ARTICLE

Copyright Preemption After the PROCD Case: A Market-Based Approach

MAUREEN A. O'ROURKE

table of contents

I. INTRODUCTION

II. THE PROCD CASE-OF SHRINKWRAP LICENSES AND PRIVATE COPYRIGHT LAW

A. The Enforceability of the Shrinkwrap License Under Contract Law

B. The Enforceability of the Use Restriction Under Federal Copyright Law

III. A DEEPER ANALYSIS OF THE PREEMPTION ISSUE

A. Two Competing Views of Copyright-Freedom of Contract v. Public Domain

B. An Evaluation of the Competing Views

C. Alternative Approaches

IV. CONCLUSION

I. Introduction

As information becomes increasingly available in digital form, a little noted yet significant legal change is occurring in the way in which information providers use the law to establish the terms under which they market their products. Electronic information providers, in contrast to their hard-copy counterparts, have continually turned to the private law of contract both to supplement and modify the public law of copyright.1 While this trend began when most users were relatively large commercial, academic, or governmental enterprises, it accelerated as software providers began to market pre-packaged software to consumers, using the infamous "shrinkwrap" as a device to alter the copyright law through private contract.2 This movement to contract has continued with the proliferation of on-line data. In particular, it has extended to contracting not just to modify copyright rights that would otherwise apply but also to create private copyright protection through contract in instances in which the public law would deny copyright protection altogether.

More specifically, since the Supreme Court's decision in Feist Publications, Inc. v. Rural Telephone Service Co.,3 compilers of facts have faced considerable risk that their compilations will be considered outside the scope of copyright protection because they lack originality.4 In the absence of certainty that public copyright will protect such data-and thereby allow recoupment of the investment made in gathering and collating it-many providers of electronic databases have attempted to create their own copyright protection through both standard form (i.e., shrinkwrap) and negotiated contracts.5 By endowing noncopyrightable data with copyright-type protection, these contractual terms governed by state law seem to be preempted since they may function to negate the federal copyright policy of promoting the free flow and use of factual information. To date, academic commentary has largely treated the contract formation and copyright preemption issues as discrete inquiries, focusing separately on the issues of the enforceability of shrinkwrap licenses6 and whether providers of copyrightable information may use contract law to supplement or modify the rights which copyright grants.7 It has had even less to say about the relationship between contract formation and copyright preemption where private parties are attempting to create their own copyright law by contract to protect information which Congress does not protect under the Copyright Act ("Act").8

This absence of academic commentary has become increasingly glaring as courts have begun to address these questions. In particular, ProCD, Inc. v. Zeidenberg9 demonstrates not only the depth of the division of opinion on these issues, but also the compelling policy arguments on each side. In ProCD, the district court held that the shrinkwrap is not a valid contract and, even if it were, its provisions which effectively confer quasi-copyright rights on noncopyrightable data are preempted by the Copyright Act.10 The Seventh Circuit reversed on both counts, holding that a shrinkwrap license may be enforceable and parties may, in certain circumstances, contract to treat noncopyrightable data as if it were copyrighted without running afoul of the Act.11

Because this case is the first of its kind and yet addresses the everyday practices of the database industry, it merits close attention. This article discusses the case in detail, analyzing the contentions of both courts. It then attempts to explain how the different decisions of the two courts reflect different understandings of the nature of copyright law. It argues that the Seventh Circuit's decision reflects a flexible freedom of contract approach to preemption which is fully appropriate given the relative novelty of the technology and congressional silence regarding preemption with respect to such technology. The article then discusses in more detail two approaches to preemption-the Easterbrook model which the Seventh Circuit used in the ProCD case and the default rules model which draws on the academic literature regarding default rules. It argues that a default rules approach may be preferable to the Easterbrook model but would require some legislative action to implement. In the interim, the Easterbrook model forms a reasonable basis from which to assess claims that provisions restricting copyright rights should be preempted.

II. The ProCD Case-Of Shrinkwrap Licenses and Private Copyright Law

In the ProCD case, ProCD marketed a CD-ROM called Select Phone, which contained both a version of its database containing listings and other information from over 3,000 telephone directories, along with search and retrieval software that facilitated the manipulation of that data.12 ProCD marketed its product to both commercial and consumer13 users, charging the general public a much lower price than it charged the trade.14 Defendant Zeidenberg purchased the consumer version of the product in 1994 and later purchased two additional packages containing updated versions of the database.15 The consumer version contained a shrinkwrap license agreement16 that was referenced on the outside of the product packaging, encoded on the CD-ROM, reproduced in hard copy in the manual accompanying the product and that appeared on the screen whenever the user ran the software.17 The license agreement contained a term limiting the use of both the copyrightable software and noncopyrightable listings to non-commercial purposes.18 Zeidenberg violated the use restriction by making the noncopyrightable listings available from his corporation (Silken Mountain Web Services, Inc.), for a fee, via the Internet.19 ProCD sued Zeidenberg and Silken Mountain Web Services in Wisconsin federal district court, setting forth a number of causes of action, including breach of contract.20 The court (and subsequently, the court of appeals as well) analyzed the breach of contract claim as involving two interrelated inquiries: (i) whether the shrinkwrap license was enforceable as a matter of contract law; and (ii) if so, whether it was rendered unenforceable as preempted by the Copyright Act.

A. The Enforceability of the Shrinkwrap License Under Contract Law

Conceptually, the issue of the enforceability of the shrinkwrap license is divided into two questions: (i) whether, as a matter of contract formation, a valid contract may be formed in the manner employed by ProCD; and (ii) assuming such a contract may be formed, whether and what terms of that contract should be enforced. However, the courts did not bifurcate the issue in this manner. Instead, both courts presented one inquiry into enforceability, arriving at opposite conclusions. An examination of the judicial reasoning of both courts suggests that the ultimate holding of the court of appeals is that contracts may be formed in the manner employed by ProCD but both the formation process and particular terms will remain subject to judicial scrutiny. An analysis of both courts' reasoning demonstrates that this holding is both consistent with the relevant law and desirable from a policy perspective.

