style formalities would not affect copyright's [*555] "enjoyment and exercise," and would, therefore, comply with our Berne obligations under the current Paris Act. This approach is attractive because it would require changes only to U.S. law; Berne, TRIPs, and the other international agreements that govern copyright and neighboring rights would remain undisturbed. Integrating new-style formalities with the current text of the Berne Convention does, however, raise several significant questions.

First, exactly what is a condition that interferes with the "enjoyment and exercise" of copyright? A solid starting point is that that language at least means that failure to comply with a formality cannot formally terminate the right, or prevent it from arising in the first place. It is unclear how much further "enjoyment and exercise" goes than that; I will return to this problem later.

Assuming for the moment that Article 5(2) allows a range of options short of formal nullification of copyright, we are still faced with a difficult problem: new-style formalities have to create a sufficient incentive for compliance to construct a reliable record of ownership and to reliably signal copyright status, but cannot use the forfeiture of rights to incent compliance.

The simplest solution would be to preserve formally voluntary registration, notice, and recordation of transfers (and reestablish a formally voluntary renewal formality) for all works, including works of foreign authors, but then incent compliance by exposing the works of noncompliant rightsholders to a "default" license that allows use for a predetermined fee. The royalty payable under the default license would be low. Ideally, the royalty to license a work that a rightsholder has failed to register, notice, reregister in the case of a transfer (i.e., record), or renew should be set to approximate the cost of complying with these formalities (i.e., the total cost of informing oneself about the details of compliance and then satisfying them). That way a rightsholder who expects his work to produce revenue exceeding the cost of complying with the relevant formality will prefer to comply with the formality, whereas a rightsholder who expects his work to produce revenue amounting to less than the cost of compliance will prefer to expose his work to the default license. (The rare rightsholder who estimates the likely revenues from his work to be equal to the cost of complying with the formality will be indifferent between compliance and exposure to the license.)

This system of formally voluntary formalities plus default licenses - which I have referred to previously as "new-style" formalities - establishes indirectly what the traditional system of compulsory formalities did directly: it eases access to commercially valueless works for which protection (or the continuation of protection) serves no purpose and focuses the system on those works for which protection is needed to ensure that the rightsholder is able to appropriate the commercial value of the expression. For the filtering function to work, of course, the government would have to maintain an easily accessible and up-to-date public registry. Given current computer database and search technology, this would not be difficult.

[*556] Importantly, the use of default licenses in a system of new-style formalities avoids the general objection to the compulsory licensing of intellectual property goods: the need for a legislature, agency, or court to set a price for the license in the absence of market negotiations. With respect to the particular use of default licenses proposed here, compliance with the formalities - or the failure to comply - serves as a price signal. Failure to comply means that the rightsholder places a minimal value on the right, a value no greater than the cost of compliance. That is all we need to know about works for which rightsholders fail to comply with formalities. And by exposing these works to a default license, we are giving these rightsholders nothing less than what they themselves expect in term of returns. The system of default licenses is therefore efficient: it removes transaction costs that would otherwise frequently prevent use, while charging an approximately optimal price (i.e., near zero) for a license. With respect to works for which rightsholders comply with formalities, the market continues to set the prices of licenses.

Default licenses can be analogized to the "penalty defaults" of the type proposed by Ian Ayres and Robert Gertner as gap-filling rules for incomplete contracts. n243 Default rules in contract theory are intended to fill gaps in contracts by providing the parties with what they likely would have contracted for. "Penalty defaults" are gap-filling rules that are designed to give at least one party to the contract an incentive to contract around the default and therefore to choose affirmatively a preferred contract provision. Penalty defaults are purposefully designed to impose what the parties would not want, in order to encourage the parties to negotiate the solution that they do want. Importantly, penalty defaults incent contracting parties to reveal information to one another that might not be revealed otherwise.

