UNIFORMITY, CHOICE OF LAW AND SOFTWARE SALES

INTRODUCTION

Over the past four years, The National Conference of Commissioners for Uniform State Laws (NCCUSL) and the American Law Institute (ALI) have drafted numerous proposals for new Article 2B of the Uniform Commercial Code (UCC 2B). The motivation for the work was the perception that UCC Article 2, which deals with sales of goods, fits poorly with licensing of software and other computer information.1 The proposed Article promoted a uniform law for licensing these products.2

From the beginning, academics3 as well as consumer groups and industry groups affected in various ways,4 strongly criticized drafts of the proposed UCC 2B. The critics focused on the extent to which the proposed statute expanded the power and scope of contracts over existing law, especially with respect to its treatment of mass-market licenses. Critics also argued that, by allowing private ordering, the proposed uniform law interfered with the operation of numerous federal laws, including those relating to intellectual property, bankruptcy, and consumer protection. Although the drafters attempted to respond to the many criticisms in subsequent drafts of UCC 2B, they could not achieve a consensus. On April 7, 1999, the ALI and NCCUSL announced that they were abandoning the attempt

* Associate Professor, George Mason University School of Law.

** Foundation Professor, George Mason University School of Law.

The authors would like to acknowledge financial support from the George Mason University School of Law, and helpful comments from Todd Zywicki. Any remaining errors are the authors’.

1 See Raymond T. Nimmer, Article 2B Preface: Meeting the Information Age (regarding the 12/01/95 Draft), available at Carol A. Kunze, The 2BGuide: A Guide to the Proposed Uniform Computer Information Transactions Act (visited Nov. 6, 1999) <http://www.2bguide.com/drafts/html>, or directly at <http://www.lawlib.uh.edu/ucc2b/pref1201.html> (visited Nov. 6, 1999).

2 For summaries of the history of the UCC 2B project, see id. See also David A. Rice, Digital Information as Property and Product: U.C.C. Article 2B, 22 U. DAYTON L. REV. 621, 627-28 (1997).

3 See infra discussion Part III. An up to date list of articles discussing the proposed UCC 2B and UCITA is available online at Carol A. Kunze, The 2BGuide: A Guide to the Proposed Uniform Computer Information Transactions Act (visited Nov. 6, 1999) <http://www.2bguide.com>. These papers, from an academic symposium, were published variously in Symposium, Intellectual Property and Contract Law in the Information Age: The Impact of Article 2B of the Uniform Commercial Code on the Future of Transactions in Information and Electronic Commerce, 13 BERKELEY TECH. L.J. 809 (1998) and 87 CAL. L. REV. 1 (1999).

4 Critics include the Motion Picture Association of America, Society of Information Management, Recording Industry Association of America, Newspaper Association of America, National Association of Broadcasters, National Cable Television Association, and Magazine Publishers of America, librarians and the Federal Trade Commission. See Larry T. Garvin, The Changed (and Changing?) Uniform Commercial Code, 26 FLA. ST. U. L. REV. 285, 348-49 (1999) (discussing opposition to UCC Article 2B). Letters and articles by opponents of the proposed laws have been collected at Carol A. Kunze, The 2BGuide: A Guide to the Proposed Uniform Computer Information Transactions Act <http://www.2bguide.com> (visited Nov. 6, 1999).

to make the proposal part of the UCC.5 Instead, the NCCUSL announced that it was moving forward with a freestanding uniform act, The Uniform Computer Information Transactions Act (UCITA).6 NCCUSL promulgated this act at its annual meeting on July 29, 1999.7 Some of the more controversial issues surrounding UCITA are discussed below.

So far, the debate on these issues has focused on the substantive terms of the proposals. Both sides implicitly assume that all states should adopt the same law on this issue, and that promulgation of UCITA or UCC 2B will eventually lead to uniformity. The debate misses the mark in three important respects. First, it is far from clear that there should be a NCCUSL or ALI-sponsored state law in this area. Even if uniformity is desirable, it should emerge spontaneously and not from the political processes of uniform lawmaking, which is aimed at maximum state law uniformity, rather than merely efficient uniformity. Moreover, even if the same law should govern all transactions, this result is possible whether or not all states adopt the same law.

Second, it is even less clear that the law will be uniform. In fact, the end of the uniform lawmaking process is just the beginning of the process of adoption by state legislatures. Given the controversy over UCITA, many states may refuse to adopt the law in any form, adopt it in a significantly modified form, or adopt the law verbatim but restrict its effect through mandatory rules in other laws. Uniform adoption of UCITA is especially doubtful now that the law is no longer linked to the ubiquitously adopted UCC or backed by the highly respected American Law Institute.8

Third, the most important issues concern the extent to which the parties should and will be able to decide for themselves which law applies to their transactions by designating the law in their contracts. As long as state law is not completely uniform, and the parties are free to choose which law

5 See NATIONAL CONFERENCE OF COMMISSIONERS OF UNIFORM STATE LAWS & AMERICAN LAW INSTITUTE, NCCUSL TO PROMULGATE FREESTANDING UNIFORM COMPUTER INFORMATION TRANSACTIONS ACT: ALI & NCCUSL ANNOUNCE THAT LEGAL RULES FOR COMPUTER INFORMATION

WILL NOT BE PART OF UCC, April 7, 1999 (joint press release), available at <http://www.2bguide.com/docs/040799pr.html> (visited Nov. 6, 1999) [hereinafter NCCUSL/ALI

April 7 JOINT PRESS RELEASE].

6 See UNIF. COMPUTER INFORMATION TRANSACTIONS ACT (Draft for Approval at NCCUSL

Meeting, July 23-30, 1999), available at <http://www.2bguide.com/drafts.html> (visited Nov. 6, 1999) or directly at <http://www.law.upenn.edu/bll/ulc/ucita/citam99.htm> (visited Nov. 6, 1999) [hereinaf

ter UCITA APPROVAL].

7 See UNIFORM COMPUTER INFORMATION TRANSACTION ACT (Approved and Recommended

for Enactment in All the States at the NCCUSL Annual Conference Meeting, July 29, 1999, revised by the NCCUSL committee on style Oct. 15, 1999) available at <http://www.2bguide.com/drafts.html> (visited Nov. 6, 1999) or directly at <http://www.law.upenn.edu/bll/ulc/ucita/cita10st.htm> (visited

Nov. 6, 1999) [hereinafter UCITA FINAL]

UNIFORM STATE LAWS, NEW UNIFORM ACT MEETS IMMEDIATE NEEDS OF THE INFORMATION AGE,

July 29, 1999 (press release), available at <http://www.2bguide.com/docs/pru72999.html> (visited Nov. 6, 1999).

8 See Memo from David Bartlett, Amy Boss & David Rice, ALI members of UCC 2B Drafting Committee, to Uniform Computer Transactions Act Drafting Committee, May 7, 1999, (declining to participate as advisors to UCITA), available at <http://www.2bguide.com/docs/50799dad.html> (visited Nov. 6, 1999).

applies, promulgation of a “uniform” law is unimportant. Indeed, the demise of UCC 2B and the resulting diminished prospects for uniformity is a positive development. It increases the probability that individual states will enact statutes that would attract and retain information technology firms.9 Even if the law does not explicitly facilitate contractual choice of law, the applicable law nevertheless will depend to some extent on party choice rather than political dictate. Because the reach of state courts and state law is limited by contracting parties’ contacts with the relevant states, firms can exercise some control over the applicable law by locating in states with favorable laws.

Part I of this article discusses both the UCC 2B and NCCUSL proposals and the debate over the content of these proposals. This Part shows that there is considerable doubt about the proper legislative approach, and that the NCCUSL result would be unsatisfactory. This Part also raises questions about whether UCITA will be, or should be, uniformly adopted.

Part II considers the costs and benefits of a uniform law. Although uniformity might reduce transaction and information costs, respond to network externalities problems and reduce the threat of federalization, the costs of imposing a uniform law on rapidly evolving transactions are likely to outweigh the benefits. Even if a uniform law is theoretically appropriate, Part II shows that an efficient proposal is unlikely to emerge from the uniform lawmaking process. We discuss possible alternatives to the NCCUSL process, including spontaneous uniformity, model laws and standardization.

Part III discusses an alternative to uniformity-explicit enforcement of contractual choice of law. Under this approach, contracting parties themselves, and not state legislatures or uniform law promulgators, determine what legal rules apply to particular transactions. An efficient rule is far more likely to emerge from this “bottom-up” process than from the “top-down” processes of state uniformity and federal law. Part III also shows that, despite any disparate approaches that state legislators take or what uniform lawmakers do, efficient law and efficient uniformity is likely to emerge through jurisdictional competition. The extent of such competition depends ultimately on each state’s rules regarding personal jurisdiction and the states’ ability to apply their laws to particular transactions. As long as the courts require deliberate contacts to obtain jurisdiction over parties and respect the parties’ legitimate expectations, firms will be able to exercise some control over which state laws apply to their transactions.

I. THE DEBATE OVER UCITA

NCCUSL promulgated UCITA at its annual meeting on July 29, 1999.10 Accordingly, it was sent to the states for adoption in Fall 1999.11

9 See Garvin, supra note 4, at 351.

10 See UCITA FINAL, supra note 7.

11 See NCCUSL/ALI April 7 JOINT PRESS RELEASE, supra note 5.

The proposed statute includes a comprehensive set of default contract rules that cover legal recognition of electronic records and proof of authentication, assent, unconscionability, formation and terms, construction, warranties, transfer of interests and rights, performance, breach of contract and remedies. In addition, the proposed statute covers broad issues such as the relationship between state and federal laws, conflict of laws, choice of law and choice of forum. As discussed above, UCITA is a slightly modified successor to the failed attempt to cover computer information licensing with a new Article 2B to the UCC.

Although many of the substantive provisions of UCITA and drafts of UCC 2B would not have changed settled existing law,12 several provisions are quite controversial. Provisions favoring broad enforceability of “shrinkwrap” licenses sold to mass-market consumers and various issues relating to the relationship between state contract law and federal intellectual property laws are the most controversial.13 This Part does not try to settle the debate over substantive issues or review in detail the enormous literature the debate has spawned. Rather, it is intended to indicate the contentious issues that remain after the long drafting process. This dissatisfaction is a byproduct of the uniform lawmaking process.14

To illustrate the process, we focus here on UCITA’s treatment of licenses sold to the mass consumer market. No issue better illustrates the great divide that separates defenders and critics of the act. Software producers, including Microsoft and other huge companies, have strong reasons for wanting to regulate uses and transfers of their products. However, it is very difficult to design a mechanism for reaching agreement on complex terms of a license prior to sale. Accordingly, the vendors commonly use licenses that are included in the product. The consumers do not see the license until they have bought and paid for the product, taken it home, and torn off the product’s “shrinkwrap.” Consumers may object to being bound by a complex contract with little opportunity to negotiate or read it. On the other hand, sellers may be concerned that if buyers are not deemed to have given consent until some time after they have begun to use the product,

12 For a summary of this issue, see DRAFTING COMMITTEE OF UNIFORM COMMERCIAL CODE,

NATIONAL CONFERENCE OF COMMISSIONERS OF UNIFORM STATE LAWS & AMERICAN LAW

INSTITUTE, DRAFT UNIFORM COMMERCIAL CODE ARTICLE 2B – LICENSES tbl. A, (Consumer Issues:

Comparison of Existing Article 2 and Other Law with Proposed Article 2B) (8/1/98 draft), available at <http://www.law.upenn.edu/bll/ulc/ucc2b/2b898.htm> (visited Nov. 6, 1999).

13 These issues were mentioned as reasons for the ALI to defer consideration of the proposed UCC 2B until May, 1999. See Letter from Geoffrey C. Hazard, Jr., American Law Inst. Ad Hoc Committee on Article 2B, to Gene N. Lebrun, President, NCCUSL, and Charles Alan Wright, President, American Law Inst. (March 26, 1998), available at <http://www.2bguide.com/docs/ghmar98.html> (visited Nov. 6, 1999)

14 See infra Part II.

those buyers might be able to reap the benefits from the product without being bound by the seller’s limits on use. In addition, corporate information technology officers face the further problem of monitoring use of the product by hundreds or even thousands of employees, any one of whom might trigger acceptance of license terms the firm does not want.

The current UCITA draft, like previous drafts of the proposed UCC 2B, provides for enforcement of standard terms included in “mass-market” licenses15 if they are not unconscionable and the licensee manifests assent to the terms before or during her initial use of or access to the information.16 The buyer’s assent may be enforced even if the terms were presented after the initial agreement.17 The buyer manifests assent by engaging in conduct with reason to know that under the circumstances the conduct indicates assent.18 This relevant conduct may include “opening a container or commencing to use information.”19

UCC critics have suggested that it always had a pro-business orientation.20 Based largely on their treatment of mass-market licenses, critics

15 See UCITA FINAL, supra note 7, § 102(a)(43) (defining “mass-market license” as “a standard form used in a mass-market transaction”)

16 See UCITA FINAL, supra note 7, § 208 (providing that a party adopts terms of records, including a standard form, if the party agrees to the record, by manifesting assent or otherwise)

17 See UCITA APPROVAL, supra note 6, § 211(a), reporters note 4. Section 211 was promulgated as Section 209(a) in UCITA FINAL, supra note 7. Section 209(b) gives the licensee in the payfirst/terms-later setting an extensive right of return, including a refund of the purchase price plus various expenses, if the licensee does not manifest assent and was obligated to pay before having an opportunity to review. See UCITA FINAL, supra note 7, § 209(b). This approach gives licensors an incentive to make terms available prior to payment whenever possible. See UCITA APPROVAL, supra note 6, § 211(a), reporter’s note 4.c.