1. The differing views of the district court and court of appeals

Both the district court and court of appeals analyzed the enforceability of the shrinkwrap under common-law contract and Wisconsin's implementation of Article 2 ("Sales") of the Uniform Commercial Code ("UCC" or "Code"). According to the district court, Zeidenberg accepted the offer-and therefore formed a contract with ProCD-at the time of sale, rather than when he began to use the product after having the opportunity to become aware of the specific contractual terms desired by ProCD.21 The district court contended that UCC provisions stating that acceptance does not occur until the buyer has a reasonable opportunity to inspect the product did not apply to the case at hand, since inspection is intended to uncover product defects, not additional written contractual terms that the seller desires.22

The district court, drawing on two of the three prior decisions relating to shrinkwraps,23 analyzed the facts under sections 2-207 and 2-209 of the UCC. Section 2-207 is the "battle of the forms" provision. It essentially modifies the common-law mirror-image rule by recognizing the commercial reality that often the competing forms of the buyer and seller contain conflicting or additional terms yet the parties still intend to contract.24 Section 2-207 contains detailed instructions on how to treat additional terms appearing in the competing forms. In contracts between merchants,25 the additional terms become part of the contract unless they materially alter it, the offer expressly limits acceptance to its terms, or notification of objection to the additional terms is given in a timely manner.26 However, the Code does not clearly address what to do with additional terms when one party is a non-merchant like Zeidenberg. The literal text of the Code seems to indicate that a contract may be formed with the additional terms to be separately accepted or rejected.27 According to the court:

Keeping in mind the legislative goal behind 2-207, it is improbable to think that the drafters wanted consumers to be held to additional proposed terms in situations in which merchants were given protection . . . [A]pplying 2-207 to the consumer transaction in this case . . . leads to the conclusion that the user agreement was not binding on defendants because they never agreed to it expressly and it never became part of the agreement between the parties.28

Hence, one can infer that the court considered the contractual use restriction a material alteration to the contract formed at the point of sale.29

The court also considered whether the shrinkwrap might be viewed as an enforceable modification to the contract formed at the point of sale. Under 2-209, while modifications to a contract do not require consideration to be binding, the parties still must agree to the modification.30 The court stated:

Section 2-209 requires the express assent of a party to any proposed contractual modifications. Assent cannot be inferred from a party's conduct in continuing with an agreement. In this case, defendants did not assent expressly to the terms of the user agreement. Their continued use of the Select Phoneä product has no bearing on whether they accepted the user agreement. Under these circumstances, 2-209 does not warrant the incorporation of the user agreement into the parties' initial sales agreement.31

The district court stated in its holding:

I conclude that because defendants did not have the opportunity to bargain or to object to the proposed user agreement or even review it before purchase and they did not assent to the terms explicitly after they learned of them, they are not bound by the user agreement.32

On appeal, the Seventh Circuit, in an opinion authored by Judge Easterbrook, informed its statutory analysis with reference to economic considerations and practical realities. From a statutory perspective, the court of appeals simply disagreed with the district court's reliance on 2-207 and its reading of the UCC as not permitting "money now, terms later" contracts.33 The court stated that 2-207 does not apply in cases involving only one form and went on to note that 2-204(1) of the Code permits contracts to be formed in any manner including conduct.34 It stated that ProCD as "master of the offer" was free both to "invite acceptance by conduct, and [to] propose limitations on the kind of conduct that constitutes acceptance."35 In the case at issue, ProCD offered a contract to be accepted by the buyer's use of the software after the buyer had an opportunity to review the offered contractual terms.36 Zeidenberg's continued use of the software after becoming aware of the terms constituted his acceptance.37

Unlike the district court, the court of appeals drew support for its holding by analogy to 2-606, which defines "acceptance of goods."38 Under 2-606(1)(b), a buyer may be deemed to have accepted where it fails to make an effective rejection, but such acceptance does not occur until the buyer has had a "reasonable opportunity to inspect [the goods]."39 While 2-606 deals with acceptance of goods and does not expressly mention acceptance of an offer, the court saw 2-606 as further evidence of the UCC's flexibility in allowing the parties to arrange their affairs in such a manner "that the buyer has a chance to make a final decision after a detailed review."40

Unlike the district court, the court of appeals discussed at length the policy implications of holding boilerplate contracts unenforceable. Its focus thus seemed to be more systemic than that of the district court, as it was concerned with the impact of its decision not merely on the immediate parties, but on all similarly situated parties.41 For example, it stressed that transaction costs may be saved through the use of standard forms42 and that, in "real life," buyers often pay for a good or service before the detailed terms are even communicated to them.43 In the consumer context, it emphasized the everyday transaction of purchasing boxed goods.44 Warranty disclaimers that states routinely enforce are often inside such boxes.45 Likewise, when a consumer purchases drugs over the counter, the consumer is charged with reading the accompanying package insert.46

The court of appeals also considered the nature of the software industry, acknowledging that the shrinkwrap is routinely used. The court suggested the conclusion that while academia has doubts about its enforceability, business does not.47 Moreover, the court emphasized the mischief that the district court's decision could work in stifling the burgeoning software distribution mechanisms of mail-order and on-line sales.48 Particularly in on-line transactions, the customer never sees a box but instead receives an electronic file which contains the relevant terms and conditions.49 As the court stated in an intuitive economic analysis:

On Zeidenberg's arguments, these unboxed sales are unfettered by terms-so the seller has made a broad warranty and must pay consequential damages for any shortfalls in performance, two "promises" that if taken seriously would drive prices through the ceiling or return transactions to the horse-and-buggy age.50

The court of appeals also argued for contract enforceability from a more sophisticated economic approach. As it noted, ProCD engaged in price discrimination, charging consumer buyers less than commercial users.51 If ProCD could not engage in such price discrimination-i.e., it had to charge all users the same amount-it would eventually have to raise the price to both consumers and commercial users.52 Fewer consumers would be able to purchase the product, while commercial users might have to pay more to make up the revenue shortfall caused by lost consumer sales.53

For price discrimination to succeed, ProCD had to find a way to prevent consumers from defeating its strategy by reselling the product to commercial users for less than what ProCD would charge such users. In the words of the court of appeals, "[t]o make price discrimination work . . . the seller must be able to control arbitrage."54 This control is particularly difficult for software vendors as they cannot easily distinguish between consumer and commercial purchasers in a garden-variety retail transaction in which a customer walks into the store, picks the software up from the shelf, and purchases it.55 ProCD could have chosen to offer different products in an effort to force users to sort themselves into consumer and commercial segments. Instead, they marketed one product, using the license to limit use of the program to non-commercial purposes.56 To hold the license unenforceable would essentially force ProCD either to raise its price or to change the way it packaged its product.