Although Ayres and Gertner formulate and apply their theory of penalty defaults in the contract law context, the theory can be applied in the intellectual property context as well. The default licenses that back new-style formalities are a kind of penalty default rule, in that they are precisely the outcome that the owner of a valuable copyright would not desire. The existence of the license encourages owners of certain works to produce information that might not be produced otherwise - i.e., that their works are sufficiently valuable that continued copyright protection makes sense.

This system of voluntary formalities backed with default licenses raises an immediate question: do they comply with the Berne Convention (and, thereby, with TRIPs)? Although there are arguments both ways, I believe that the better reading of Berne would permit new-style formalities.

Article 5(2) and economic rights. The first issue is whether new-style formalities offend the Berne Article 5(2) proscription of formalities that [*557] interfere with the "enjoyment and exercise" of copyright. There are two senses in which that phrase may be interpreted. The first relates to enjoyment and exercise of the author's economic rights. Under Article 9(1) of the Berne Convention, authors of literary and artistic works have the exclusive right of authorizing the reproduction of those works "in any manner or form." This includes traditional photocopying, digital copying, or any other form of copying of the entire work or any part thereof. These exclusive rights include, as a necessary corollary, the right to refuse to authorize reproduction of a protected work. This right to exclude is the mainspring of the author's economic right: by restricting reproduction, the author may reduce output of his work and thereby realize supracompetitive returns if his work lacks ready substitutes.

Nevertheless, authors who fail to comply with new-style formalities and thereby lose their previously existing right to exclude are likely not, as a category, deprived of any aspect of the "enjoyment and exercise" of the economic rights appertaining to their copyright. An author who fails to comply with new-style formalities is merely converting an entitlement that is initially protected by a property right (the right to exclude, realized through injunctions and infringement damages) into an entitlement protected by a liability right (the right to recover revenues from use via a default license). n244 Even though new-style formalities set up the liability rule as the default, and require authors to opt out to preserve their ability to exploit the property rule, the system - unlike the usual system of compulsory licenses - is still voluntary. And if the royalty payable under the license is set correctly, owners of copyrights with projected values lower than the cost of complying with a formality should actually prefer the liability right as a means of exploiting their copyrights.

Creating an exploitation option based on a liability rule is a modest extension of what the current regime provides. In the copyright system we have now, rightsholders are allowed to choose the best approaches within the existing set of property rules for exploiting their copyrights, whether via exclusion and collection of infringement damages, a program of licensing and collection of license royalties, or a mixture of the two. New-style formalities would expand the existing process by establishing a liability rule option in the form of a default license for works not valuable enough to justify customized licensing. Again, whether the default license applies is within the control of the rightsholder, and therefore a rightsholder's decision to rely on that liability rule, rather than on a property rule, to protect the enjoyment and exercise of his copyright is not a forfeiture of rights. It is, rather, a signal that a particular rightsholder believes that a one-size-fits-all liability rule based on a default license is preferable, because of low transaction costs, to a property right [*558] exploited through (expensive) customized licensing or enforced through (expensive and protracted) infringement litigation.

The function of new-style formalities, in sum, would be to establish decision points at which rightsholders would be obliged to choose whether to stay within the typical system of property rules or to switch to a liability rule as the means to exercise and enjoy their copyright. Thus, new-style formalities do not touch on the existence or continuation of copyright, but merely on the manner in which rights are exploited. The difference is important, as the WIPO commentary to Article 5(2) makes clear: "what is at issue here is the recognition and scope of protection and not the various possible ways of exploiting the rights given by the law." n245

Of course, the "signal" that authors send via compliance or noncompliance with new-style formalities is unlikely to accurately reflect the underlying value of a work in every instance. Some authors will doubtless underestimate the future revenues that their works may bring in, and will mistakenly opt not to comply with formalities and expose their work to a default license. It is important to note that the possibility of error cuts both ways: some authors will mistakenly opt to invest in compliance with formalities for works that are unlikely to produce revenues greater than the cost of compliance. But despite the certainty that some authors will make the wrong decision, it is nonetheless true that authors (and assigns such as publishers) are the parties best placed to decide whether the likely returns from a particular work are great enough to merit investment in compliance with formalities, so the signal-to-noise ratio is likely to be usefully high. In any event, if an author is uncertain regarding his work's future value, he retains the option of making the relatively small investment required to comply with formalities as a form of insurance against incorrectly valuing his asset.