18 See UCITA APPROVAL, supra note 6, § 112.

19 Id. § 112, reporter’s note 5.

20 See Kathleen Patchel, Interest Group Politics, Federalism, and the Uniform Laws Process: Some Lessons from the Uniform Commercial Code, 78 MINN. L. REV. 83, 86 (1993)

have labelled UCC 2B and UCITA as pro-software licensor/anti-consumer acts.21 A closer examination of these issues, however, suggests that enactment of rules that reflect widespread industry practice is likely to promote efficiency and is not likely to harm consumers.

By enforcing the licensor’s standard form terms without express agreement at the time the transaction occurred, UCITA and the proposed UCC 2B adopt Judge Easterbrook’s view in ProCD v. Zeidenberg.22 ProCD compiled and supplemented information from more than 3,000 telephone directories into a computer database at a cost of more than ten million dollars, with substantial additional costs for updating.23 ProCD then marketed a CD-ROM version of the database, along with a copyrighted search and proprietary decompression software program, to businesses for use in compiling customer lists. ProCD also marketed the database to the general public at a much lower price. For the general public version, the outside of the box provided notice that the software came with restrictions stated in an enclosed license. This license limited use of the application program and listings to non-commercial purposes. Zeidenberg bought a consumer package, ignored the terms of the license, and made the information available on a web site for a fee. He later bought two more consumer packages containing identical licenses to obtain the updated listings. ProCD sued to enjoin further commercial dissemination of the database. The district court held that license terms not appearing on the outside of the packages were unenforceable under Wisconsin’s version of the UCC, and that, in any event, the federal copyright laws preempted enforcement of the licenses under state law. The Seventh Circuit reversed, holding that the terms of the shrinkwrap license were enforceable under the UCC and that the Copyright Act did not preempt enforcement of these terms. UCITA adopts this holding.24

Commentators have criticized UCITA and ProCD as inconsistent with the UCC and existing precedents .25 The cases on enforceability of shrinkwrap licenses decided under the UCC and prior to ProCD are distinguishable and do not support this claim .26 For example, Step-Saver v.

But see A. Brooke Overby, Modeling UCC Drafting, 29 LOY. L.A. L. REV. 645, 645-46 (1996).

21 See supra notes 3 and 4 and accompanying text.

22 86 F.3d 1447, 1449 (7th Cir. 1996). See also Hill v. Gateway 2000, Inc., 105 F.3d 1147, 1149-50 (7th Cir. 1997).

23 See ProCD, 86 F. 3d at 1449.

24 See UCITA APPROVAL, supra note 6, § 210, reporter’s note 3. Section 210 was promulgated as Section 208 in UCITA FINAL, supra note 7.

25 For a listing of articles and cases critical of ProCD on contract grounds, see Mark A. Lemley, Beyond Preemption: The Law and Policy of Intellectual Property Licensing, 87 CAL. L. REV. 111, 120 n.20 (1999) [hereinafter Lemley, Beyond Preemption]

26 See ProCD, 86 F.3d. at 1452. See also Brian Covotta & Pamela Sergeeff, Intellectual Property: ProCD v. Zeidenberg, 13 BERKELEY TECH. L.J. 35 (1998)

Wyse27 involved a “battle of the forms” between two commercial entities rather than the validity of a consumer’s acceptance of a licensor’s standard form. In such a case, UCITA directs the court to “review the entire circumstances” of the transaction and the course of dealings between the two parties.28 This position gives the court discretion not to enforce conflicting terms in the standard form, as happened in Step-Saver under UCC § 2-207 .29 Thus, UCITA does not alter the result in such cases.30

Another case decided prior to ProCD, Arizona Retail Systems v. Software Link, Inc.,31 is even less authoritative. In this case, the buyer purchased multiple copies of software. He admitted that, when the initial software copy contract was formed, he knew the terms of the license. Applying UCC § 2-207, the court enforced the terms of the shrinkwrap license for the initial sale because the buyer manifested knowing consent to the relevant term (specifically, a limitation on warranties).32 However, the court refused to enforce the same terms contained in identical license agreements on subsequent telephone orders for the same product between the same parties.33 The court reasoned that the parties formed the subsequent transactions at the time the buyer made the telephone orders. Therefore, these transactions were proposals for modification of the contract under UCC § 2-209. The court here was misguided. It is “schizophrenic”34 to enforce the terms of the initial license transaction because the buyer knew the terms while failing to enforce identical terms between the same parties in subsequent transactions because the buyer was informationally disadvantaged.

A final example of a case supposedly inconsistent with ProCD is Vault v. Quaid Software Ltd.35 The district court and Fifth Circuit con

tract Formation and Notions of Manifested Assent in the Arena of Shrinkwrap Licenses, 92 NW. U. L. REV. 379 (1997). For lists of prior cases see ProCD, 86 F.3d at 1452 and Lemley, Beyond Preemption, supra note 25, at 120 n.20 (listing cases).

27 939 F.2d 91, 103 (3d Cir. 1991).

28 See UCITA APPROVAL, supra note 6, § 212, reporter’s note 3. Section 212 was promulgated as Section 210 in UCITA FINAL, supra note 7.

29 Many of the cases critics cite involve the battle of the forms and interpretations under UCC 2-107. See Lemley, Intellectual Property and Shrinkwrap, supra note 25, at 1249.

30 See UCITA FINAL, supra note 7. If there is an acceptance with varying terms, UCITA Section 204 applies. See id. § 204. If the acceptance does not materially alter the offer, a contract is formed. If an acceptance materially alters an offer, then UCITA Section 210 applies, and the court examines the parties’ conduct. See id. § 210. Conditional offers are controlled by UCITA Section 205. See id. § 205. In “battle of the forms” cases, conditional language in a standard term precludes the formation of a contract only if the party proposing the form acts consistently with the language, as by refusing to perform, refusing to permit performance, or refusing to accept the benefits of the contract, until the proposed terms are accepted. If both parties act inconsistently with their conditional language, UCITA Section 210 applies.

31 831 F. Supp. 759 (D. Ariz. 1993).

32 See id. at 761.

33 See id. at 764-66.

34 See Lemley, Intellectual Property and Shrinkwrap supra note 25, at 1254 (noting that “[b]ecause of its schizophrenic result, Arizona Retail is at best limited authority for enforcing shrinkwrap licenses”). It might be argued with equal strength that the case is also limited authority for refusing to enforce such licenses.

35 655 F. Supp. 750 (E.D. La. 1987), af’d, 847 F.2d 255 (5th Cir. 1988).

cluded, without analysis, that the shrinkwrap license was “a contract of adhesion,”36 which could be enforced only under the Louisiana Software License Enforcement Act (SLEA).37 The assertion that contracts of adhesion are not valid or enforceable is questionable under contract law and the UCC.38 The courts’ holdings that federal law would preempt SLEA would also fall in the presence of an enforceable contract.39

Even assuming a legitimate conflict between ProCD and prior cases, the fundamental issue for legislators is a policy question: which approach to follow? There are several potential arguments for limiting freedom of contract in this setting. All of these arguments are questionable, none is conclusive. The first is that, because the terms in form contracts are not explicitly bargained for, they are “contracts of adhesion,” where one party dictates terms on a “take-it-or-leave-it” basis. Merely because explicit bargaining is not feasible, however, does not mean that consumers are unprotected. As long as the parties know of the existence and effect of the terms, and the buyer has a viable alternative to buying the product, the value of the terms may be reflected in the price .40 Moreover, users can voice complaints on the Internet .41 Thus, even if bargaining over non-price terms were impossible, a rational seller would have an incentive to design an efficient contract .42 Additionally, a rule that condemns most or all form contracts as “adhesion” contracts penalizes an efficient form of contracting that clarifies the parties’ expectations at a low cost .43 The costs of negotiating terms in each transaction occurring in mass consumer markets would be prohibitive.

A second argument against enforcing these licenses is that it is inefficient to enforce form contracts against individuals who have less than full information about their terms. In other words, if consumers are uninformed, they will pay too much for unfavorable terms. A sufficient number

36 Vault, 655 F. Supp. at 761, af’d, 847 F.2d at 269.

37 LA. REV. STAT. ANN. 51, §§ 1961-66 (West 1987 & Supp. 1999).

38 See RESTATEMENT (SECOND) OF CONTRACTS, § 210

also Deborah Kemp, Limitations upon the Software Producer ‘s Rights: Vault Corp. v. Quaid Software

Ltd., 16 RUTGERS COMPUTER & TECH. L.J. 85 (1990).

39 See infra text accompanying note 146.

40 Suppose, for example, that a buyer is willing to pay up to $100 dollars for a license to use a database. Suppose that the seller adds a term that is worth $20 to the seller but costs the buyer $30. The amount the buyer is willing to pay will fall by $30. If the seller is already extracting most of the surplus from the bargain, the price will fall by approximately $30. The net result is that the seller gains $20 from addition of the term, but loses $30 from the reduction in the price of the license. See David Friedman, In Defense of Private Orderings: Comments on Julie Cohen ‘s “Copyright and the Jurisprudence of Self-Help “, 13 BERKELEY TECH. L.J. 1151, 1158 (1998). To be sure, it might be argued that sellers can impose terms at unfavorable prices not because of the way the transaction is structured but because they have “excessive” market power. However, this argument would extend beyond the consumer context, and indeed beyond contract law, into the realm of antitrust.

41 See Robert W. Gomulkiewicz, The License is the Product: Comments on the Promise of Article 2B for Software and Information Licensing, 13 BERKELEY TECH. L.J. 891, 898 (1998).

42 Id. at 1156. Note that the above argument does not depend upon the market position of the seller-i.e., we did not have to assume that the seller was in a competitive market. This result illustrates the general principle that even monopolists benefit by producing efficiently.

43 See Friedman, supra note 40, at 1158.

of informed consumers transacting in a mass market, however, would result in efficient contracts for both the informed and uninformed.44 Given the prohibitive costs of discriminating between informed and uninformed in this setting, competition for the marginally informed consumer protects the uninformed consumer from unwanted terms .45 It is reasonable to assume that there are enough informed consumers in the market to provide these market effects, at least for computer information transactions, for which readily accessible information abounds and where many consumers are highly educated and technically adept.

Thus, the case law is at best ambiguous on how enforceability of shrinkwrap licenses is treated under general contract and commercial law. More importantly, because UCITA’s mission is to deal with fundamental policy issues rather than count cases, court rulings should not be determinative. The arguments against enforcing terms contained in standard forms do not make a convincing case for per se condemnation of form contracts .46 Nevertheless, the UCITA approach has already engendered strong opposition from those who seek limits on freedom of contract .47 The anti-contractarian opposition to UCITA does not mean that UCITA is likely to please all of those who favor freedom of contract. Perhaps most importantly, UCITA imposes strong limits on contractual choice of law, a critical, although not fatal, departure from contractual freedom.48

The controversy over UCITA illustrates the difficulty of making the right policy choice, as well as the peril uniform lawmakers face when determining the set of rules in a complex and dynamic area of the law. The outcome reflects the politics and defects of the uniform lawmaking process . 49 Because uniform lawmakers seek to maximize adoptions, they must cater to the preference of state legislators. These legislators are subject to conflicting incentives. On the one hand, state legislators seek to adopt laws that will not harm the state in competing for firms and transactions. On the other hand, to hang onto campaign contributions and other benefits they receive from promoting wealth-redistributing mandatory rules, legislators might resist these competitive forces. The conflict between pro- and anticontractual considerations may result in compromises like those in the choice of law provisions discussed below.

II. THE BENEFITS AND COSTS OF A UNIFORM LAW

The debate over UCITA summarized above largely ignores the bigger

44 See Alan Schwartz & Louis L. Wilde, Intervening In Markets on the Basis of Imperfect Information: A Legal and Economic Analysis, 127 U. PA. L. REV. 630 (1979).

45 Sellers’ reliance on standardized form contracts makes it costly to discriminate between the informed and the uniformed. See id. at 662-63. This consequence is another benefit to consumers of standardized contracting.

46 See Nimmer, supra note 26, at 872-79.

47 Indeed, the opposition caused UCITA’s severance from the UCC upon the ALI’s withdrawl of the project.

48 See infra Part III.

49 See infra Part II.C.

issues concerning the promulgation of a uniform law of computer information transactions. This Part seeks to fill that gap by discussing whether NCCUSL and ALI uniform law proposals promote efficiency.50 We show that, although there may be some benefits to state law uniformity, it is questionable whether the NCCUSL and the ALI should have a role in securing this uniformity. Regardless, given the contentious debate on UCITA and its separation from the UCC, it is unlikely that state law will be uniform. It therefore makes sense to focus on the state competition that will proceed with or without UCITA. We turn to that analysis in Part III.