Based on these economic considerations, the court of appeals held the shrinkwrap to be an enforceable contract. However, it stopped short of a blanket statement that all terms of all shrinkwraps would be enforceable, stating that "[s]hrinkwrap licenses are enforceable unless their terms are objectionable on grounds applicable to contracts in general (for example, if they violate a rule of positive law, or if they are unconscionable)."57

2. An evaluation of the competing views

The court of appeals' decision upholding the enforceability of shrinkwrap licenses is likely to be considered both remarkable and disturbing by academia. The weight of academic commentary prior to ProCD argued primarily for a refusal to enforce such agreements or at least parts thereof for diverse reasons.58 Judicial authority in this area has been quite scant and has yielded no clear cut answer. 59 However, Professor Lemley characterizes the precedent that does exist as reflecting a "general refusal of . . . United States courts to enforce shrinkwrap licenses" which brings U.S. law into harmony with that of the many other countries that do not enforce shrinkwrap licenses or do so only with restrictions.60 The court of appeals' decision then seems truly revolutionary, flying in the face of academic thought and, to a lesser extent, the views of other courts.

However, as a matter of statutory interpretation, the court's decision is eminently supportable. While it is tempting to avoid the federal copyright preemption issue by holding either the shrinkwrap itself (or certain of its terms) unenforceable, there is nothing in the UCC which precludes its enforcement. More importantly, both the UCC's policy and provisions provide ample support to argue both for allowing contracts to be formed in the manner used by ProCD and for enforcing the provisions of such contracts.

Any UCC analysis of both contract formation and enforcement should be anchored in a firm understanding of the Code's overall statutory goals, as well as those contained specifically in Article 2. The Code's overall philosophy is reflected in 1-102(2) which states:

Underlying purposes and policies of this Act are

(a) to simplify, clarify and modernize the law governing commercial transactions;

(b) to permit the continued expansion of commercial practices through custom, usage and agreement of the parties;

(c) to make uniform the law among the various jurisdictions.61

In the context of Article 2, the UCC's drafters sought to decrease contracting costs by adopting legal rules that recognized commercial reality and de-emphasized the formalities of common-law contract.62 Further, Article 2 is based on the belief that law can be and in fact is revealed by what parties actually do.63 Thus, Article 2 often looks to commercial practice in defining the legal rule.64 Moreover, the UCC, in implementing its drafters' belief in freedom of contract, seeks to decrease costs by employing legal rules that give effect to the parties' agreement.65 Thus, the statutory orientation of Article 2 is one in which the parties are given maximum flexibility to draft their own agreement which a court should generally enforce.66

Any particular dispute must be interpreted under the literal wording of the Code placed against this perceived intent. In the case of the shrinkwrap license, such an interpretation argues for its enforceability.67 Simply put, holding the shrinkwrap enforceable based on the economic considerations which Judge Easterbrook identified illustrates the very flexibility in accommodating commercial reality that the drafters designed the Code to achieve.

As a matter of statutory interpretation, the court of appeals has the better argument in asserting that 2-207 is not the governing provision. Section 2-207 governs narrow situations in which the parties have either exchanged conflicting forms or discussed the transaction with one sending a confirmation adding terms which had not previously been considered.68 The shrinkwrap transaction does not fit this paradigm. The very point that some critics argue should render the shrinkwrap unenforceable-its unbargained-for nature69-serves to remove it from the ambit of 2-207 since 2-207 contemplates some type of communication between the parties.

Further, even if one adopts the district court's view that the contract is formed at the time of payment, the conclusion that the shrinkwrap is an unenforceable modification under 2-209 is not compelled. The district court went beyond the statutory wording in asserting that a modification requires the express assent of the party whom it is to bind and cannot be inferred by conduct.70 Section 2-209 requires an "agreement" for an effective modification.71 Under 1-201(3), an agreement is "the bargain of the parties in fact as found in their language or by implication from other circumstances including course of dealing or usage of trade or course of performance."72 According to 1-201's Official Comments, "As used in this Act the word ['agreement'] is intended to include full recognition of usage of trade, course of dealing, course of performance and the surrounding circumstances as effective parts thereof."73

Under the Code's wording then, a modification may be inferred from conduct.74 If one accepts the district court's view that the contract is formed at the time of purchase, then the original terms consist of the rights which the Copyright Act would grant75 plus the UCC's gap-filling provisions.76 The shrinkwrap terms would be considered a modification to which the purchaser may "agree" through his conduct in continuing to use the product after being put on notice of the relevant terms.77

In the ProCD case, usage of trade supports contracting in the manner that the parties employed,78 as well as enforcing the particular terms contained in the shrinkwrap. Simply put, "Everybody does it." More technically, "A usage of trade is any practice or method of dealing having such regularity of observance in a place, vocation or trade as to justify an expectation that it will be observed with respect to the transaction in question."79 It is common practice in the software industry to contract in the manner which ProCD adopted, as well as to employ the terms which it did. This method of contracting facilitates a mass market that might otherwise not exist by saving the transaction costs of face-to-face bargaining. The particular terms help the provider recoup its investment while not forcing it to charge an exorbitant price.

Of course, merely because "everybody does it" does not make it right. The Code's willingness to incorporate trade practices is based on an economic argument. The drafters assumed that commercial parties would not long persist in an inefficient practice.80 The mere fact that a trade practice is time-tested is evidence that it is "efficient" and therefore desirable from an economic perspective.

This contention is less persuasive in cases in which one party is a consumer-buyer who does not bargain with the seller. In such situations, the seller, in the interest of maximizing its own gain from contracting, may foist inefficient terms off on the buyer on a "take it or leave it" basis. For this reason, the Code contains an array of safety valves to help guarantee that buyers are not bound by onerous terms.

For example, the relevant usage of trade binds only those who should be aware of it.81 This, however, may not assist the consumer. The shrinkwrap has been so pervasive for so many years that even a novice buyer should probably be aware of the usage of trade. Still, even if the usage of trade is binding, 2-302 provides for court reformation of unconscionable contracts.82 This section, though, is also unlikely to assist the consumer since, as the court of appeals noted, the use restriction was reasonable under the circumstances.83

The court of appeals' analysis concerning the time of contract formation is equally, if not more, persuasive than the argument that the shrinkwrap constitutes an enforceable contractual modification. The Code, in effectuating its goal of flexibility, encourages the parties to arrange their contractual relations as they see fit. The rules of offer and acceptance as set forth by the Code are not rigid but rather are subject to contrary agreement of the parties. The method employed by ProCD-an offer to be accepted by the user's continued use of the software-is reasonable and, as discussed above, reflects the relevant usage of trade.

Additionally, there is a tenable argument that 2-606(1)(b), which provides for inspection prior to acceptance of goods, offers direct rather than merely tangential support for the Seventh Circuit's holding. As the court noted, "[t]erms of use are no less a part of 'the product' than are the size of the database and the speed with which the software compiles listings."84 The user therefore should have a right to inspect those terms before accepting the goods. Section 2-606(1)(b) affords that right. Nothing in the Code states that acceptance of an offer cannot occur simultaneously with acceptance of the goods.