There is a rejoinder to these arguments that proceeds from a different view of what exactly Berne protects the "enjoyment and exercise" of. What Berne protects, this objection would argue, is not the author's overall ability to exploit his copyright, but something more specific: the enjoyment and exercise of an author's right to exclude, a right that is granted in Article 9(1) of the Convention as well as in domestic law. The right to exclude, according to this objection, is not merely an initial entitlement, but is the subject of Berne's protection for the entire term of copyright. Accordingly, even if rightsholders are free to alienate their right to exclude (via, for example, a voluntary license or a dedication to the public domain), Berne prohibits governments from conditioning the right to exclude on compliance with a formality. Berne prohibits, in other words, using noncompliance with a formality as a trigger to shift a work from property rule to liability rule status.

[*559] This is a colorable argument, but is, I believe, an overreading of Berne. The right to control reproduction - to exclude others from making copies - is, as Article 9(1) of Berne makes clear, the core right granted under copyright. So copyright is structured to protect an author's interests, using a property rule as the initial entitlement. But there is nothing inevitable about this choice; it is, rather, a practical one. Authors' interest in their works could have been protected by using a liability rule as the initial entitlement, except that structuring the entitlement in that way would require government to set a price for the use of copyrighted works. In general, we do not believe that government is well placed to set prices. In the absence of any indication that government would be able to set a more accurate price for the right to reproduce a copyrighted asset than would be set by a market transaction (and to do it more cheaply as well), it makes sense to base copyright in a property rule, at least with respect to the initial entitlement. n246 It also makes sense to limit government's ability to mandate access to copyrighted works to the extent that we fear that the nonmarket pricing decisions that such mandated access would make necessary are unlikely to accurately track the market value of the asset. Article 9(2) of Berne limits such government-mandated "exceptions" from the author's exclusive right; I will return to that provision shortly. But the important point is that the purpose of the copyright system is not to protect a rightsholder's property right qua property right. The purpose of the copyright system is to protect a rightsholder's ability to use his initial entitlement, which comes in the form of a property right, as a lever to pursue the exploitation strategy best suited to his particular interests.

Seen in that light, it makes little sense to lump default licenses, in which use is priced with the assistance of robust information produced by the author himself, together with typical compulsory licenses, in which price is determined by fiat (or, at best, fiat following administrative hearings at which rightsholders and would-be compulsory licensees offer contending and self-serving accounts of the worth of the assets at issue). Unlike in the case of ordinary compulsory licenses, the default licenses attending new-style formalities do not threaten to interfere with the exclusive rights of any rightsholder who does not consider the use of a default license to be in his interest. The objection to new-style formalities, then, is reduced to a narrow protest that government should not force authors to decide whether to exploit their right via a property rule or a liability rule, or, alternatively, that if government does force the choice, the default should be a property rule rather than a liability rule (i.e., rightsholders should have to affirmatively "opt in" to a liability rule). These arguments would make the right to exclude truly totemic. Berne (and domestic copyright law) would no longer be focused on protecting authors' abilities to exploit their works, but would be bound up instead in [*560] enforcing a particular property rule, not just as an initial entitlement, but as a perpetual entitlement, even in instances where authors (and the public) would benefit from the use of an alternative means of exploitation.