This Part proceeds as follows. Sections A through C discuss the costs and benefits of widespread state adoption of official uniform law proposals. Section D discusses the potential impact of uniform law proposals by NCCUSL and the ALI. Sections E and F evaluate potential alternatives to the uniform laws process-spontaneous or unguided uniformity, model laws, and federal law. The overall conclusion is that, although there are potential benefits from uniformity, NCCUSL’s efforts are likely to generate more costs than benefits. The real solution, discussed in Part III, is enforcing contractual choice of law.

A. The Case for Uniformity

This Section discusses the benefits of UCITA’s widespread adoption. Many of these arguments should be familiar to advocates of state law uniformity. The function of this discussion is to lay the foundation for discussing alternative methods to obtain these benefits.

1. Transaction and Information Costs

Just like diverse state laws concerning other types of sales, state laws on software sales require buyers and sellers to learn the law that applies to each transaction. These costs involve not only learning different legal rules, but also determining which law applies. As discussed below, the default choice of law rules are vague and indeterminate, and it may be unclear whether the parties can solve this problem by agreement.51 It may also be costly to determine the material facts for choice of law purposes. Buyers may have to determine where their sellers are located, or vice versa. Perhaps worst of all, rather than being able to guide their conduct according to the applicable law, the parties may not know which law applies until the case is litigated.

The UCC covers transactions where the parties seek to handle the transaction quickly without researching the applicable law. It is particularly important to economize on transaction and information costs in these situations and the UCC accomplishes this. It is primarily for this reason

50 This discussion draws on our more general work, Larry E. Ribstein & Bruce H. Kobayashi, An Economic Analysis of Uniform State Laws, 25 J. LEGAL STUD. 131 (1996).

51 See infra Part III.

that we have characterized the Uniform Commercial Code as among the more efficient uniform laws.52

2. Better Legislative Drafting

In addition to immediate benefits for parties to individual transactions, uniform laws can foster efficient legal rules that encourage socially beneficial commerce. First, the uniform law drafters serve as expert consultants to state legislators, particularly for those who serve part-time53 and who may lack expertise both in drafting and in specific substantive areas. This expertise may increase the quality of the resulting laws. Indeed, it is highly unlikely that a part-time inexpert legislature would attempt to deal with a project as complex as UCITA.

Second, whether or not legislators have enough expertise to develop a law like UCITA, they may not have the incentive to do so. A legislator who invests in drafting a law like UCITA may not receive enough reward to compensate for the distraction from the rest of her agenda. Indeed, given the controversy surrounding the law and the risk that the legislative body will not enact it, the legislator may well lose more from the effort than she stands to gain. Also, because the legislator has no property right in the resulting law, other lawmakers may easily appropriate the fruits of a successful project.54

3. Mandatory Rules and Spillovers

States may pass laws authorizing conduct by local residents that impose costs on residents of other states. Conversely, a state may pass a mandatory law, ostensibly to protect its residents, whose reach is broad enough to effectively preclude conduct based primarily in another state. The out-of-state costs of state laws are called “spillovers.” In mass-market licenses, for example, spillovers could be laws either facilitating sales of shrinkwrap licenses in states that prohibit them, or effectively preventing such sales even in states that authorize them. The existence and nature of such spillovers would depend to a significant extent on the type of conduct that triggers a state’s asserting jurisdiction.55 But if the states coordinated around a uniform law, presumably the law would reflect overall costs and benefits in all states and not just individual enacting states.56

The problem, is that legislators who have an incentive to pass legislation that imposes spillovers would, for the same reason, have an incentive

52 See id. at 150.

53 See id. at 140.

54 See Susan Rose-Ackerman, Risk-Taking and Reelection: Does Federalism Promote Innova

tion?, 9 J. LEGAL STUD. 593, 610-11 (1980).

55 See infra text accompanying notes 226-229.

56 The flip-side to this argument, however, is that through uniform laws states can deter exit from ineficient mandatory laws. See infra text accompanying note 105.

not to adopt a uniform law that precluded this spillover.57 Thus, uniform laws are unlikely to effectively deal with the spillover problem.58 The most complete response to any such problem would be a federal law, and not uniform state laws, but federal laws have problems of their own.59

4. Facilitating Development of Case Law and Private Forms

A uniform law can facilitate the development of a repository of judicial decisions and privately developed forms and devices that aid interpretation and application of specific statutory terms. On the critical issues concerning assent to shrinkwrap licenses,60 judicial decisions may clarify issues that UCITA raises. These issues include whether the licensee assented “before or during the party’s initial performance or use of or access to the information,” whether the licensee had an “opportunity to review” the license before the obligation to pay and whether the program had to be installed to enable review of the license to determine reimbursement of the costs of restoring.61 Non-uniform provisions for the terms of assenting to shrinkwrap licenses reduce the interpretive value of precedents on assent.

An example of the effects of non-uniform provisions is UCITA’s adoption of the implied warranty of merchantability for computer software. This implied warranty expressly recognizes that a computer program may be merchantable even if it contains bugs.62 Under this standard, judicial decisions would determine what types of defects are inconsistent with merchantability. These decisions may or may not be relevant under statutes that either do not apply the merchantability standard to software or do not recognize that such a standard tolerates bugs. Indeed, the whole law of computer information transactions may develop very differently under UCITA’s UCC-type approach than under statutes that do not apply analogous standards.

Firms following the uniformity of UCITA may privately develop terms and devices that ensure enforcement of contracts. For example, the Act provides for effectuation of pre-transaction disclosure of licenses where a licensor that “makes its computer information available to a licensee electronically from its Internet or similar electronic site. . . . [and] affords an opportunity to review the terms of a standard form license . . . .”63 This provision applies only if the licensor complies with certain requirements, including: displaying the information “in close proximity to a de

57 See infra text accompanying note 105.

58 See Edward J. Janger, Predicting When the Uniform Law Process Will Fail: Article 9, Capture, and the Race to the Bottom, 83 IOWA L. REV. 569, 577-81, 592 (1998).

59 See infra Part II.F.

60 See supra Part I.A.

61 See UCITA FINAL, supra note 7, § 209(a)-(b).

62 See UCITA APPROVAL, supra note 6, § 403, reporter’s note 3.a (noting that given the complexity of computer programs, “[t]he presence of some defects may be consistent with the merchantability standard”).

63 See UCITA FINAL § 211, supra note 7.

scription of the computer information, or to instructions or steps for acquiring it” or “in a prominent place on the site from which the computer information is offered,” and if the licensor does not affirmatively prevent downloading or copying of the license terms. Uniform adoption of such statutory terms will give licensors the incentive to develop their web sites accordingly. Similarly, uniform adoption of UCITA’s authorization of electronic self-help enforcement of licenses64 may give rise to industry practices in this regard that would not arise, or would arise more tentatively, without such adoption.

Any commonly adopted standard, whether or not proposed by NCCUSL, can achieve these benefits. The question then becomes whether NCCUSL is likely to lead to development of the most efficient standards.65

5. Impact on Preemption

The question discussed here is whether state adoption of UCITA might affect the likelihood of federal law preempting UCITA.66 To the extent that federal law seeks to promote uniformity, uniform state law arguably reduces the need for federal law. If so, this situation may persuade Congress not to adopt preempting federal law in the first instance, and may dissuade courts from interpreting existing laws to preempt state law. Indeed, the uniform law movement originated during the pre-Erie era as a way to help stem the complete federalization of state law.67 Post-Erie, the fear of federal encroachment continues to motivate uniform laws, including the UCC.68

Although states’ promulgation and wide adoption of a uniform law may reduce the need for federal law, it may not reduce Congress’s incentive to legislate. On the one hand, a move to state uniformity may persuade federal legislators that there are political costs to encroaching on state law in this area.69 On the other hand, passage and wide acceptance of a uniform law might demonstrate the existence of interest group support for a law and thereby encourage Congress to act. In other words, the uniform laws

64 See id. § 605. With regard to the value of such enforcement, see Kenneth W. Dam, Self-Help in the Digital Jungle, 28 J. LEGAL STUD. 393, (1999).

65 See infra Part II.E. For an analysis suggesting that the UCC process does not lead to efficient norms, see Robert E. Scott, Rethinking the Uniformity Norm in Commercial Law: Optimal Institutional Design for Regulating Incomplete Contracts, University of Virginia School of Law Working Paper 99-9 (June 1999), Paul B. Stephan, The Futility of Unification and Harmonization in International Commercial Law, University of Virginia School of Law Working Paper 99-10 (June 1999) and compare Jody Kraus & Steven Walt, In Defense of the Incorporation Strategy, University of Virginia School of Law Working Paper Series 99-4 (June 1999).

66 See infra Part II.F. for a discussion of whether federal law does preempt UCITA.

67 See WALTER P. ARMSTRONG, JR., A CENTURY OF SERVICE: A CENTENNIAL HISTORY OF THE NATIONAL CONFERENCE OF COMMISSIONERS ON UNIFORM STATE LAWS 14 (1991).

68 See Patchel, supra note 20, at 93-98 (noting that the Uniform Commercial Code was partly motivated by Congressional efforts to federalize the law of sales).

69 Cf. generally Jonathan R. Macey, Federal Deference to Local Regulators and the Economic Theory of Regulation: Toward a Public-Choice Explanation of Federalism, 76 VA. L. REV. 265 (1990).

process might serve as a stalking horse for federal law. Also, adoption or enactment of the uniform law proposal might spur interest groups that oppose it to redirect their efforts toward Congress.70 And states’ rejection of the uniform proposal might encourage Congress to act by showing that state legislatures have little enthusiasm for acting.

This analysis begs the question of whether preemption would be a good thing. Indeed, it seems odd that NCCUSL and other advocates of uniformity oppose preemption, because federalizing the law can be a more secure route to uniformity than hoping that the states will line up behind a single law.71 Yet federal uniformity may be less efficient than the weaker uniformity that results under state law.

B. Costs of a Uniform Law

Although uniformity of state laws on software sales may have some benefits, uniformity also has potential costs. The basic problem is that NCCUSL is attempting to fashion law in an area where there is little consensus and much uncertainty. Although Part I focused on issues of enforceability of shrinkwrap licenses, there are many other equally difficult issues, such as regulation of self-help mechanisms and the role of electronic agents in contract formation.

Even if there were no policy debate, the UCITA drafters would still have to keep up with fast-moving technology, without knowing how well particular legal rules will function. Consider the problems raised by the evolving concept of defects in computer software

The uniform lawmaking process is ill-suited to deal with such a dynamic legal context. The basic problem is that it proceeds by punctuated equilibrium rather than gradual evolution. When NCCUSL decides its work is done and promulgates a uniform law proposal, the states are then supposed to adopt the law. If evolving circumstances dictate a need for change, NCCUSL may or may not decide to expend its limited resources on making the change. If they do, there is a long delay while the laborious

70 See Garvin, supra note 4, at 350-51 (noting risk that FTC action could preempt UCITA).

71 See infra Part II.F.

72 See Holly K. Towle, The Politics of Licensing Law, 36 HOUS. L. REV. 121, 177 (1999) (discussing Section 2B-118, which creates a new consumer defense to electronic contracts involving an “electronic error,” and noting that, because the rules depend on the existence of means for correcting or avoiding the error is not reasonably provided, they may affect web site design).

uniform lawmaking process is cranked up and states deliberate over the revised act. Because uniformity might decrease during the revision process,73 NCCUSL must balance the need for change against the goal of uniformity.

Even if it is clear that there should be some uniform provisions dealing with software sales, there is an additional question concerning which provisions need to be uniform. For some types of provisions, the costs of stifling continued experimentation and local variation may outweigh the benefits of uniformity. There is, for example, probably a strong case for uniformity in mass-market transactions where there is a high need for standardization and minimizing per-transaction costs. Uniformity makes less sense, however, for transactions that are likely to occur with customized bargaining. NCCUSL’s program of promoting uniform adoption of entire laws frustrates this type of differentiation.74

C. The Politics of Uniformity

Even if uniform state laws on computer information transactions could increase social welfare, it is important to consider the type of uniform proposal likely to emerge from the uniform lawmaking process. The process itself leads to ineffective legislation. Several commentators have discussed the deficiencies of this process.75 As two of these commentators, Alan Schwartz and Robert Scott, have observed, “whether a subject is best regulated by a uniform law or a federal act should be influenced by a perception of how uniform laws are actually made.”76 This study of how the laws are made does not “support the assumption that the [uniform lawmaker] possesses superior legislative capacity to a public legislature-an assumption that has led to the uncritical adoption of UCC provisions by state legislatures.”77 In fact, the very objective of uniformity is likely to reduce the value of these proposals.

UCITA began as a proposed amendment to the UCC. UCC revisions, including the drafting of proposed UCC 2B, are a joint product of NCCUSL and ALI. The institutional differences between NCCUSL and ALI are significant. NCCUSL is a quasi-public body appointed by state governors and funded in part by the states. By contrast, ALI is a private organization whose members choose new members and whose primary purpose is to influence judges through its restatements of the law. UCC’s

73 See Ribstein & Kobayashi, supra note 50, at 188-93 (showing data on states’ adoption of amended uniform law proposals).