From a systemic perspective then, the UCC seems to support the enforcement of a shrinkwrap as a valid method of contracting. Again, this conclusion is relatively neutral with respect to the enforcement of particular terms of the contract.85 Neither court was completely clear on this distinction. Upholding the shrinkwrap as a valid method of contracting seems desirable from a policy perspective and consistent with decisions in other contexts upholding contracts which the buyer discovers after purchase. However, merely stating that the shrinkwrap is a valid contract in the abstract says little or nothing about whether particular provisions would be enforceable. For example, a reasonable choice of law provision would probably be enforceable86 while a clause limiting time for complaints might be set aside in the case of a latent defect.87 The relevant questions are: (i) as a general matter, how does the law decide which contractual clauses should be enforced and which should not; and (ii) where on that spectrum of enforceability does the ProCD use limitation fit?

Under UCC 2-302, the courts can use unconscionability as the primary mechanism to police contractual terms. Section 2-302 grants a court broad discretion in deciding whether or not, and how, to enforce a contract that contains an unconscionable clause. Unfortunately, it offers little guidance to courts on how to make the threshold determination of whether a particular clause is unconscionable.88 According to the Official Comments, "The principle is one of oppression and unfair surprise . . . and not of disturbance of allocation of risks because of superior bargaining power."89 The question then often reduces to an assessment of what the buyer's expectations were under all the facts and circumstances.90 Unfortunately, this inquiry is inherently speculative. To inject some certainty, courts have often attempted objectively to determine what the reasonable expectations of the buyer would be given the relevant market.91

The market factors which the Seventh Circuit considered in upholding the contract itself and the use restriction it contained are applicable in the unconscionability assessment. The court emphasized the fact that the use restriction was reasonable under the circumstances-price discrimination between commercial and consumer users allowing the seller to market the product at a lower price. In other words, the court implied that for the price the buyer paid, the buyer should have expected the use restriction and therefore should be held to it despite the fact that, as a practical matter, it may never have read the license agreement.

The importance of this market inquiry cannot be overemphasized. The UCC is based on an economic rationale and clearly contemplates the introduction of relevant market evidence in Code cases. Thus, the market inquiry in the context of contract formation is obviously an appropriate one for a court to make. That market inquiry suggests that the use restriction should be upheld as a matter of contract law. However, just as stating that the shrinkwrap is an enforceable contract says little or nothing about the enforceability of particular provisions, using market evidence to uphold a particular provision as a matter of contract law says little or nothing92 about whether that provision should be upheld as a matter of federal copyright law.

B. The Enforceability of the Use Restriction Under Federal Copyright Law

The issue of whether the use restriction was unenforceable as preempted by the Copyright Act would have been rendered moot if the shrinkwrap were held unenforceable as a matter of contract law. However, the district court, which did in fact hold the agreement unenforceable, still commented on copyright preemption.93 In contrast, the court of appeals had no choice but to face the preemption issue since it held the shrinkwrap enforceable, reversing the contractual holding of the district court.

In the ProCD case, the conflict between federal copyright policy and state enforcement of a private contract is easy to perceive. This conflict arises because of the Supreme Court's decision in Feist Publications, Inc. v. Rural Telephone Service Co.94 The Court in Feist held that the particular compilation of data in the white pages directory at issue was not copyrightable.95 Enforcement of the contract in ProCD seems rather obviously to be at odds with this holding as it gives copyright rights to noncopyrightable data. Thus, a review of Feist is helpful in framing the issue prior to discussing the respective courts' preemption holdings in ProCD.

In Feist, Rural Telephone Service Co. was a public utility which published a conventional phone book containing white pages listings of its subscribers.96 Feist sought to publish an area-wide phone book which would encompass Rural's subscribers as well as those of other utilities.97 Rural refused to license its white pages listings to Feist and sued Feist for copyright infringement when a number of its listings were allegedly copied into Feist's area-wide directory.98

The Court began its decision by noting that "[F]acts are not copyrightable . . . [while] compilations of facts generally are."99 Facts are not copyrightable because they lack the originality which is constitutionally required before a work may merit copyright protection:

"No one may claim originality as to facts." This is because 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 or she has merely discovered its existence.100

The Court went on to hold that while a compilation of facts may possess the requisite originality in the selection and arrangement of the data, Rural's "garden-variety white pages directory, devoid of even the slightest trace of creativity," lacked originality and therefore could not qualify for copyright as a compilation.101

In support of its decision, the Court interpreted the Copyright Act as embodying a strong policy of "encourag[ing] others to build freely upon the ideas and information conveyed by a work."102 While it may "seem unfair" that a second-comer could reap the benefits of another's initial investment in gathering the particular facts, the Court stated that "[t]his result is neither unfair nor unfortunate. It is the means by which copyright advances the progress of science and art."103

The Court's decision in Feist effectively overturned a long line of judicial authority, often labeled as "sweat of the brow" cases.104 Courts in "sweat of the brow" jurisdictions were willing to afford compilers copyright protection based on the expenditure of labor inherent in the effort to collect and arrange the particular facts.105 In fact, interestingly, the Seventh Circuit (which upheld the ProCD contract) has been described as a "bastion of the 'sweat of the brow theory.'"106 The Court in Feist rejected the "sweat of the brow" theory, stating:

In summary, the 1976 revisions to the Copyright Act leave no doubt that originality, not "sweat of the brow," is the touchstone of copyright protection in directories and other fact-based works. Nor is there any doubt that the same was true under the 1909 Act.107

In ProCD, ProCD attempted by private contract to recover the copyright protection which Feist had clearly withdrawn. The preemption issue is thus rather sharply defined in ProCD: may a party create, by contract, copyright-type rights in data which is not afforded copyright protection under the Supreme Court's interpretation of the Act? Put another way, may a party by private contract alter the policy balance struck by Congress with respect to copyrightable subject matter?