Article 9(2), TRIPs Article 13, and copyright "exceptions." So the best reading of Article 5(2), in my view, would be one which allows new-style formalities. But assume for the moment, contrary to the arguments laid out above, that the right to exclude is in fact a totemic right under Berne. Assume then that the Article 5(2) prohibition of formalities that interfere with the "enjoyment and exercise" of copyright focuses narrowly on maintaining inviolate, throughout the term of copyright, the right to exclude, i.e., the property rule. Assume also that new-style formalities impermissibly interfere with the enjoyment and exercise of that right. The default licenses that enforce new-style formalities may nonetheless still be permissible under Article 9(2), which permits exceptions to the exclusive reproduction right in certain "special cases," provided that the excepted reproduction "does not conflict with a normal exploitation of the work and does not unreasonably prejudice the legitimate interests of the author." Article 13 of the TRIPs accord contains similar language and generalizes Berne's exceptions to all of the exclusive rights granted under Berne and TRIPs (e.g., the rights to create derivative works, to authorize public performances, and to authorize broadcasts). n247

Ricketson states that exceptions permitted under Article 9(2) are subject in all cases to respect for the author's moral rights; n248 that much is evident from the three-step test in Article 9(2), which mixes economic and moral rights concerns. It is also clear that the three criteria for restricting exclusive rights must all be met in order for restrictions to be permissible. Further guidance in how to apply the Article 9(2) test is available from the Report on the Berne Revision Conference held in Stockholm in 1967:

If it is considered that reproduction conflicts with the normal exploitation of the work, reproduction is not permitted at all. If it is considered that reproduction does not conflict with the normal exploitation of the work, the next step would be to consider whether it does not unreasonably prejudice the legitimate interests of the author. Only if such is not the case would it be possible in certain special cases to introduce a compulsory license, or to provide for use without payment. A practical example might be photocopying for various purposes. If it consists of producing a very large number of copies, it may not be permitted, as it conflicts with a normal exploitation of the work. [*561] If it implies a rather large number of copies for use in industrial undertakings, it may not unreasonably prejudice the legitimate interests of the author, provided that, according to national legislation, an equitable remuneration is paid. If a small number of copies is made, photocopying may be permitted without payment, particularly for individual or scientific use. n249

There is another more recent, and perhaps more useful, interpretation of the exceptions language. In June 2000, a dispute resolution panel of the World Trade Organization (WTO) issued a report, n250 in an action brought by the European Communities, holding that Section 110(5) of the Copyright Act, n251 a provision establishing royalty-free compulsory licenses for the public performance of radio or television transmissions of nondramatic musical works for businesses, including restaurants and bars, below a certain size or using certain "homestyle" stereo and television equipment, did not qualify as a permitted exception under Article 13 of the TRIPs accord. n252 The WTO panel stated that Berne Article 9(2) and TRIPs Article 13 were to be construed in a manner that avoided conflict, n253 and it limned the scope of each of the three elements of the Berne Article 9(2)/TRIPs Article 13 test.

"Normal exploitation." Whether a system of default licenses conflicts with the "normal exploitation" of a work can be thought of in two different ways. One would reflect a totemic view of the right to exclude: that right, this argument would hold, is the "normal" way in which works are exploited, and therefore default licenses, which substitute a liability rule for the exclusionary property rule, are incompatible with "normal exploitation." But by that reasoning, all compulsory licenses would run afoul of the "normal exploitation" element of the Article 9(2) test, a result which makes no sense against the background of a Berne provision intended to regulate, but not to prohibit, the use of exceptions such as compulsory licenses.

The second, and better, construction of the "normal exploitation" element would ask whether a default license would result, in comparison with normal copyright remedies, in rightsholders in the aggregate realizing lower returns from their works - or, as the WTO panel framed the question, whether the excepted use would "enter into economic competition with the ways that right holders normally extract economic value from ... the work ... and thereby [*562] deprive them of significant or tangible commercial gains." n254 As outlined above, default licenses would apply only when a rightsholder failed to comply with a low-cost formality. Failure to comply is a signal that the net present value of expected future revenues from a work is lower than the cost of compliance. The fee payable under the default license (i.e., the fee payable by each user) is set to approximate the cost of compliance. Thus, for works that fall under the default license, on average the rightsholder's ability to "exploit" the work will be, if anything, enhanced.