74 Thus, where there is preexisting uniformity, as where jurisdictional competition has produced spontaneous uniformity, the promulgation of a uniform act that is not widely adopted can actually reduce uniformity. See id.

75 See Patchel, supra note 20

76 See Schwartz & Scott, supra note 75, at 652.

77 See Scott, supra note 75, at 1822.

permanent editorial board, consisting of ALI and NCCUSL members, meets semi-annually and recommends revisions to the UCC. If ALI and NCCUSL accept the board’s recommendations, ALI appoints a study group of law professors and practitioners. The study group sends its reports to ALI for approval, and to NCCUSL for drafting into statutes.78 After approval by both ALI and NCCUSL, NCCUSL presents the proposed statute to state legislatures just like any other uniform act. Because of its genesis as a UCC project, UCITA reflects the influence of both ALI and NCCUSL. We discuss the potential marginal impact of the ALI and the likely consequences of the ALI’s withdrawal from the project below.

The NCCUSL process seeks to achieve maximum state law uniformity.79 Thus, NCCUSL adopts proposals in as many areas of the law as possible, including those where diversity would be more efficient. To accomplish this objective, NCCUSL does not generally act until the direction of state legislation is clear, by which time the NCCUSL proposal may no longer be useful. Because state governors appoint NCCUSL Commissioners, these Commissioners are likely to have political clout. Although state governors can appoint numerous commissioners, each state has only one vote. This voting system helps ensure that enough states favor the resulting law to give it a good chance to achieve uniformity.

In general, uniform lawmaking reflects the interests and views of three broad groups. First, there are the uniform legislators themselves. At the NCCUSL level these “legislators” are primarily practitioners who, as discussed above, have some political influence, while ALI participants include practitioners, academics and judges. Schwartz and Scott argue that those who are likely to participate in the process are also relatively likely to have a high commitment to maintaining the status quo because of their own preferences, and to preserve their reputations for prudence and responsibility.80 Perhaps most importantly, the participants in the uniformity process have an interest in maximizing the states’ adoption of their proposals, and therefore are likely to craft their proposals to achieve this result.

Second, representatives from various interest groups seek to influence the work of uniform lawmakers. There is an initial question of why interest groups would bother at the uniform lawmaking level rather than seeking to influence state lawmakers directly. As discussed below, state legislators may be especially inclined to adopt a uniform law proposal.81 Also, uniform laws may be more durable than laws adopted other than through the uniform laws process because once state legislators commit to uniformity they may be reluctant to make changes that destroy uniformity. Addition

78 For a review of this process, see Schwartz & Scott, supra note 75, at 600-01.

79 We discuss this process in detail in Ribstein & Kobayashi, supra note 50. Schwartz & Scott discuss a more generalized version of the process that applies to ALI and NCCUSL and their joint work, where proposals emanate from study groups and then proceed to a more general quasi-legislative body. At least with respect to UCITA, the influence of the initial study group proposal is unclear.

80 See Schwartz & Scott, supra note 75, at 611.

81 See infra Part II.D.

ally, uniform lawmakers may be especially receptive to credible interest groups because the lawmakers lack resources for investigating interest group claims.82 Interest groups may therefore get higher net returns on their lobbying efforts at the uniform lawmaking level than they would by working on a state by state basis.

Third, “reformers” may seek to have their views enacted into law through the uniform law process. The term is used here to refer to individuals who advocate their own views rather than those of an interest group. Of course reformers may be allied with interest groups

In short, it is unrealistic to expect the product of the uniform lawmaking process to reflect the “public” interest any more than the product of any other legislature. The outcome instead depends on the interaction of the above groups in the specific context. The context, in turn, depends, first, on whether NCCUSL, ALI or some other law drafting organization is involved and, second, on what type of law is being drafted. With respect to the drafting organization, ALI is generally focused on writing restatements of the law. It follows that ALI will rely on legal academics, who have broad expertise on what the law is, much more heavily than NCCUSL. This distinction is significant because academics are much more likely to be reformers than the practitioners who participate in NCCUSL.

As for NCCUSL, the pursuit of uniformity may increase the influence of cohesive interest groups. NCCUSL will heed groups that can influence enactment in states to maximize the widespread adoptions of its proposed laws.84 NCCUSL will also seek to engineer compromises among competing interest groups that can lead it to adopt convoluted rules85 and to address controversial matters in commentary that has questionable authority.86 The NCCUSL drafting process appears to enhance the power of interest groups because the work is done in lengthy drafting committee meetings held all over the country. These expensive meetings stretch the travel budgets of poorly funded lobbyists.87 NCCUSL’s concern about

82 See id. at 630

83 See Schwartz & Scott, supra note 75, at 610.

84 See Garvin, supra note 4, at 359-60 (also noting the potential impact on drafters of focused interest group criticism of proposed laws)

85 See Ribstein & Kobayashi, supra note 75, at 975-79 (discussing ULLCA dissolution provisions).

86 See Ribstein & Kobayashi, supra note 75, at 979-80.

87 See Garvin, supra note 4, at 353 (noting the difficulties presented by drafting in committee

enactment indicates that NCCUSL proposals are not likely to be very effective in reforming the law, because NCCUSL will be reluctant to adopt proposals that states will shun.88

The critical variables for the law NCCUSL drafts are whether a single interest group is dominant and how informed the participants in the process are about the subject matter. In Schwartz and Scott’s model, uniform laws are more likely to include specific provisions where a single dominant interest group is active and where the uniform lawmakers are relatively well informed. Such specific provisions are analogous to speed limits rather than a duty to drive carefully. Uniform lawmakers are more likely to adopt vague proposals or adhere to the status quo where interest groups compete, or there is no dominant group, and where the uniform lawmakers are relatively uninformed.89 These factors operate in tandem. An uninformed participant’s lack of incentive or resources to become informed increases the influence of a dominant interest group or the participant’s inclination to support the status quo when groups compete. For instance, the dominance of banks and finance companies has resulted in clear rules in UCC Articles 3, 4 and 9 that favor these interests. An example is the emphasis on filings in Article 9.90 On the other hand, interest group competition in the revision of Article 6 on bulk sales led to adherence to the status quo, while the lack of a dominant interest group invited Article 2’s vague “good faith” and “reasonableness” rules.91

UCITA involves a specialized area about which uniform lawmakers may not be well informed. The act also strongly affects several interest groups. It is not clear that software manufacturers are the most dominant group in this area. Moreover, each group may be affected in several different ways. For example, the Society of Information Management (SIM) includes members who are both buyers of software and sellers of other products. They must be concerned that protections applying to software

meetings and recommending drafting by email and computer bulletin board)

88 See Janger, supra note 58, at 585-86. This reluctance bears on whether a NCCUSL-promoted uniform law will be effective to restrain the states from imposing costs on other states. Some argue that states have incentives to adopt uniform laws that restrain their own ability to export costs in order to avoid costs imposed on them by other states. See Larry Kramer, Rethinking Choice of Law, 90 COLUM. L. REV. 277, 314 (1990) (noting states’ incentives to act reciprocally in deciding choice of law issues). However, legislators in many situations may lack incentives to act reciprocally. See Ribstein & Kobayashi, supra note 50, at 140. For example, reciprocity would break down where states can gain from cost-externalizing legislation without incurring costs imposed by other states. As discussed in Part III, we believe that concerns about spillovers in this area have been overblown. The important point for present purposes is that, even if there is a potential for spillovers, uniform laws are likely to be a poor solution.

89 See Schwartz & Scott, supra note 75, at 609

90 See Garvin, supra note 4, at 308, 344 (discussing Article 9)

91 See Schwartz & Scott, supra note 75, at 645.

purchases will not be applied to them as sellers.92 Also, because UCITA involves important issues and has been heavily publicized, it is hard for a single interest group to seize the debate. Under the Schwartz and Scott model, these factors might suggest that UCITA would either adhere to the status quo or adopt vague rules that enable reformers to achieve some of their objectives. Given the fundamental question concerning whether Article 2 applies in this context, evaluating the status quo prediction is difficult because there is no status quo for software sales.93

Perhaps the most idiosyncratic aspect of UCITA is that it is a hybrid product rather than either NCCUSL-only or an ALI/NCCUSL product. Its development reflects ALI involvement, including greater involvement of academic reformers and experts, and greater notoriety than the average NCCUSL product. However, ALI withdrawal means that approval of the final product did not involve the vague compromises that probably would have resulted had UCITA remained a UCC project. These compromises are provoked in a UCC project by reformers relatively unconcerned about enactment and unwilling to defer to experts. One reformer was Professor Lawrence Lessig, who strongly criticized a draft of UCC 2B for reversing the supposedly long-held notion that “contract provisions-especially those in a standard contract-that are surprising to a reasonable person are not binding unless they are brought to the signer’s attention.”94 Also, Professor Harvey Perlman, a member both of NCCUSL and the ALI Council, proposed to bar enforcement as “impermissible,” terms that are “contrary to public policies relating to innovation, competition, and free expression.”95 Although NCCUSL adopted the Perlman proposal in a “sense of the house motion,”96 neither proposal made it into any UCC 2B or UCITA draft.97

The clearest indication of the impact of ALI’s withdrawal is NCCUSL’s rejection of proposals currently being considered in the revision of UCC Article 2. Just prior to NCCUSL’s final approval, a task force of the Article 2 committee and representatives of consumer groups presented a memo at a May 1999 harmonization meeting.98 The memo in

92 The Reporter challenged a SIM representative urging a substantive product quality standard [“regulated standard form” in the report] by asking whether SIM would favor application of such a rule to SIM as sellers and the Chair pointedly suggested discussion of these issues in Article 2 drafting committee meetings. See Report of November 13-15, 1998, Drafting Committee Meeting, (discussing Section 208, which is Section 209 in the UCITA FINAL version), available at <http://www.2bguide.com/nov98rpt.html> (visited Nov. 6, 1999).

93 See Towle, supra note 72, at 133-35.

94 See Lawrence Lessig, Sign It and Weep, THE INDUSTRY STANDARD (Nov. 20, 1998)

<http://www.thestandard.net/articles/article_print/0,1454,2583,00.html>. For criticism of this position see Towle, supra note 72, at 155-56 (quoting Lessig, supra this note).

95 See Charles R. McManis, The Privatization (or “Shrink-Wrapping”) of American Copyright Law, 87 CAL. L. REV. 173, 189 (1999).

96 Id. at 188.

97 UCITA provides only for invalidation of terms violating “a fundamental public policy.” See UCITA FINAL, supra note 7, § 105(a)

98 See Memorandum from Connie Ring and Ray Nimmer to UCITA Drafting Committee on

cluded the following unconscionability provision, which the task force advocating the Article 2 draft wanted the UCITA committee to consider:

(b) In a consumer contract, a nonnegotiated term in a standard form record is unconscionable and is not enforceable if it:

(1) vitiates the essential purpose of the contract

(2) subject to Section 2-202, conflicts with other material terms to which the parties have expressly agreed

(3) imposes a manifestly unreasonable risk or cost on the consumer in the circumstances.

(c) If a court as a matter of law finds that a consumer contract or any term thereof has been induced by unconscionable conduct or that unconscionable conduct has occurred in the collection of a claim arising from a consumer contract, the court may grant appropriate relief.99

Although this language does not yet appear in the Article 2 draft, it is consistent with language already there. Section 2-206(a) currently provides that “[i]n a consumer contract, a court may refuse to enforce a standard term in a record the inclusion of which was materially inconsistent with reasonable commercial standards of fair dealing in contracts of that type . . . .”100 Rather than adopting such strong substantive protections, UCITA relies on traditional notions of unconscionability and consumers’ procedural rights,101 including the right to obtain a full refund before using the product if they object to terms of a shrinkwrap license.102 A drafting committee member commented that the unconscionability standard had, for fifty years, been accepted in sales law without expansion, reflecting uniform lawmakers’ preference for the status quo.103

The forces favoring markets and private ordering were not entirely victorious in the UCITA process. As discussed in more detail below, a critical defeat for private ordering came in the rejection of clear rules for enforcing contractual choice of law and forum.104 Choice of law and forum provisions are important issues, particularly because UCITA’s defense of contract is subject to state consumer protection and other laws that apply to

Possible Revisions of UCITA Draft 6 (July 15, 1999) (discussing the May, 1999 proposed language), available at <http://www.2bguide.com/docs/ucita722m.pdf> (visited Nov. 6, 1999).

99 Id.

100 UCC § 2-206 (Revision, March, 1999 Proposed Draft).

101 See Report of November 13-15, 1998 Drafting Committee Meeting (discussing the shrinkwrap license provision, then § 208, which is Section 209 in the UCITA FINAL, supra note 7), supra note 92. Similarly, the Chair and Reporter of the UCITA Drafting Committee, commenting on the unconscionability language proposed by the UCC task force, noted that “UCITA, of course, has several protections against abuse in standard form contracting that are not present in revised Article 2. These include standards of assent and opportunity to review, as well as the fundamental public policy rule.” See Memorandum from Ring & Nimmer to UCITA Drafting Committee, supra note 98, at 6.