1. The differing views of the district court and court of appeals

Both the district court and court of appeals addressed the question of preemption under 301(a) of the Copyright Act.108 Section 301(a) provides that "all legal or equitable rights that are equivalent to any of the exclusive rights within the general scope of copyright . . . and come within the subject matter of copyright . . . are governed exclusively by this title."109 The issue which occupied most of the analysis of both courts was whether the breach of contract claim implicated rights "equivalent" to any of the exclusive copyright rights. The test for equivalence has often been stated as whether or not the noncopyright claim contains an "extra element" which renders it "qualitatively different" from a pure copyright cause of action; if it does not, then the noncopyright cause of action is preempted.110

According to the district court, ProCD's breach of contract claim was preempted by the Copyright Act. The court stated: "[ProCD's] breach of contract claim is nothing more than an effort to prevent defendants from copying and distributing its data, exactly what it sought to bar defendants from doing under copyright law."111 In other words, the very act that breached the license agreement infringed the copyright. The district court rejected the argument that merely because the breach of contract claim requires a showing of the additional element of breach it is qualitatively different from a copyright claim.112

The district court also discussed copyright policy in its preemption holding. In particular, it characterized the use restriction as "an attempt to avoid the confines of copyright law and of Feist."113 It found the use restriction to be fundamentally inconsistent with the copyright policy of favoring accessibility to information expressed in Feist:

It is only when a contract erects a barrier on access to information that under copyright law should be accessible that 301 operates to protect copyright law from individually crafted evasions of that law. . . . Plaintiff cannot use a standard form contract to make an end run around copyright law.114

Transaction costs for ProCD itself would have been high if the compilers of each of the 3000 directories which ProCD used to assemble its database had included a similar use restriction.115

The court of appeals took a much different view of the preemption question, making much of the distinction between rights established by law and rights between parties. The court characterized copyright as a legal right against the world which "restrict[s] the options of persons who are strangers to the author."116 Contracts, on the other hand, "generally affect only their parties; strangers may do as they please, so contracts do not create 'exclusive rights.'"117 The court analogized to contracts involving trade secrets which often require the recipient of the trade secret to treat noncopyrightable data as if it were copyrighted.118 The court characterized relevant Supreme Court precedent in the trade secret area as "hold[ing] that contracts about trade secrets may be enforced-precisely because they do not affect strangers' ability to discover and use the information independently."119

The court of appeals did not expressly use the "extra element" test to explain why it thought the breach of contract action to be qualitatively different from a copyright infringement claim. It emphasized, however, that daily transactions often involve promises involving intellectual property, and such promises are routinely enforced.120 For example, video rental stores often limit the use of the rented tape to home viewing and require the tape's return in a specified number of days.121 No one would suggest that the customer could keep the tape longer because 301(a) renders the customer's promise unenforceable.122 In a context closer to ProCD, the court noted that LEXIS-NEXIS often places use restrictions on the data it makes available.123 The court compared the two, stating:

ProCD offers software and data for two prices: one for personal use, a higher price for commercial use. Zeidenberg wants to use the data without paying the seller's price; if the law student [, using LEXIS under a license priced for educational use only, may not resell to law firms for commercial use,] neither can Zeidenberg [avoid the ProCD use restriction].124

The court stopped short of a per se rule that all contracts would survive a preemption analysis: "[W]e think it prudent to refrain from adopting a rule that anything with the label 'contract' is necessarily outside the preemption clause: the variations and possibilities are too numerous to foresee."125 The court did attempt to answer the policy question of how the use restriction could be consistent with the copyright policy articulated in Feist of maintaining a viable public domain: "Enforcement of the shrinkwrap license may even make information more readily available, by reducing the price ProCD charges to consumer buyers."126 It concluded by stating that "whether a particular license is generous or restrictive, a simple two-party contract is not 'equivalent to any of the exclusive rights within the general scope of copyright' and therefore may be enforced."127

2. An evaluation of the competing views

The confusion and inconsistency of preemption law makes preemption issues notoriously difficult to resolve. Preemption issues implicate broad policy questions and require courts to articulate the rationale for a particular legislative enactment and to decide how that enactment fits within the overall legislative plan and the common law generally. In part, the reason for the two courts' disparate holdings on preemption is that the district court concentrated primarily on the relationship between copyright and contract at an abstract level, while the court of appeals addressed these two areas from both a theoretical and practical perspective. In particular, the court of appeals considered how the preemption decision would impact the market as well as how the market should influence the preemption decision.

Commentators and courts have never definitively resolved the question of whether a breach of contract claim in which the act which constitutes breach is also the act which would infringe the copyright is preempted.128 However, as a general rule, breach of contract claims are not preempted by copyright because the additional element of breach of promise distinguishes the contract claim from one sounding in copyright.129 The district court rejected precedent to this effect while the court of appeals characterized it as "sound."130

Thus, one may infer that the difference between the two courts' decisions was disagreement over the existence of an "extra element" that would render the breach of contract action qualitatively different from copyright and thereby save it from preemption. The district court saw no such extra element, stating that "[c]ontracts that seek to protect reproduction and distribution rights step into the territory already covered by copyright law."131 In contrast, one may infer from the court of appeals' opinion that it found an extra element in the breach of promise. In the nonnegotiated shrinkwrap context, this promise or consent must be inferred by conduct since it is never expressly given. The court's discussion of market factors supports this reading of its decision. Market factors may help a court decide whether it is reasonable to assume that a purchaser either actually assented to the particular terms or would have assented had it been aware of them. However, the question remains as to whether such an inquiry is authorized under the Act.

III. A Deeper Analysis of the Preemption Issue

The preemption decision is not as clear cut as either court presented it. In fact, it is a difficult issue with compelling arguments on both sides. Because of the importance of its resolution to the computer software industry, it merits further analysis. Specifically, lawmakers should identify and weigh competing policy considerations in an effort to form a consistent rule-even if that rule incorporates flexible standards rather than articulating a bright-line test. In the meantime, in the absence of legislative action, courts should continue to consider relevant market evidence in defining the relationship between copyright and contract in particular cases.

A. Two Competing Views of Copyright-Freedom of Contract v. Public Domain

The particular ProCD issue-defining the allowable range of contract with respect to noncopyrightable data-brings two extreme competing views of the purpose of copyright into focus.132 One view (for simplicity, referred to as the "freedom of contract" view) contends that copyright merely provides a bundle of rights to copyright owners to help them avoid the transaction costs of contracting with each purchaser of the copyrighted material.133 In this model, copyright functions much like Article 2 of the UCC-as a boilerplate contract to govern the parties' relationship in the absence of a contrary agreement.134 Nothing prevents the parties from contracting around the rights granted by copyright or, in the case of data not protected by copyright, from creating their own property rights through private contract. In the freedom of contract model, any distinction between the contract and preemption inquiries collapses. As long as the contract passes muster under traditional common law contract and the UCC, its provisions-even those affecting copyright-type rights-are not subject to preemption.