The argument that new-style formalities will not impair rightsholders' abilities to exploit their works finds powerful support in the fair use doctrine, which often has been characterized as a limitation on exclusive rights that immunizes uses for which the transaction costs of negotiating a license exceed the potential return to rightsholders from the license. n255 Articles 10 and 10bis of the Berne Convention list certain fair use-type exceptions to the exclusive reproduction right; these include limited rights to make quotations n256 and to use works in aid of teaching n257 and news reporting. n258 Berne's provision for fair use shows that the Convention permits incursions on the exclusive reproduction right when transaction costs make negotiated arrangements too costly. The default licenses that back new-style formalities can be supported on the same grounds - in fact, they are in some respects less invasive than fair use, because use under a default license results in compensation, whereas use under the fair use doctrine does not.

"Legitimate interests" and moral rights. The second requirement - that an exempted reproduction "not unreasonably prejudice the legitimate interests of the author" - is susceptible, at least in part, to the same analysis. To the extent that the author's "legitimate interests" are taken to mean his ability to capture whatever rents his exclusive rights will return, the default license does not interfere. n259 To the extent, however, that the phrase "legitimate interests" refers to the author's moral rights, additional analysis is required.

[*563] The "legitimate interests" of copyright holders (as well as the "enjoyment and exercise" of copyright) is tied, as well, to the moral rights that Berne requires signatories to grant to authors. n260 Article 6bis of the Berne Convention requires member states to grant authors rights of paternity (i.e., the "right to claim authorship") and integrity (i.e., the "right to object to any distortion, mutilation or other modification" that would prejudice the author's reputation). n261 These rights are independent of the author's economic rights, survive the transfer of those economic rights, n262 and must, in most instances, persist for at least the expiry of the economic rights, even following the death of the author. n263

It must be noted that since acceding to the Berne Convention, the United States has refused to fully incorporate into its domestic law the moral rights set out in Article 6bis. Instead, the United States has relied on a number of different sources, including an author's right under copyright law to control derivative works; state unfair competition, defamation, and privacy laws; and the Visual Artists Rights Act of 1990 (VARA), n264 an amendment to the copyright law granting limited rights of paternity and integrity to a narrowly defined class of "works of visual art," n265 to approximate the Berne requirements. Whether the United States currently complies with Berne Article 6bis is a subject beyond the scope of this Article. n266 But it seems likely that current U.S. law, if it complies at all, does so only minimally. The question is whether subjecting certain works to the default licenses that attend new-style formalities would subtract meaningfully from a level of protection for paternity and integrity rights that is already stinting. If so, then the United States may fall out of compliance with Article 6bis (or, perhaps, make its continued noncompliance no longer tolerable).

[*564] Regardless of whether current U.S. protection of paternity and integrity rights meets Berne minima, installing a system of new-style formalities need not touch these rights at all: although the works of noncompliant rightsholders would be subject to default licensing, the copyrights on works affected by the regime would nonetheless be left formally intact. Accordingly, there is no reason why the law could not specify that use of works under the default licenses is subject, in all cases, to whatever (narrow) protection current U.S. law affords to paternity and integrity rights.

But new-style formalities could go further, in a way that would strengthen the United States's commitment to facilitating the exercise of Berne-mandated moral rights. One method would be to infuse into new-style formalities the type of "some rights reserved" copyright customization that Creative Commons provides now. n267 New-style registration, notice, recordation of transfers, and renewal could be designed to allow rightsholders to signal exactly which rights they wish to retain, and which freedoms are allowed. The difference is that instead of relying on a Creative Commons license, rightsholders' choices about which rights to reserve would be enforceable as a matter of positive law.