102 See UCITA FINAL, supra note 7, § 209. The ABA proposed this provision. See Report of April 11-13, 1997 Drafting Committee Meeting (discussing ABA’s proposal on Mass Market Licenses, then Section 308, which is Section 209 in the UCITA FINAL, supra note 7), available at <http://www.2bguide.com/apr97mtg.html> (visited Nov. 6, 1999).

103 See Report of November 13-15, 1998 Drafting Committee Meeting (discussing the shrinkwrap license provision, then Section 208, which is Section 209 in the UCITA FINAL, supra note 7), supra note 92.

1 04 See infra Part III.

software licensing.105 The parties might seek to avoid these laws by contracting for application of the law of a more permissive state, and ensuring enforcement of this contract by contracting to have the case tried in the permissive state.106 Uniform lawmakers, seeking to ensure widespread enactment, understandably hesitate to threaten interest group deals and cherished policies of enacting states. More importantly, focusing on individuals’ choices about what law applies is inconsistent with the goal of eliminating that choice and imposing uniformity across the states. In this important respect, the very objective of uniformity impedes NCCUSL’s ability to improve the efficiency of the law.

D. The Impact of Uniform Law Proposals

One might ask whether any problem with uniform law proposals really matters, because the states are always free to disregard the proposals. Indeed, we have shown that the states tend not to adopt NCCUSL proposals for laws where uniformity is likely to have greater costs than benefits.107 Because UCC-type transactions, which include those covered by UCITA, arguably belong in the category of those where uniformity is relatively efficient,108 our data appear to predict that the states will adopt UCITA and thereby gravitate toward efficient uniformity. If uniformity is efficient, NCCUSL-provided uniformity may be less efficient than uniformity achieved through states’ spontaneous movement toward a single type of law. Assuming the states will move toward uniformity only if it is efficient, the important question is whether NCCUSL’s promotion of a uniform law proposal can lead to a less efficient equilibrium than would have resulted without NCCUSL’s intervention.

There are reasons to expect that UCITA will not be very successful. It is significant that the act has lost the imprimatur of being part of the UCC, because that is an important factor in ensuring the widespread adoption of a uniform act. While NCCUSL-only products have not been, on average, widely adopted,109 the adoption by state legislatures of joint ALI/NCCUSL UCC revisions has almost been “pro forma.”110 Moreover, the controversy over UCITA discussed in Part I strongly suggests that state law will not be uniform after promulgation of UCITA. 111

Despite these problems, NCCUSL’s influence alone may cause states to enact UCITA, even if they would not have done so on their own. UCITA may even promote greater uniformity than an evolutionary process

105 See UCITA FINAL, supra note 7, § 105(c) (providing that, “if this [Act] conflicts with a consumer protection statute [or administrative rule], the consumer protection statute [or rule] governs”).

106 See infra Part III.

107 See generally Ribstein & Kobayashi, supra note 50.

108 See id. at 150.

109 Id. at 134.

110 See Patchel, supra note 20 at 136-43. For a comparison of UCC and non-UCC adoption rates, see Ribstein & Kobayashi, supra note 50, at 194-99, tbl. A2.

111 See Garvin, supra note 4, at 351.

would have produced.112 First, NCCUSL can provide camouflage for interest group legislation by presenting it on behalf of the seemingly disinterested and officially accredited group that produced the respected Uniform Commercial Code. Second, a NCCUSL or ALI-backed uniform law could serve as a “focal” provision113 that greases the skids toward uniformity.114 Third, NCCUSL invites advisors from various groups to participate and lobby state legislators to enact their proposals so as to actively promote coordination.

Given the political defects of the NCCUSL process, its influence on state laws is likely to be perverse.115 For example, UCITA may promote widespread adoption of the inefficient limitations on contractual choice of law discussed in Part II.C and in Part III. At the same time, the uniform law process is unlikely to deal effectively with the most important negative consequence of state law diversity-the spillover problem.116 States have no more incentive to refrain from imposing spillovers through uniform laws than they do in the absence of a uniform law proposal.

E. Undirected Uniformity: Spontaneous Uniformity, Standardization and Model Laws

Despite the potentially perverse effects of a NCCUSL proposal, one still may argue that NCCUSL-generated uniformity is better than uniformity that might arise without NCCUSL, or no uniformity at all. The question is whether there are alternative uniformity-producing mechanisms with benefits similar to those for NCCUSL-led uniformity, but with lower costs. This Section considers the feasibility of “undirected” uniformity that arises without the efforts of an agency, such as NCCUSL, dedicated to enactment of uniform laws. Undirected uniformity is a product of state competition and, therefore, is an important byproduct of contractual choice of law.117 This Section also compares the likely efficiency of a spontaneously emerging standard with one that emerges from the NCCUSL process.

Undirected uniformity arises in several ways. First, there is evidence that the states move toward uniformity on their own, without the benefit of NCCUSL or ALI proposals.118 Indeed, our data show that, at least in the absence of a uniform law proposal, states tend to make efficient choices,

112 See Ribstein & Kobayashi, supra note 50, at 146-48 (discussing NCCUSL’s ability to cause the adoption of its proposals).

113 See generally Eric Rasmusen, GAMES AND INFORMATION: AN INTRODUCTION TO GAME

THEORY 36 (1989) (noting that a focal point is an example of a Nash equilibrium in which no player will deviate from cooperation as long as the other players do not deviate)

114 See Lea Brilmayer, CONFLICT OF LAWS: FOUNDATIONS AND FUTURE DIRECTIONS (1991).

115 See supra Part II.C.

116 See supra text accompanying note 58.

117 See infra Part III.

118 See generally Bruce H. Kobayashi & Larry E. Ribstein, Evolution and Spontaneous Uniformity: Evidence from the Evolution of the Limited Liability Company, 34 ECON. INQ. 464 (1996), re

printed in UNCERTAINTY AND ECONOMIC EVOLUTION (John Lott, ed. 1997).

choosing uniformity where uniformity is most beneficial.119 This tendency indicates that relying on the flexibility, collective wisdom, and experience of numerous legislatures, bar drafting committees, and business groups may be superior to NCCUSL-led uniformity. This fact is especially true given NCCUSL’s insistence that the states adopt entire proposals, and therefore accept NCCUSL’s own judgment as to which provisions should be uniform.120

Second, undirected uniformity might arise where an industry or lawyers’ group provides a model law. Private groups can internalize the costs of this effort because their members gain from improving the law or receive reputational benefits by participating in drafting. Although interest groups may be heavily involved-indeed specific industries may promulgate model laws-the result may be positive. States can choose clear, expert-written provisions. To garner general support and adoption, the proponent industry has a strong incentive to propose a balanced law. The expertise of industry specialists might be particularly useful for a law, like UCITA, addressing a technical area.

Third, a state legislature might take the lead in developing a comprehensive statute that other states can copy. Delaware’s efforts in corporate law are a prominent example of this occurrence. Virginia and Connecticut, anticipating the adoption of UCITA, have already taken steps in this direction by proposing laws modeled on then-current versions of the proposed UCC 2B.121 Some states have law revision committees and state bar groups active in law reform.122 The state and its legislators might benefit from this effort by luring firms to relocate.123

Fourth, the benefits of uniformity are achievable even if state laws remain diverse, as long as firms or individuals contract to apply a single law to their transactions. This form of uniformity, or standardization, arises from the action of a single legislature without any coordination among the states. A single state’s adoption of a standard serves the same purpose as uniformity in facilitating development of a network of case law, private forms, and technology. The critical factor is the number of transactions governed by the same law rather than the number of states adopting the law.

Undirected uniformity may be superior to uniformity arising from the NCCUSL process because the resulting standard does not share the political defects of a NCCUSL proposal. As noted, the goal of achieving uniformity results in interest group compromises and reliance on politiciangeneralists rather than informed experts.124 By contrast, the model or other law that ultimately emerges as the standard does not involve the same sort

119 See id.

120 See supra text accompanying note 79.

121 See H.J. 38, 1998 Sess. (Va. 1998) (unenacted)

122 See Garvin, supra note 4, at 357.

123 See discussion infra Part III.

124 See supra Part II.C.

of compromise process

Undirected uniformity may arise more quickly than NCCUSL-led uniformity. Reliance on the uniform law process may actually delay uniformity because the process must wait for a consensus to develop. This delay is exacerbated if other groups are reluctant to step in while deliberations are pending.126 Furthermore, NCCUSL interest group compromises might lead to the refusal to incorporate widely adopted state laws into the uniform law proposal, even though it would produce uniformity. Our data show that an ABA model limited liability company law actually produced greater uniformity over the time span of the study than the equivalent uniform law.127

One might expect the uniform law drafting process to reach results similar to undirected uniformity because uniform law drafters anticipate the effects of state competition to maximize state adoptions of their proposal.128 Accordingly, uniform lawmakers have some incentive to enact existing model or state laws that they expect will become widely adopted. This incentive does not produce the same effects as state competition, however, because one cannot assume all states want to participate actively in competition to attract firms and transactions. States’ legislative priorities may reflect a desire, at least at the outset,129 to resist competition and to favor strong interest groups within the state. Moreover, the conflicting groups that participate in the NCCUSL process might cause adoption of proposals that would not maximize uniformity. 130

The important question is whether there are potential problems in unguided uniformity that might justify the uniform law process despite its defects. One potential problem is that the states might tend to adopt the wrong standard. This problem can occur because of imperfect information about the correct standard or because states tend to disregard their own information due to some sort of “herd” behavior.131 It is not clear, how

125 See id.

126 In particular, the ABA has long had a close relationship with both NCCUSL and ALI

127 See Larry E. Ribstein & Bruce H. Kobayashi, supra note 67 (showing that, in the Uniform Limited Liability Company Act, NCCUSL failed to mimic widely adopted state laws even though doing so would maximize the likelihood of uniformity).

128 See Janger, supra note 58, at 591-92.

129 Over time, state competition may break down mandatory rules. See Bruce H. Kobayashi &

Larry E. Ribstein, Contract and Jurisdictional Freedom, in FALL AND RISE OF FREEDOM OF CONTRACT (F.H. Buckley, ed., forthcoming 1999).

130 See supra note 127 and accompanying text.

131 For general discussions of herd behavior as applied to investments, see Abhijit V. Banerjee, A Simple Model of Herd Behavior, 107 Q.J. ECON. 797 (1992)

ever, why state legislators would flock to a standard they believed was inferior, or why they would be less, rather than more, likely to do so if uniform lawmakers proposed the standard. In any event, our data contradict any herding impulse by state legislators, at least with the law of unincorporated business associations.132

Another potential impediment to developing efficient undirected uniformity are “network externalities.” Network externalities occur when a standard, such as a software operating system, can generate a network of benefits that increase the value of the “networked” product. With software operating systems, the network of benefits includes programs written for the system. For computer information transactions law, the network may consist of case law, standard forms, business practices and technologies like those discussed in Part II.A. Even if a market would be better off moving to a new “network,” this movement does not necessarily occur.133 Even though society gains from a user’s move to a new network, the new user incurs the entire cost, including losing the benefits of the old network. Thus, users may not move to the new network and the old network may be “locked in.” Commentators have argued that there are similar effects with contracts and statutes.134 For example, Klausner suggests that the long dominance of Delaware law might be due to lock-in effects rather than the superiority of the Delaware regime.135

The network externalities theory is highly speculative. Commentators criticize the theory and data refute it.136 One reason to doubt the impor

1012-13 (1992)

132 See Kobayashi & Ribstein, supra note 118.

133 For general discussions of network externalities and lock-in see Joseph Farrell & Garth Saloner, Standardization, Compatibility, and Innovation, 16 RAND J. ECON. 70, 71-72 (1985) (characterizing the problem as one of excess inertia)

134 See Marcel Kahan & Michael Klausner, Standardization and Innovation in Corporate Contracting (Or, “The Economics of Boilerplate”), 83 VA. L. REV. 713 (1997)

135 See Klausner, supra note 134.

136 For criticisms of the theory as applied to products and services see S. J. Liebowitz & Stephen E. Margolis, The Fable of the Keys, 33 J.L. & ECON. 1 (1990). See also S. J. Liebowitz & Stephen E.

Margolis, WINNERS, LOSERS, AND MICROSOFT: COMPETITION AND ANTITRUST IN HIGH

TECHNOLOGY (1999) (showing evidence that Microsoft’s victory in software markets is due to the superiority of their products rather than network externalities)

tance of externalities and lock-in is that entrepreneurs can play a role in developing new networks in legal and other products and in motivating users to move to the new networks.