Another view (for simplicity, referred to as the "public domain" view) argues that copyright represents a legislative scheme carefully balanced to advance the public interest by providing an incentive to authors to create while safeguarding the free flow of the information on which such creativity is based.135 In this model, there are immutable rules around which the parties cannot contract because the public interest cannot be sacrificed on the altar of two-party agreements.136 The problem is in defining exactly which provisions of the copyright law are immutable and which are not, in a manner more definitive than simply stating "We know immutable rules when we see them."137 The public domain model thus distinguishes between the contract and preemption inquiries. Merely because a contract is enforceable under traditional contract law does not mean that it is not preempted by copyright law. The difficulty lies in identifying under what circumstances provisions of an otherwise enforceable contract are preempted.

B. An Evaluation of the Competing Views

The freedom of contract view has some intuitive appeal. The copyright system itself is primarily based on an economic rationale: correcting defects inherent in the market for public goods.138 The motivating factor behind the pre-Feist sweat of the brow holdings139 was in some sense market failure. The compiler of facts could invest a substantial sum in the discovery and collation of facts. Copyright provided a ready mechanism to allow the factual compiler to recoup its investment.

After Feist, the withdrawal of copyright protection for factual compilations resulted in a market defect in that compilers could no longer recoup their investments. To counteract that market imperfection, compilers began to enter into contracts privately, to establish the copyright protection that the public law had withdrawn. Adherents of the freedom of contract view would argue that, since copyright itself is a response to market imperfections, it should be permissible to use private contract to overcome a market imperfection caused by the withdrawal of that copyright protection.

The supporters of the public domain model would respond in two ways, with both economic and non-economic arguments. From an economic perspective, while it is true that facts are public goods requiring some incentive mechanism to encourage their discovery, facts also constitute the building blocks for further creativity. If copyright protection extended to facts, some creative expression would be foregone. Second-comers would be unable to build on those facts without incurring the transaction costs of locating the copyright owner, negotiating with her and paying her a license fee. From a systemic perspective, then, the refusal to protect facts may be economically justified because protection may result in a decrease rather than increase in creative activity. To allow parties to contract otherwise would frustrate the carefully crafted balance which Congress has constructed.

Noneconomic justifications may also be offered for refusing to enforce use restrictions applicable to facts. The decision to withhold copyright protection from facts is based, in part, on the belief that they should be free for all, irrespective of whether protection would encourage or discourage creativity at the end of the day. Individual parties should not be free to create barriers to this information which Congress has deemed should be readily accessible. Another way to put it is that copyright law is based on other policy considerations in addition to economics. Therefore, any particular contract which is enforceable under, for example, Article 2 of the UCC, may still be subject to preemption if, under the particular circumstances, other policies trump those of freedom of contract.

The concerns of the public domain theory are particularly evident in the shrinkwrap context. Because the forms are both standard in content and pervasive, they resemble private copyright law. A state clearly could not enact a valid statute which extended copyright protection to facts, yet by enforcing ProCD type provisions in its courts, the state effectively achieves the same result. The rights set up by the shrinkwrap thus begin to resemble copyright rights against the world rather than mere contractual rights between two parties.

Moreover, a long-standing doctrine of intellectual property law states that the purchaser of a product is free to do with it as he wishes, except for engaging in acts which would violate intellectual property rights.140 For example, no one would suggest enforcing a boilerplate contract for a new car that provided for breach if the car were used on the Massachusetts Turnpike or a label on a Coke can that prohibited reverse engineering. More to the point, it seems likely that if the conventional phone book were distributed with a tear-off strip containing a clause that attempted to limit the recipient's use of the book, such a clause would not be enforced. Why then can't the purchaser of the factual database use and dispose of the facts it just bought in the same way that the car purchaser can drive the car on the Massachusetts Turnpike, the Coke purchaser may reverse engineer the soda, and the recipient of the hard copy phone book may do with it as he pleases? This illustration encapsulates the essence of the public domain argument-enforcement of the shrinkwrap use restriction frustrates fundamental intellectual property doctrine.

C. Alternative Approaches

The optimal position on the spectrum running from the freedom of contract model to the public domain model probably lies somewhere in between the two. To reach this position, courts must descend from the realm of the purely theoretical to decide concrete cases and reconcile competing interests. In so doing, they must use a decision-making model that integrates considerations such as economic and other public policy concerns. The following discusses the decision-making models that courts could use in deciding whether to enforce contractual terms which restrict the copyright rights that purchasers would otherwise have.

1. The Easterbrook approach

Obviously, one model is that which Judge Easterbrook employed. Under this model, contractual terms are generally not preempted since they affect only their parties and do not create rights against the world in the same manner that the public copyright law does. Thus, there would be little reason to preempt provisions of contracts modifying copyright rights where those contracts are entered into by two knowledgeable, informed parties.141 Such a freely bargained contract would, as Judge Easterbrook said, "generally affect only its parties"142 and would not set up by private contract a copyright scheme that competes with federal law.

Standard form contracts are more troublesome because they come closer to establishing a competitive private copyright scheme. Yet standard form contracts are not invidious in and of themselves, but may simply be a mechanism enabling a mass market to evolve where otherwise one might not exist. Under the Easterbrook model, a court assessing restrictive terms contained in a standard form contract should analyze relevant market factors as Judge Easterbrook did in ProCD.143 If the restriction is reasonable under the circumstances, the court should infer the purchaser's consent. This consent would provide the "extra element" which would save the agreement from preemption under 301. As Judge Easterbrook said, paraphrasing the Supreme Court:

Terms and conditions offered by contract reflect private ordering, essential to the efficient functioning of markets. Although some principles that carry the name of contract law are designed to defeat rather than implement consensual transactions . . . the rules that respect private choice are not preempted by a clause [such as 301 which is aimed at] prevent[ing the] states from substituting their own regulatory systems for those of the national government.144

Under the Easterbrook model, where the particular term modifying copyright rights is reasonable, it is not preempted even if it is not bargained for and even if the party whom that term is to bind is unaware of its existence.

Presumably, if a party could demonstrate that a market is not functioning efficiently, a court could make a deeper inquiry into enforceability. Judge Easterbrook left the nature and extent of that inquiry unspecified. However, one may infer that it would focus on the flexible doctrine of unconscionability.145 If a party could show that a market were inefficient, under the Easterbrook model, a court might set aside a contract or particular terms thereof as unconscionable.

The Easterbrook model embodies the freedom of contract approach to copyright by collapsing the contract and preemption issues. Courts should, however, recognize the fact that the theoretical underpinnings of contract and copyright are different. A more nuanced approach that takes this difference into account might therefore be preferable.

2. The "Default Rules" approach

Another decision-making model that a court might use in assessing use restrictions is one based on an analogy to the models described in the default rules literature.146 This literature classifies rules as either default or immutable.147 Parties are free to contract around default rules but cannot vary immutable ones.148 Historically, the default rules literature has primarily been concerned with filling contractual gaps.149 However, it could also be used to help address the issue of whether the norm for contractual provisions which purport to restrict copyright rights should be one of enforcement or nonenforcement.