There are two potential advantages of the integration of new-style formalities with the Creative Commons approach. First, it could be used to strengthen U.S. compliance with Berne-mandated moral rights. Authors who comply with new-style formalities could be permitted, in exchange for a blanket grant of permission to use their works, to demand attribution in all cases, even in instances, such as fair use, where the rightsholder would not currently have the power to enforce such a demand. Similarly, authors who comply with new-style formalities would be able to protect their integrity rights, by permitting reproduction but restricting derivative uses. n268

The second advantage is normative: by disaggregating economic from moral rights, and the moral rights of paternity and integrity from one another, new-style registration, notice, recordation, and renewal would allow us to understand over time what people want in terms of rights for different types of works, and how those desires change (if they do at all). That information would be useful in the debate over future changes to the copyright laws.

[*565]

Figure 6: License Distribution Choices of Creative Commons License Users

[SEE FIGURE 6 IN ORIGINAL]

The data set out in Figure 6 - data which was provided by Creative Commons - shows the choices that Creative Commons licensors have made over the first two years of the organization's existence regarding which rights to reserve and which to give away. Although rightsholders who seek out, or become informed about, Creative Commons and decide to enter into a Creative Commons license are certainly not representative of rightsholders as a broader group, the Creative Commons license distribution data gives us some insight into what the world might look like when copyright is no longer an on/off switch, but is more finely variegated.

Perhaps most unexpectedly, the data shows that a significant majority (67%) of Creative Commons licensors allow the use of their content in the creation of derivative works. This data suggests that many rightsholders would voluntarily abandon control over derivative works, which is an element both of the author's economic right and his right of integrity - although most who do so (again 67%) would limit use to noncommercial derivative works.

In contrast, the Creative Commons data shows that almost all licensors (97%) require attribution in exchange for permission to use their works. That number suggests that the norm favoring attribution is strong, and, consequently, that we may succeed in moving copyright closer to rightsholders' expectations and simultaneously free a large amount of creative work if we install a [*566] mechanism for exchanging the right to control reproduction for a stronger commitment to provide attribution.

"Certain special cases." The final element of Article 9(2), the "certain special cases" requirement, is exceedingly difficult to pin down. The first area of difficulty is determining whether the element imposes a separate constraint at all. The official WIPO commentary on Article 9(2) does not lay out any guidelines for determining whether an exception fits under the "certain special cases" language. Indeed, the commentary fails to mention "certain special cases" as a separate requirement, and instead discusses only the "two conditions" of protecting normal exploitation and avoiding unreasonable prejudice to legitimate expectations - suggesting, perhaps, that the "certain special cases" language does no more than reflect a situation in which the two principal factors are met. n269 The same sense is conveyed in the extract, quoted above, from the Report of the Stockholm Revision of the Berne Convention, which focuses on the "normal exploitation" and "legitimate interests" elements, and suggests again that the "certain special cases" language refers entirely to instances in which neither element is impinged upon.

The WTO panel took a different approach. Interpreting each of Article 13's three requirements to avoid any "redundancy or inutility," n270 the panel held that the "certain special cases" language imposed a separate constraint on exceptions and, in fact, that exceptions must first be shown to meet its requirements before analysis of the other criteria is undertaken. n271 The panel read the "certain special cases" language to require that exceptions be "clearly defined" n272 and "narrow in [a] quantitative as well as a qualitative sense." n273 Applying these standards, the panel invalidated Section 110(5)(B)'s broad grant of compulsory licenses for nondramatic musical works for businesses, holding that the exception could not qualify as "narrow" when, according to evidence before the panel, 70% of all restaurants, 73% of all bars, and 45% of all retail stores qualified for compulsory licenses under the provision. n274 In contrast, the panel upheld the provisions of Section 110(5)(A) granting compulsory licenses for dramatic musical works where the standards set out in the statute would allow only 16% of restaurants, 13.5% of bars, and 18% of retail stores to qualify for royalty-free compulsory licenses for a narrower class of copyrighted works. n275

[*567] For reasons that have been discussed earlier in this Article, we would expect a large number of authors to fail to comply with registration and notice requirements, and, similarly, the majority of rightsholders to fail to comply with a renewal formality. Accordingly, a large number of works (in both absolute and percentage terms) will be exposed to default licenses in a system of new-style formalities. So one might read the WTO panel's holding that exceptions must be "narrow in a quantitative as well as a qualitative sense" as ruling default licenses out as an exception permissible under Berne 9(2) or TRIPs 13.