Even if the network externalities are viable, its implications for uniform laws in general, and UCITA in particular, are ambiguous. Network externalities might theoretically prevent a move to a more efficient standard, unless the move is promoted by NCCUSL. Therefore, despite the defects of the NCCUSL process, that process might lead to a more efficient equilibrium than would arise without NCCUSL.137 Notwithstanding the possibility of increased efficiency, network externalities also might operate to lock in an inefficient NCCUSL standard because the costs of abandoning the existing network and moving to a new one are prohibitive.138 In other words, the network externalities theory says nothing about the relative efficiency of NCCUSL/ALI or spontaneously generated standards, but rather only deals with the extra costs from locking in a standard arising under either regime. This approach is reminiscent of the debate over keyboards and videotape in the network externalities literature. While it has been asserted that the QWERTY typewriter keyboard and the VHS videotape formats were inferior to the alternatives of Dvorak and Betamax,139 there is also convincing evidence that the market did not get it wrong in either case.140

In short, even if state law uniformity has benefits, we need not rely on NCCUSL, with all of its political defects, to produce these benefits.

F. Federal Law and Preemption

A final means of achieving uniformity in this area is through federal law, either under existing intellectual property laws or under new federal laws. This Section first discusses the debate on whether federal law, in fact, preempts UCITA. Our analysis shows that this result is unclear. Then it shows that any debate on this score should be settled against preemption. Accordingly, Congress should not enact legislation explicitly preempting state contract law on software sales.

Some commentators have argued that the federal copyright law may preempt state law enforcement under UCITA. This preemption would include restrictive contract terms like those in the ProCD license forbidding commercial use.141 This analysis is obviously important in understanding

with authors).

137 See supra Part II.C.

138 See Garvin, supra note 4, at 349-50 (noting that if, through uniformity, “the law is frozen incorrectly, it may alter greatly the continued development of an important industry”). See also Michael Froomkin, Article 2B as Legal Software for Electronic Contracting-Operating System or Trojan Horse?, 13 BERKELEY TECH. L.J. 1023, 1031 (1998) (concluding that UCC 2B is technologically premature in its standards for electronic contracting).

139 See supra commentary note 133.

140 See Liebowitz & Margolis, supra note 136.

141 See Dennis S. Karjala, Federal Preemption of Shrinkwrap and On-Line Licenses, 22 U. DAYTON L. REV. 512 (1997)

the extent to which UCITA would change the law.

The Copyright Act contains what is known as a “field preemption” provision. Section 301(a) of this Act reads:

[A]ll legal or equitable rights that are equivalent to any of the exclusive rights within the general scope of copyright . . . are governed exclusively by this title. Therefore, no person is entitled to any such right or equivalent right in any such work under the common law or statute of any State.142

Courts have held that contracts are not equivalent to rights within the scope of the Copyright Act under the “extra element” test, and therefore are generally not subject to preemption under section 301(a).143 The extra element test holds that a state law is not preempted if an extra element is required instead of, or in addition to, the acts of reproduction, performance, distribution, or display for a state-created cause of action.144 Such an element includes the parties’ agreement.145 Conversely, absent an enforceable contract, a state law that imposes terms on parties is subject to preemption.146 As Judge Easterbrook said in ProCD:

Rights “equivalent to any of the exclusive rights within the general scope of copyright” are rights established by law-rights that restrict the options of persons who are strangers to the author . . . . A copyright is a right against the world. Contracts, by contrast, generally affect only their parties

The general terms of the Supremacy Clause can also preempt contract

Contract into Expand, 87 CAL. L. REV. 19 (1999). Conversely, some argue that state law enforceability of waivers of owners’ fair use rights may lead to copyright protections without the limitations imposed by federal law. See McManis, supra note 95.

142 17 U.S.C. § 301(a) (1994)

143 See ProCD, 86 F.3d at 1450

144 See Lemley, Beyond Preemption, supra note 25, at 139-40. However, application of UCITA would likely result in the existence of an enforceable contract. The extra element test arguably allows courts to “find” the extra element needed for the state law to survive preemption, and allows legislatures to manipulate preemption analysis by adding extra elements to state laws. See Patrick McNamara, Note, Copyright Preemption: Efecting the Analysis Prescribed by Section 301, 24 B.C. L. REV. 963, 984-85 (1983). In the context of UCITA, the drafters’ adoption of Judge Easterbrook’s analysis of Section 301 might be regarded as an attempt to settle the preemption issue in favor of state law.

145 See Nimmer, supra note 38, at 862-63

146 See Vault v. Quaid Software Ltd., 655 F. Supp. 750 (E.D. La. 1987), af’d, 847 F.2d 255 (5th Cir. 1988). Some commentators have suggested that the Fifth Circuit’s holding in Vault, in which federal law preempted a Louisiana statute that enforced terms contained in shrink-wrap licenses, sets an ominous precedent for the preemption of contracts under UCITA. See id.

147 See ProCD, 86 F.3d at 1454.

terms even if they are not within the Copyright Act’s “field preemption.” This preemption occurs when the contract terms directly conflict with the law, or impede accomplishment of Congress’s objectives. The Supreme Court used this general preemption theory during the 1960s to invalidate state unfair competition laws. The Court held that the state laws interfered with federal laws by purporting to protect ideas left unprotected by federal patent law.148 The Court has also invalidated the state common law doctrine of licensee estoppel149 and a Florida statute that prevented copying of boat hulls by means of any “direct molding process.”150 However, the Court has not consistently extended these holdings to other areas. It upheld a state law protecting sound recordings that were not then covered by copyright law,151 and allowed a state trade secrets law to co-exist with federal patent law.152

For contract terms, a general principle is hard to discern. For example, the Court preempted contract terms that “extended” a patent’s term by structuring payments beyond the patent’s expiration date,153 but enforced a royalty contract for use of an unpatented design for a key ring.154 One might approach these cases by considering the costs imposed by enforcing the contract in terms of federal law.155 Licensing terms that limit use, such as the one litigated in ProCD, can be seen as attempts to opt-out of the first-sale doctrine.156 However, Judge Easterbrook noted in ProCD that terms restricting or limiting uses or resale, coupled with lower prices, allow licensors to price discriminate between low-value and high-value users.157 Without restrictions on use or resale, copies licensed at the low

148 See Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225 (1964)

149 See Lear, Inc. v. Adkins, 395 U.S. 653 (1969).

150 See Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141 (1989). For criticism of the Court’s approach in these cases, see Douglas G. Lichtman, The Economics of Innovation: Protecting Unpatentable Goods, 81 MINN. L. REV. 693 (1997)

151 See Goldstein v. California, 412 U.S. 546 (1973).

152 See Kewanee Oil v. Bicron Corp., 416 U.S. 470 (1974).

153 Brulotte v. Thys Co., 379 U.S. 29 (1964). Although the contract in Brulotte did not technically “extend” the term of the patent, the Court held that the contract could not do so. See id. at 32

154 Aronson v. Quick Point Pencil Co., 440 U.S. 257 (1979).

155 See David McGowan, Free Contracting, Fair Competition, and Article 2B: Some Reflections on Federal Competition Policy, Information Transactions, and “Aggressive Neutrality,” 13 BERKELEY TECH. L.J. 1173 (1998) (examining the intersection of contract, intellectual property and antitrust law)

156 17 U. S.C. § 109 (1994). The committee report to Section 109(a) of the Copyright Act of 1976 indicates that Congress anticipated that parties might contract out of a first-sale right. See Lemley, Intellectual Property and Shrinkwrap, supra note 25, at 1273 n.156 (citing Notes of Committee on the Judiciary, H.R. Rep. No. 94-1476, at 79 (1976), reprinted in 1976 U.S.C.C.A.N. 5659, ). To the extent that the committee report indicates that Congress intended the first sale doctrine to operate as a default rule rather than as a mandatory rule, license terms that provide for a de facto waiver of the first sale doctrine, such as the one at issue in ProCD, would not directly conflict with federal copyright law, nor would it impede Congress’s objectives.

157 See ProCD, 86 F. 3d at 1449-50. For an analysis of information that is shared by multiple users, see Yannis Bakos, et al., Shared Information Goods, 42 J.L. & ECON. 117 (1999)

value price will flow to high-value users, forcing sellers to employ costly technological fixes or increase prices charged, or even abandon, low-value users.158 One commentator has argued that contracts that facilitate price discrimination or otherwise increase the licensor’s profit should be discouraged as a matter of public policy and perhaps enforcement of such contracts under state law should be preempted because they alter the “delicate balance” between producers and users that federal copyright laws achieve.159 Indeed, successful third-degree price discrimination always increases the licensor’s profit.160 But a cost-benefit analysis must move beyond the static view and consider the dynamic benefit from increasing the probability that a seller will produce a costly-to-compile database.161 Because copyright law does not protect databases like the one involved in ProCD,162 database compilers’ return on their investment, and therefore willingness to engage in this effort, depends on whether they can contractually limit use.163

A related objection to allowing contractual protection of databases is that such protection conflicts with the federal copyright laws. It allows contracting parties to protect works that should be in the public domain. The contract does not take the white page listings used by ProCD out of the public domain, however, because a competitor is free to independently compile another database. In economic terms, because ProCD has no right to prevent others using the public domain listings in their database, allowing contractual protection of databases does not increase the cost of expression to other potential ProCD competitors.164

Contracts may complement rather than interfere with the intellectual

Besen & Sheila N. Kirby, Private Copying, Appropriability, and Optimal Copying Royalties, 32 J.L. & ECON. 255 (1989)

158 See ProCD, 86 F. 3d. at 1450. But see Michael J. Meurer, Price Discrimination, Personal Use and Piracy: Copyright Protection of Digital Works, 45 BUFF. L. REV. 845 (1997) (arguing that facilitating price discrimination is not necessary given the existences of superior substitutes such as digital self-help). For an economic analysis of self-help, see Kenneth W. Dam, Self Help in the Digital Jungle, 28 J. LEGAL STUD. 393 (1999). For an analysis criticizing broad self-help rights, see Julie Cohen, Copyright and the Jurisprudence of Self-Help, 13 BERKELEY TECH. L.J. 1089 (1998).

159 See Ramona L. Paetzold, Contracts Enlarging a Copyright Owner ‘s Rights: A Framework for Determining Unenforceability, 68 NEB. L. REV. 816 (1989). See also Lemley, Beyond Preemption supra note 25

160 See Meurer, supra note 158.

161 See Michelle M. Burtis & Bruce H. Kobayashi, Intellectual Property and Limitations on Contract: Antitrust, Preemption, and the First Sale Doctrine, in COMPETITION IN DYNAMIC ECONOMIES, (J. Ellig ed., forthcoming 2000). With regard to the use-creation tradeoff, see generally Frank H. Easterbrook, Insider Trading, Secret Agents, Evidentiary Privileges, and the Production of Information, 1981 SUP. CT. REV. 309 (1982)

162 See Feist Publications, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340 (1991) (rejecting the “sweatof-the-brow” theory of copyright).

163 For a more complete discussion of intellectual property protection for databases, see Jane C. Ginsburg, Copyright, Common Law, and Sui Generis Protection of Databases in the United States and Abroad, 66 U. CIN. L. REV. 151 (1997).

164 See Landes & Posner, An Economic Analysis of Copyright Law, 18 J. LEGAL STUD. 325, 332 (1989) (noting increases in the cost of expression a reason for limitations on copyright).

property laws.165 Congress’s failure to protect certain types of information may simply indicate that technology has outstripped its foresight rather than reflect a conscious decision to allow such information to fall into the public domain.166 Any policy that leaves everything to Congress’s limited foresight might create a centralized, uninformed and inflexible system that provides inadequate incentives for the creation of intellectual property.167 This type of system does not serve the underlying utilitarian goals of the intellectual property laws, to “promote the progress of the sciences and useful arts.”168

Thus, whether current federal law preempts UCITA is far from clear. The question then arises whether Congress should enact a law explicitly regulating this area and preempting inconsistent state law. The argument for federalization focuses on the potential for spillovers under state law.169 For example, with software sales, states that want to attract large software companies have an incentive to avoid strong protections for software buyers. While the state would reap all of the benefits of a lax law, buyers in other states might incur some of the costs if they cannot establish sufficient contacts to get jurisdiction over remote sellers.170 More plausibly, given the ease of establishing jurisdiction, states may seek to impose burdens on remote software companies for the benefit of local consumers, lawyers, and corporate software users.

Federal law is potentially superior to state law uniformity in preventing spillovers. This superiority results because gaps in state uniformity can permit variations that lead to spillovers. Not only might some states fail to adopt the uniform law proposal or adopt a modified version, but the uniform law itself will not necessarily cover the field. For example, UCITA provides that it is subject to mandatory laws of adopting states.171 These may include consumer protection laws that adversely affect sellers in other states. Conversely, some states may have weak, or no, mandatory laws and so these laws would favor local software sellers to the possible detriment of their remote buyers. It has thus been argued that federal law is the ap

165 See Gomulkiewicz, supra note 41 (discussing the diversity and innovation software licenses enable)

166 This conclusion is indicated by Congress’s recent attempt to fill the gap regarding protection of databases involved in cases such as ProCD. See Harvey Berkman, Congress Tackles Database Law,

NAT’L L.J., July 22, 1999, at B1.