An adherent of the public domain view would contend that the Act's refusal to provide protection for facts is an immutable rule. Generally, immutable rules may be justified when "unregulated contracting would be socially deleterious because parties internal or external to the contract cannot adequately protect themselves."150 In other words, either paternalism or externalities can be used to justify classifying a particular rule as immutable.151

Neither of these two considerations offers a convincing case for branding the Act's rule against protection for facts immutable. Parties who agree to contracts which restrict copyright rights are usually protected in some manner. When parties are informed, they are protected by their own knowledge. When they are uninformed, they are protected by the impersonal workings of the market; the market helps to ensure that the provisions of even standard form contracts are reasonable. Paternalism, then, does not support labeling lack of protection for facts an immutable rule.

The externality argument for an immutable rule is somewhat stronger. Since facts constitute the building blocks for creative activity, restrictions on the use of facts (particularly if they become pervasive) could substantially increase costs. Any benefit from a use restriction would be offset by the costs of expression foregone. Unfortunately, this argument is inherently speculative. Virtually all contractual terms impose costs on someone or something. If this were the only criterion for making a rule immutable, then it would be a short trip down the slope to making all contract rules unchangeable. This result is not desirable generally or in this case. In the absence of empirical evidence on the impact of use restrictions which demonstrates that their costs exceed their benefits, labeling the copyright rule immutable seems unwise without evidence of congressional intent to that effect.

If the Act's rule against protecting facts is not immutable, then the issue devolves to the question of whether the default rule should be one which enforces use restrictions or not. At first glance, the default rules literature seems to argue for the Easterbrook approach. The conventional theory is that the default rule should be set at what the parties would have agreed to had they negotiated over the particular issue.152 Evidence of usage of trade and considerations of economic efficiency help a court to decide what rule the parties would have agreed to had they bargained.153 In this context, the conventional theory would argue for a default rule of enforcement. Usage of trade and economic efficiency both support the proposition that the parties would have agreed to the use restriction had they bargained over it.154 Therefore, while standard form agreements may resemble private legislation, their terms in fact mirror what the parties would have agreed to in a negotiation and therefore should be enforced.

However, another principle revealed in the default rules literature is that the default should be drawn against the relatively informed party to give that party an incentive to reveal information.155 In the ProCD context, this test may argue for a default rule of nonenforcement of use restrictions. The ordinary consumer purchaser of software is likely to be uninformed about the relevant law. To the extent that such purchaser is informed, its expectations are likely to be set by reference to the bargain struck by the Copyright Act156-a bargain which permits the free use of facts.157 If purchasers expect copyright to apply to the data they purchase and fail to notice the use restriction, they may pay too much for the product. A penalty default of nonenforcement would correct this market imperfection of asymmetric information by encouraging sellers to disclose information. The use restriction would generally be enforceable if the seller brought it to the purchaser's attention. In a negotiated contract, the seller would expressly bring the desired term to the buyer's attention in the course of writing it into the contract. In the case of the standard form contract, the seller could not expressly make the buyer aware of the use restriction but could put the buyer on notice of it by making the term conspicuous. The default rules model, then, would support a rule of nonenforcement unless the use restriction were conspicuous.

3. Objections to the default rules model

There is an objection to this suggestion of a nonenforceability norm, which Judge Easterbrook noted in assessing the use restriction's enforceability. In rejecting the argument that the license or its terms should have been conspicuous, he stated, "Competition among vendors, not judicial revision of a package's contents, is how consumers are protected in a market economy."158 In other words, in an efficient market, there is no reason to require conspicuousness as a legal rule. The terms which the seller offers are desirable from both the buyer's and seller's perspective, regardless of whether or not the buyer notices a particular term. The product price adjusts in light of the package of terms offered even where the buyer is uninformed.

Moreover, under certain conditions, the market itself will move sellers to make certain terms conspicuous. For example, sellers often bring warranties to the buyer's attention as a way to differentiate their products.159 In the ProCD case, ProCD had an incentive to make the use restriction obvious. As the Seventh Circuit noted, the use restriction was the primary means through which ProCD put its price discrimination strategy into effect.160 ProCD therefore had an interest in making sure the purchaser understood that restriction. Of course, ProCD could always seek to enforce the restriction judicially after breach by a purchaser; hence the suit against Zeidenberg. However, litigation is expensive and damages are not always compensatory. Moreover, the damage may be irreparable. Electronic information is much easier to copy and make broadly available than hard copy publications. A user can upload data to the Internet and make it accessible to millions of users in just a few keystrokes, substantially decreasing the size of the information originator's market.161 ProCD therefore had a strong incentive to inform purchasers of its CD-ROM that such activity would not be tolerated.162 If the market functions to encourage the seller to disclose information then a legal rule with the same end seems at best superfluous.

These objections to a default rule of nonenforcement reflect different assumptions about the market. The default rules approach makes the most sense when the market is inefficient, while the Easterbrook approach is most suitable for an efficiently functioning market. Markets fall into both categories, but legal rules are usually cast in general terms. Thus, a default rule of nonenforcement might be preferable because it protects purchasers in inefficient markets while not materially raising costs or otherwise interfering with the operation of an efficient market.

The overall impact on litigation costs under the alternative rules is speculative. A default rule requiring a use restriction to be conspicuous could be cheaper because sellers who failed to comply with it would not survive a summary judgment motion; otherwise, both models would require some evidence regarding market condition. The Easterbrook model would enforce the terms in an efficient market and make a deeper market inquiry in deciding whether a particular term is unconscionable in the case of an inefficient market. The default rules model would also still consider market factors. Making the term conspicuous would not provide a safe harbor-other contractual doctrines such as unconscionability would continue to apply. The two tests might therefore often arrive at the same results.163

However, other policy reasons may favor adopting the default rules analysis. Even if the market were efficient, it may be appropriate to put purchasers on notice of the fact that the copyright rights they are accustomed to in the hard copy world are being modified in the soft copy one. The default rules approach would further the Code philosophy of encouraging efficient contracting while also recognizing the copyright policy favoring the free use of facts. Under the default rules model, purchasers would be less likely inadvertently to surrender the rights they have under copyright law since their relinquishing such rights would be brought to their attention. This notice would help to make the private legislation of the shrinkwrap more closely resemble a true negotiated agreement in which parties have agreed to restrict their copyright rights in exchange for the product they purchased.