It should be noted, however, that the WTO panel's report was issued in a case involving royalty-free compulsory licensing, which means that, unlike in the case of default licenses, the WTO panel was dealing with an exception that eliminates rightsholders' ability to receive compensation for their work for the excepted uses. In that context, and in the absence of any particular provision in Berne or TRIPs approving the particular compulsory licenses at issue, an independent "special cases" element makes sense. If the touchstone of the exceptions provision is preserving rightsholders' abilities to profit from their works, application of the "normal exploitation" and "legitimate interests" elements might theoretically provide all the information needed to determine whether an exception should be permitted. But, in practice, whether a particular exception interferes with "normal exploitation" or a rightsholder's "legitimate interests" may be difficult to determine with certainty. The "certain special cases" element can therefore be seen as providing an easily administered threshold test: does this exception affect a large enough share of the potential licensing market that it is likely to interfere with a rightsholder's ability to exploit his work?

But what role should the "certain special cases" element play in a case like default licenses, where there is no systematic interference with rightsholders' economic interests? In such an instance, the "certain special cases" language may be both theoretically and practically coterminous with the "normal exploitation" element, i.e., the number of works to which an exception applies is required to be sufficiently limited that the market for a particular work is not substantially affected. For works that have a ready market, even a relatively small amount of excepted usage might affect the rightsholder's ability to exploit the work. But because the default licenses connected to new-style formalities apply only after a rightsholder signals that his work does not enjoy a substantial commercial market, even a relatively large number of excepted uses made under a default license would not interfere with the rightsholder's own understanding of his ability to exploit the work. Put differently, there is a strong argument that because default licenses arise only when the author or rightsholder sends a signal that his work lacks significant commercial value, the default license applies only in the "special case" where a liability rule is preferred by the rightsholder.

[*568] Perhaps the most that can be said at this point about the "special cases" element - and indeed about the Berne Article 9(2)/TRIPs Article 13 test altogether - is (1) that on a proper reading of Berne Article 5(2), the need to justify new-style formalities as an "exception" will never arise, and (2) if the exceptions provision is applicable, the elements of the test are sufficiently indeterminate (at least as they apply to default licenses, a mechanism that the Berne drafters could not have had in mind when they formulated the Article 9(2) test) that the application of the test will involve not so much testing new-style formalities under the formal elements as an evaluation of whether the system of new-style formalities serves (or undermines) the foundational principles of the Berne Convention. If Berne is focused on protecting authors' ability to exploit their works, new-style formalities will be compatible with the Convention. If Berne is focused, instead, on enforcing a particular property rule, regardless of authors' interests, then new-style formalities may fail under the current text of the Convention.

Conclusion

My hope in writing this Article is that reformalization might reduce the current friction over copyright to a level conducive to a broader discussion about intellectual property reform. As the discussion above makes clear, the stifling of creativity - as well as free speech - created by the current unconditional copyright regime will only become worse with the passage of time. Unlike more radical proposals for reforming copyright law, reformalization is a task that can be accomplished with a few manageable and realistic changes to international and domestic law and without endangering the interests of any particular segment of the copyright community. Just as importantly, unlike more limited proposals, reformalization would ensure that creative material lacking commercial value becomes available for reuse immediately.

With clear rules governing copyright protection and a formalities-compliance system making use of the best technology available, a reformalized copyright regime would reinstate the best aspects of the old U.S. copyright system while leaving behind the bureaucratic difficulties that turned "formalities" into a despised term decades ago. Though this Article does not offer a political road map for achieving the legislative changes needed for reformalization, the proposal outlined above shows that significant reform can take place without damaging the interests of copyright owners who would otherwise have strong incentives to oppose the creation of a less restrictive copyright regime. The challenge now is to turn these academic concepts into actual changes in domestic and international laws.

****