167 See Lichtman, supra note 150, at 694-95

168 U.S. CONST., art. I, § 8, cl. 8.

169 For discussions of spillovers as a justification for federal law, see Frank H. Easterbrook, Federalism and European Business Law, 14 INT’L REV. L. & ECON. 125 (1994)

170 See infra discussion of jurisdiction in text accompanying notes 222-229.

171 See UCITA FINAL, supra note 7, § 105(c).

propriate antidote to a state law race to the bottom.172

The main argument against federal law is based on precisely those costs of uniformity that cast doubt on the efficiency of state law uniformity including reduced variation, competition, and experimentation.173 In other words, state laws may be better because they are less uniform. Most importantly, the opportunity for exit that exists under state law may provide a greater assurance of an efficient law than imposing federal law everywhere. For example, federal law may impose strong consumer protections that reduce value-increasing transactions or redistribute wealth between sellers and buyers. To evaluate the efficiency of federal law in preventing spillovers, it is necessary to know whether federal law would prevent a state law race to the bottom or frustrate a race to the top. With respect to UCITA, this question turns in part on the debate discussed earlier over whether shrinkwrap licenses harm consumers.174 If, as we argue, market forces adequately protect consumers, federal mandatory laws may be inefficient. Even if uniformity is potentially efficient, the sort of uniformity that would result under federal law may not be. Congress probably would not enact a comprehensive law of computer information transactions. Still, if Congress had the constitutional power to enact such a law, the law would be politically infeasible.175 The result might be an incoherent patchwork of state and federal law.176

Finally, even if federal law theoretically might provide a more efficient brand of uniformity than what would result under state law, it is far from clear that Congressional legislation will be better than the NCCUSL/ALI product. Federal law-making does not avoid politics, but simply involves a different sort of politics than arises under the uniform laws process. For example, because federal legislators work on a variety of bills over time, they can engage in logrolling, or vote-trading.177 The possibility of trades means that strong language in one bill can be traded for strong language in another. While this option may reduce vagueness at the federal level it also may facilitate even more inefficient wealth redistributions than state uniform laws accomplish. Also, in contrast to NCCUSL, federal legislation does not involve the sort of private, geographically dispersed meetings that favor industry groups.178 Therefore, consumer groups may have more power at the federal level.

172 See Janger, supra note 58, at 628.

173 See supra Part II.B.

174 See supra Part II.A.

175 See Macey, supra note 69.

176 For discussions of analogous problems integrating federal bankruptcy law and the state law of business associations, see Larry E. Ribstein, Partner Bankruptcy and the Federalization of Partnership Law, 33 WAKE FOREST L. REV. 795 (1998)

177 Cf. Schwartz & Scott, supra note 75, at 613 (discussing the effect of the absence of logrolling in drafting uniform laws). By contrast, NCCUSL drafters work on one law at a time over several years, which makes logrolling difficult or impossible.

178 See supra text accompanying note 87.

III. A BETTER WAY: CONTRACTUAL CHOICE OF LAW

The above discussion shows that there are significant objections both to the general concept of uniformity and to achieving such uniformity through a NCCUSL/ALI-promoted law. Federally provided uniformity of part of the law may lead to an even less efficient outcome. A large number of states will likely reject UCITA anyway. States often adopt modified versions of uniform acts or to ignore uniform acts altogether,179 and the loss of a connection with the UCC decreases the prospects for widespread adoption.180 Even a uniformly adopted state law or comprehensive federal law would be subject to myriad judicial interpretations by state or federal courts.

This result does not, however, have to be the end of prospects for standardization of the law governing software sale or of significant law reform in this area. As long as at least one state adopts a UCITA-type law -perhaps a state seeking to attract software manufacturers-software sellers and their customers can contract for application of this law.181 The law might thereby become a standard even if only one state adopts it, just as Delaware law serves as a standard for corporate law. Accordingly, contractual choice of law might solve many of the problems that uniform laws address. At the same time it might do so without the attendant costs. As long as the states do not seek to conform to a uniform standard, the contracting parties can decide what law applies and are not held hostage to a perverse political process.

The problem with this scenario is that states may seek to block evasion of their mandatory laws by limiting the enforcement of contractual choice of law. Indeed, UCITA itself follows this approach. This Part shows that restrictive rules on contractual choice are not justified on efficiency grounds. Moreover, it shows that legal restrictions on enforcement of contractual choice of law need not preclude active jurisdictional competition and eventual erosion of rules restricting contractual choice.

Section A outlines the advantages of enforcing contractual choice of law, while Section B discusses some of the arguments against broad enforcement of choice of law contracts.182 Section C evaluates the UCITA choice of law provisions in light of this analysis. It shows that UCITA has qualified enforcement of contractual choice. One expects this approach from a statute whose main objective is for the law, and not contracts, to provide uniformity. But Section D shows that, despite lawmakers’ best efforts, a regime broadly permitting contractual choice of law will nevertheless develop.

179 See Ribstein & Kobayashi, supra note 50.

180 See supra text accompanying note 110.

181 See Garvin, supra note 4, at 351.

182 For a more extended discussion of costs and benefits of enforcing contractual choice and a general review of the state and constitutional law see Larry E. Ribstein, Choosing Law By Contract, 18

J. CORP. L. 245 (1993).

A. Benefits of Enforcing Contractual Choice of Law

Choice of law clauses can solve many of the problems that serve as arguments in favor of uniformity.183 First, contractually clarifying choice of law reduces the transaction costs of determining the applicable law at the time of the contract and at the time of trial.184 Although it may be costly for customers to sort out choice of law when confronted with form contracts where the sellers essentially designate the law, customers are no worse off than if choice of law were left to confusing common law rules. Also, these costs are eliminated if the firms and individuals move to a common standard.185 The parties have a strong incentive to do so if there are benefits from standardization. One benefit is that adoption of a common standard serves the same purpose as uniformity in facilitating development of a network of case law, private forms and technology. The critical factor is the number of transactions governed by the same law rather than the number of states adopting the law.

Moreover, contractual choice of law solves the spillover problem associated with perverse mandatory laws.186 The freedom to contractually choose the applicable law allows parties to escape inefficient mandatory state laws without having to completely avoid contacts with the state.187 This flexibility reduces interest groups’ incentive to lobby for these laws and legislators’ incentives to pass them.

Uniform laws may be a second-best solution to the problems that diverse state laws create. In other words, state law uniformity may be efficient only to the extent that contractual choice of law is not fully enforced. This result occurs because, where contracts are practicable, enforcing choice of law clauses would be a better way to reduce the costs of state law diversity than a uniform law proposal. Contractual choice of law allows parties to determine which law suits their transactions best, rather than being forced to accept a one-size-fits-all statute. Also, jurisdictional choice provides a mechanism for “voting” on various legal approaches. Rather than being forced to accept a flawed uniform act produced by a centralized process, the parties can choose the best statute from the collective wisdom of numerous bar drafting groups and legislatures.

Finally, contractual choice of law provides better predictability concerning the applicable law. A uniform law regime, at best, ensures that a particular statute will apply to all transactions in a given category. Even that result assumes that all states do, in fact, adopt the uniform law without variation, which is unlikely given the controversy over UCITA. But even with complete state law uniformity, there would still be variations in the states’ judicial interpretations. More importantly, mandatory laws outside

183 See supra Part II.A.

184 See supra Part I.A.1.

185 See supra Part II.E.

186 See supra Part II.A.3. Contractual choice of law potentially involves spillover costs associated with evasion of mandatory rules. This problem is discussed in Part III.B.

187 But, as discussed below, this option arises if contractual choice is not enforced.

the scope of the uniform statute would vary from state to state.188 In contrast, by contracting for the law of a particular state, the parties choose the entire law as well as the judicial decisions of that state.

B. Objections to Enforcing Contractual Choice

The objections to enforcing contractual choice of law are based fundamentally on the potential for a “race to the bottom,” where states’ competition to have their laws designated in contracts produces inefficient laws and evasion of benign internal rules.189 The race-to-the-bottom argument raises two questions. First, why would parties contract for application of inefficient laws? Second, why would legislators adopt such laws in the first place?

The arguments for inefficient contracting for the applicable law echo those relating to contracting generally, including those relating to enforceability of shrinkwrap licenses discussed above.190 One potential problem is the presence of informational asymmetries in choice of law clauses in standard form contracts. With respect to software sales, the licensor responsible for drafting the term is arguably better informed about the effect of choosing the applicable law. The primary effect of a choice of law selection clause, however, will likely be to enforce literally the terms of the contract. Thus, a party arguing surprise would have to argue that she believed she could resist enforcement under the law of a jurisdiction different that the one agreed upon in the contract. This resistance is plausible only for the small category of clauses that are simultaneously regarded in most states as being against public policy and that are not banned under federal law. An example is waivers of a licensee’s rights to use copyrighted material. Moreover, the best way to eliminate surprise and clarify expectations would be a strong legal rule favoring enforceability.

Contracting for choice of law also theoretically may be inefficient because of third-party effects. One could argue, for example, that enforcing choice of a state law that limits use of information has external effects on other users by allowing contracting parties to keep private information that would otherwise fall into the public domain. As discussed above however,191 this analysis is incomplete because it ignores the importance of the offsetting third-party effect of providing adequate incentives to produce the information in the first place. The real question is whether the contractually selected law that would enforce the contract is less likely to get this balance wrong than the state whose mandatory law is evaded.

Whether the race is to the top or the bottom may depend partly on

188 SEE UCITA FINAL, supra note 7, § 105(c) (providing that UCITA is subject to contrary mandatory laws).

189 For an analysis of the race to the bottom theory in relation to uniform laws see Janger, supra note 58, at 578-82, 588-92.

190 See supra Part I.

191 See supra Part II.F.

whether the contractually selected state imposes spillovers. Specifically, the contractually chosen state may have excessively lax laws because it has inadequate incentives to consider out-of-state harms. For example, Delaware might be said to have won a race to the bottom in corporate law192 because out-of-state shareholders bear the costs of any inefficient corporate laws. One way to deal with this problem is to enforce choice of law clauses only when the parties choose the law of a state where they have significant contacts. Another way is to place substantive limits on contractual selection of the law of an otherwise unconnected state, such as refusing to enforce the clause where it contravenes “fundamental” policies of the connected state.

There are several reasons to be skeptical of the spillover argument and the constraints on contracting that appear to follow from it. First, it may not be clear a priori whether enforcing contractual choice of law creates or avoids spillovers. One of the benefits of contractual choice is that it permits the parties to avoid inefficient laws.193 These laws may themselves be attributable to spillover effects. For example, laws that are liberal to licensees may have just as great an effect on incentives to produce information as laws that permit strong licenses have on the dissemination of existing information.

Second, contractual choice of law does not allow parties to make a contract that is condemned by all jurisdictions

Third, non-enforcement of contractual choice of law is not necessarily the last line of defense against spillovers. As discussed above,196 federal

192 See generally William L. Cary, Federalism and Corporate Law: Reflections Upon Delaware, 83 YALE L.J. 663, 665-66 (1974).

193 See supra Part III.A.

194 See supra Part II.C.

195 See generally Gary Becker, A Theory of Competition Among Pressure Groups for Political Influence, 98 Q.J. ECON. 371 (1983).

196 See supra note 169 and accompanying text.

law is available for this purpose. The parties must not only choose the law of some state, but also a law that federal law does not preempt. Interest groups frustrated at the state level can always seek help in Congress or in an administrative agency such as the FTC. Indeed, the threat of federal preemption might deter states from going too far in enacting permissive laws or authorizing evasion of such laws by contractual choice of law.197 It seems unlikely that a very large category of overly permissive laws would survive scrutiny at both levels.

Fourth, even if an objectionable contract manages to survive regulation at both the federal and state level, choice of law clauses are subject to discipline in product markets. For example, shares in publicly-held Delaware firms are priced in efficient securities markets and must therefore be made attractive to investors to minimize firms’ cost of capital.198

Finally, any benefits of refusing to enforce contractual choice of law must be weighed against the costs-i.e., the benefits of enforcement. Barriers to contractual choice of law impede the efficiency-promoting aspects of jurisdictional choice. The more difficult it is for contracting parties to exit wealth-redistributing laws, the greater is the legislators’ ability to enact these laws.

C. Contractual Choice of Law and UCITA

Extending the principle of freedom of contract to contractual choice of law seems consistent with the general principles underlying UCITA. Many different interests wanted this approach. These interests included not only big manufacturers seeking to standardize license terms favoring contractual choice of law in UCC 2B, but also small software producers, publishers and film companies concerned about being forced to litigate in non-U.S. courts under non-U.S. law hostile to their interests.199 Nevertheless, there are reasons to predict that UCITA would limit contractual choice of law. First, individual parties’ choice of law is fundamentally inconsistent with NCCUSL’s goal of having its proposals adopted verbatim in every jurisdiction. Second, contractual choice of law encourages states to compete for contract business with their own laws, and therefore can be seen as an impediment to widespread adoption of NCCUSL’s proposals. Third, groups other than those from industry, particularly consumer groups, also had influence. Thus, as recently as a February, 1997 meeting, the Drafting Committee adopted a broad contractual choice provision, eschewing an exception that would prevent consumers from receiving benefits of a law

197 With respect to the effect of state adoption of UCC § 2B or UCITA on the possibility of preemption, see supra Part I.A.5.

198 See Ralph Winter, GOVERNMENT AND THE CORPORATION (1978). For evidence that corporate

laws respond to this discipline, see Peter Dodd & Richard Leftwich, The Market for Corporate Charters: “Unhealthy Competition” Versus Federal Regulation, 53 J. BUS. 259 (1980).