The default rules approach thus offers a flexible model for courts to use. At the outset, a court could consider whether the particular copyright rule around which the parties are contracting is immutable or not. If it is not, courts may still consider copyright policy in determining under what conditions the terms which vary the social bargain struck by the Copyright Act should be enforced.

4. Judicial authority for adopting a decision-making model

Both the Easterbrook and default rules models have difficulties. Relevant market evidence would have to be introduced under both, as well as evidence of congressional intent under the default rules model. More troubling, however, is the issue of whether either a market or a detailed policy inquiry is an appropriate one for the judiciary to make. Normally, the balancing of competing interests is a matter for the legislature to consider when drafting a statute. The legislature has the institutional competence to weigh competing viewpoints and arrive at a conclusion. The unelected judiciary largely lacks such institutional competence.

Moreover, 301 by its literal terms does not authorize a detailed market analysis.164 Its inquiry is rather mechanical. However, in cases like breach of contract where the "extra element" is breach of promise, the market analysis cannot be avoided in the context of nonnegotiated agreements. The promise is not express, yet it may in fact be perfectly reasonable to infer it. Thus, unless 301 is interpreted to allow market evidence as an aid in determining whether a real promise exists, the provisions of most standard form agreements affecting copyright-type rights would be preempted. This result does not seem to be desirable in light of congressional silence on the topic.

In reality, courts have for years been conducting market analysis in copyright cases, suggesting that they do have the institutional competence to balance competing interests in cases like ProCD. In particular, courts historically have examined market evidence in assessing alleged infringers' claims of fair use under 107 of the Act. Fair use is an equitable defense to an infringement claim, allowing a court to use public policy grounds to excuse conduct that would otherwise be infringing.165 In conducting a fair use inquiry, a court is to balance the four non-exclusive statutory fair use factors:

(1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; (2) the nature of the copyrighted work; (3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and (4) the effect of the use upon the potential market for or value of the copyrighted work.166

Although the Supreme Court has emphasized the necessity of weighing all four factors in light of the Act's statutory purpose before making a fair use determination,167 courts have emphasized the first and the fourth.168

For example, a fair-use-type inquiry would ask whether the ProCD use restriction was reasonable in light of the market for fact-based databases. It would also ask if Zeidenberg's conduct were to become widespread, what impact would that have on the database market and the incentives of others to create similar works? These are exactly the types of considerations which the Seventh Circuit emphasized. The Seventh Circuit's decision in ProCD could be viewed as almost a "reverse" fair use inquiry.169 Rather than using market factors to assess whether an infringer's conduct should be excused, the Seventh Circuit used market factors to determine whether a copyright owner's use of contract to buttress its copyright rights-or lack thereof as in ProCD-would constitute "fair use" under the Act and therefore survive preemption. Its analysis, rather than being revolutionary, seems fully justified given that 301 must authorize a market inquiry or else many standard form contracts would be preempted. Moreover, the already existing 107 market inquiry forms a principled basis from which the court could conduct its analysis. In other words, the Seventh Circuit, in balancing competing interests, was not exceeding its institutional competence but rather using the flexible approach already implicitly authorized in the statutory structure of the Copyright Act.

The default rules approach is more difficult to support under current statutory wording. There is scant authority in either 301 or the UCC for a court to impose a requirement of conspicuousness on the parties. A court might seek to do so under the general requirement that "[e]very contract or duty within [the UCC] imposes an obligation of good faith in its performance or enforcement."170

There are, however, difficulties with this approach. Courts are split on the issue of whether a mere allegation of lack of good faith, without more, may state a cause of action under the UCC.171 Second, using good faith to imply such a requirement seems fundamentally at odds with a vision of the Code as a comprehensive statement of the law. If courts may use good faith to imply new contractual duties, the certainty and concomitant cost reduction which the Code was designed to achieve may be seriously compromised. Moreover, it would be unfair to impose a new requirement on an unsuspecting seller, at least in the first instance.

Thus, implementation of the default rules approach may require an amendment to the UCC to make the conspicuousness requirement express. This amendment could be easily incorporated into the proposed Article 2B which addresses licensing.172 In the meantime, however, the Easterbrook model, while perhaps not ideal, does provide room for protecting purchasers against unreasonable use restrictions.

IV. Conclusion

Electronic technology is different from the hard copy publications for which copyright law was designed. The sheer volume of academic commentary debating the merits of applying the existing copyright regime or devising some new form of protection for electronic works is testimony to the difficulties of adapting the law to deal with technology which did not exist at the time the law was enacted.173 Where a genuinely new technology emerges, it may be appropriate to defer to the market in the absence of specific congressional intent to the contrary. Congress first comprehensively considered the impact of new technology on the copyright laws in 1979.174 In computer science, 18 years is an eternity, encompassing a number of hardware and software generations. The technology and its uses have changed a great deal since 1979. It seems appropriate therefore for a court to consider the nature of the subject matter and how the market for that subject matter works, rather than blindly applying copyright principles that may no longer fit.

In fact, Congress has attempted to "catch up" with the technology. The recent report of the National Information Infrastructure Task Force175 and flurry of legislative proposals176-including one to extend a new form of protection to on-line databases177-demonstrate that at least some members of Congress believe that adjustments should be made to the Copyright Act to clarify the manner in which its rules should be applied to electronic technology. Prior to congressional action, allowing parties to structure their deals as they see fit, within the confines of traditional contract law, seems reasonable.178

Moreover, allowing parties to enter into contracts is not synonymous with granting them a license to enforce all the terms of such contracts, no matter how onerous or how much at odds with public policy they may be. The governing law, be it common law contract or the UCC, has never been one of unfettered freedom of contract. The same limits that have always applied continue to apply to contracts involving electronic data. Additionally, even in making a market inquiry, the courts should be willing to look outside of contract doctrine-for example, to the antitrust laws-in making the determination as to whether or not a particular agreement or term thereof should be enforced given a particular market and its defects.179

Certainly, both the Easterbrook and default rules approaches lean much more toward the freedom of contract model than to the public domain model of copyright law. This seems appropriate for the reasons set forth above, but it also highlights the need for systematic treatment of the issue of how to define the contract/copyright boundary. In continuing to address legislative proposals dealing with electronic technologies, Congress should consider this question thoughtfully and perhaps offer more guidance to the judiciary than the current 301. In particular, Congress should clarify its intent with respect to which copyright rules are immutable and which are subject to change by private agreement, either by amending the Copyright Act or by enacting new legislation to address electronic technology. In the absence of that clarification, courts should continue to inform their decisions with reference to market considerations and by analogy to other areas of law. This flexible inquiry seems consistent with the Copyright Act's overall approach and well within judicial expertise.