199 See Report of February, 1997 Drafting Committee meeting. See also Scott, supra note 75, at 1826-29 (discussing adoption of choice of law provision in UCC Article 9 because of support by dominant interest group).

that would have applied absent the choice of law clause.200 In the end, consumers were able to qualify enforcement of contractual choice of law. UCITA provides as follows regarding choice of law:

The parties in their agreement may choose the applicable law. However, the choice is not enforceable in a consumer contract to the extent it would vary a rule that may not be varied by agreement under the law of the jurisdiction whose law would apply under subsections (b) and (c) in the absence of the agreement.201

In the absence of an enforceable choice of law term, the following rules apply:

An access contract or contract providing for electronic delivery of a copy is governed by the law of the jurisdiction in which the licensor is located when the agreement is made.

A consumer contract that requires delivery of a copy on a physical medium is governed by the law of the jurisdiction in which the copy is or should have been delivered to the consumer.

In all other cases, the contract is governed by the law of the jurisdiction having the most significant relationship to the transaction. 202

For purposes of this section, a party is located at its place of business if it has one place of business, at its chief executive office if it has more than one place of business, or at its place of incorporation or primary registration if it does not have a physical place of business. Otherwise, a party is located at its primary residence. 203

To the extent that this section allows parties to contractually choose the applicable law in commercial contracts, the UCITA would implement fewer limitations on contractual choice of law than would application of general choice of law rules. Among other things, it would eliminate the requirement that the chosen jurisdiction have a “reasonable relationship” to the transaction.204 But choice of law is not enforceable in “consumer

200 See Report of February 31-23, 1997 Drafting Committee Meeting (discussing choice of forum options), available at <http://www.2bguide.com/feb97rpt.html> (visited Nov. 6, 1999).

201 UCITA FINAL, supra note 7, § 109(a).

202 Id. § 109(b)(1)-(b)(3).

203 See id. § 109(d).

204 Enforcing such contracts rests on the basic conflicts rule except in the “absence of effective choice” by the parties. This basic rule is subject to two major limitations. A contractual choice of law will not be enforced as to issues of validity if:

(a) the chosen state has no substantial relationship to the parties or the transaction and there is no other reasonable basis for the parties’ choice

RESTATEMENT (SECOND) CONFLICTS, § 187 (1971). The UCC provides for enforcement of contractual

choice of law unless the contract has no “reasonable relation” with the chosen law. See U.C.C. § 1-105(1). For a critique of these limitations, see Ribstein, supra note 182, at 261-67. For a comparison between UCITA and the general law on contractual choice, see UCITA APPROVAL, supra note 6, § 109, reporter’s note 3.

contracts” if it would vary a mandatory rule of a jurisdiction whose law would otherwise apply under section 109(b). This rule requires application of mandatory laws of the licensor’s state if the delivery is electronic, and mandatory laws of the licensee’s state when a physical copy is involved. UCITA’s consumer rule is therefore actually more restrictive than the general rule on contractual choice of law. The general rule allows contracting around a mandatory rule applied under default choice of law rules unless there is no minimum relationship with the chosen state or the chosen law would contravene a “fundamental policy” of the default choice of law state and that state has a “materially greater interest” in the issue than does the chosen state.205

UCITA also might restrict contractual choice of law under section 105(b), which allows a court to refuse to enforce any contract term if application of the term would violate a “fundamental public policy.”206 It is not clear whether this provision applies only to enforcing contract terms under UCITA itself, or whether it applies to enforcing a contractual choice of law clause that chooses a state law that violates “fundamental public policy.” If section 105(b) applies in this context, it could serve as an openended qualification on contractual choice of law enforcement even in nonconsumer cases.

The analysis of the costs and benefits of contractual choice does not support UCITA’s limitations on contractual choice of law. The only reason the Reporter’s Notes give for the deviation from “freedom of contract” is that firms should not be able to evade mandatory consumer laws.207 In dropping the “reasonable relationship” requirement for commercial contracts, the Reporter’s Notes state that in a “global information economy, limitations of that type are inappropriate and arbitrary,” and concede that the limits of contractual choice of law for consumer transactions will impose “significant costs on Internet commerce.”208 Given the analysis of market and other constraints on contractual choice of law, it is far from clear that any consumer benefits of this restriction outweigh the costs to consumers, sellers and others. It is also not clear why licensors, by locating in a state that does not have mandatory laws, can escape the effects only if they rely exclusively on electronic delivery rather than distribution of physical copies. The rule could have the perverse effect of causing licensors to shift from physical distribution to electronic distribution even where the latter is less efficient. The Reporter’s Notes cite the costs of complying with the inconsistent laws of many jurisdictions as the reason for mandating application of the law of the licensor’s state in electronic transactions.209 However, the Reporter ignores the fact that a mass market software seller faces similar problems with physical delivery of the same

205 See RESTATEMENT (SECOND) OF CONFLICTS § 187 (1971).

206 See UCITA FINAL, supra note 7, § 105 (b).

207 See UCITA APPROVAL, supra note 6, § 109, reporter’s note 3.

208 Id.

209 See id. reporter’s note 2.

products or information. Perhaps the concern is that it is harder to block the sale of products to jurisdictions with onerous laws where the delivery is electronic rather than physical.210 Blocking technology might be able to solve this problem,211 but once again, the law trails technology.212

D. The Efect of Legal Constraints on Contractual Choice of Law

The previous Section shows that UCITA contains significant restrictions on contractual choice of law for consumer transactions. This Section shows that jurisdictional competition might emerge despite these or other restrictions on contractual choice of law.213

The prerequisite for state competition is enforcement of choice of law contracts in at least one state.214 The enforcement might be recognized in a state choice of law statute,215 a choice of law provision in a modified version of UCITA, or a judicial decision. For example, a state might adopt UCITA with a modified section 109 providing for enforcement of contractual choice of law irrespective of whether the contract involves a commercial or consumer contract, and whether the delivery was electronic or physical. States’ incentives to enforce contractual choice of law are discussed below. For now, it is enough to make the reasonable assumption that at least one state will enforce these clauses. The chosen state’s law would apply to contracts to the extent that the parties contractually designate the law of this state and the case is tried in the chosen state.

The hitch in the process, of course, is that an action on the contract might be tried in a jurisdiction that does not enforce contractual choice of law. Both UCITA and the general rule on enforcing contractual choice of law leave courts with substantial discretion not to enforce contractual choice of law, by either holding that enforcement would frustrate a fundamental state policy or that the chosen state lacks a sufficient connection to the transaction.216 Because state legislators control judges’ power, salary, and tenure, judges have strong incentives to preserve legislative bargains. Judges accomplish this goal by blocking exit through choice of law.217

210 See infra text accompanying notes 221-226 (discussing jurisdictional rules in Internet transactions).

211 See infra text accompanying note 229.

212 See supra text accompanying note 167.

213 For a more extended discussion of this process, see Kobayashi & Ribstein, supra note 129.

214 The enforcement might be by a federal court exercising diversity jurisdiction. For data indicating that federal courts have tended to enforce contractual choice to a greater extent than state courts, although they are required to apply the same rules, see Ribstein, supra note 182, at 285.

215 For a discussion of statutes in several large commercial jurisdictions explicitly recognizing enforcement of contractual choice, see Larry E. Ribstein, Delaware, Lawyers and Choice of Law, 19

DEL. J. CORP. L. 999 (1994).

216 See supra Part III.C.

217 See Ribstein, supra note 182, at 274-79 (discussing the judicial and legislative politics of enforcing contractual choice). That is not to say that courts necessarily will refuse to enforce. Among other things, courts might see enforcement as consistent with the legislature’s wish to avoid excessive costs to a powerful interest group. For a discussion advocating judicial interpretation of mandatory rules to permit enforcement of contractual choice unless explicitly prohibited by the legislature, see

It is, therefore, significant whether the party seeking enforcement of contractual choice can control which court decides the case. In litigation over whether a licensee violated terms limiting use, the licensor is normally the plaintiff, and can file a suit for breach of contract in the forum chosen by the contractual choice of law. On the other hand, in cases involving enforcement of terms that limit warranties, the plaintiff is likely to be the licensee, who can sue in a state that refuses to enforce contractual choice of law. Predicting the forum can be complicated. This difficulty results from the availability of declaratory relief in advance of a suit for breach of contract and the possibility of removal to federal court where enforcement of choice of law clauses is more likely.218

A party seeking enforcement of a contractual choice of law clause might attempt to increase the likelihood of enforcement through contractual clauses that designate the forum in which the case will be tried and the contracting parties’ consent to jurisdiction in that forum. But courts of non-designated states might refuse to enforce these clauses for reasons similar to those underlying non-enforcement of contractual choice of law.219 For example, UCITA provides that “[t]he parties in their agreement may choose an exclusive judicial forum unless the choice is unreasonable and unjust.”220

Courts may be somewhat more willing, however, to enforce choice of forum than choice of law clauses because the former, at least facially, do not involve the courts in an evaluation of legal rules, but only a consideration of the parties’ convenience. The UCITA Reporter’s Note explicitly adopts the test in two prominent cases favorable to enforcement of choice of forum clauses, Bremen v. Zapata Of-shore Co.221 and Carnival Cruise Lines, Inc. v. Shute .222 The Reporter notes that “a contractual choice of forum that responds to a valid commercial purpose is not invalid simply because it has an adverse effect on a party, even if bargaining power is unequal. The burden of establishing that the clause fails lies with the party asserting its invalidity.”223 He also quotes Carnival Cruise:

Because it is not unlikely that a mishap in a cruise could subject a cruise line to litigation in several different fora, the line has a special interest in limiting such fora. Moreover, a clause establishing [the forum] has the salutary effect of dispelling confusion as to where suits may be brought . . . .224

The Reporter adds: “In an Internet transaction, choice of forum will often be justified on the basis of the international risk that would otherwise

Michael J. Whincop & Mary Keyes, Statutes’ Domains in Private International Law: An Economic Theory of the Limits of Mandatory Rules, 20 SYDNEY L. REV. 435 (1998).

218 See supra note 214.

219 See Ribstein, supra note 182, at 282-83.

220 See UCITA FINAL, supra note 7, § 110(a) (emphasis added).

221 407 U.S. 1 (1972). See also UCITA APPROVAL, supra note 6, § 110, reporter’s notes 2-4.

222 499 U.S. 585 (1991). See also UCITA APPROVAL, supra note 6, § 110, reporter’s notes 2-4.

223 UCITA APPROVAL, supra note 6, § 110, reporter’s note 3.

224 Id. reporter’s note 4 (quoting Carnival Cruise, 499 U.S. at 585).

exist. Choice of a forum at a party’s location is reasonable.”225 As discussed below, this risk exists because of the problem of predicting which courts will be able to exercise jurisdiction over those who do business on the Internet.

The only completely reliable way licensors and other parties have to stay out of the reach of unfavorable states is to avoid contacts with them if those states would be able to exercise jurisdiction over the parties or apply their law to the case. This fact forces the parties to pay particular attention to the rules regarding jurisdiction in Internet transactions. Some commentators claim that Internet transactions might leave the parties vulnerable to suit everywhere .226 Another commentator has pointed out that no court has exercised jurisdiction over a defendant who did not affirmatively act to create effects there .227 There may, therefore, be no jurisdiction in a receiver’s state when a receiver downloads from a website and the seller remains passive in the receiver’s state. The jurisdictional rule for Internet transactions may ultimately depend on the technologies that emerge.228 For example, blocking software might enable a web broadcaster to control

downloads from the website .229

Jurisdictional competition ultimately will depend on the dynamics created by the legal rules on jurisdiction and choice of law. These dynamics exist along two dimensions. First, to increase the likelihood that a favorable state’s law will be applied to their transactions, firms may physically locate in those states. Physical location in a state increases the likelihood that that state’s law will be applied under both default choice of law rules and rules governing enforcement of choice of law contracts, as indicated by the UCITA provisions quoted above. Under the UCITA rule, the applicable law in the absence of an enforceable choice is the licensor’s state230 – i.e., place of business if the party has only one, chief executive office if it has more than one place of business, or place of incorporation or primary registration if it does not have a physical place of business .231 Information technology firms, whose value consists of intellectual property and human capital, may have lower costs of relocating than firms relying on physical capital such as factories. At the same time, attracting high revenue firms with high-paid employees can increase the tax base of the states with the most favorable laws. The prospect of higher tax revenues gives state lawmakers an incentive to lighten regulatory burdens on taxpaying industries.

Second, under the jurisdiction rules discussed above, licensors avoid

225 Id.

226 See David R. Johnson & David Post, Law And Borders-The Rise of Law in Cyberspace, 48

STAN. L. REV. 1367 (1996).

227 See Jack L. Goldsmith, Against Cyberanarchy, 65 U. CHI. L. REV. 1199, 1218 (1998).

228 See id. at 1224.

229 Note, however, that software triggered by the recipient’s address might be thwarted by users who download remotely from a different address.

230 See UCITA FINAL, supra note 7, § 109(b)(1).

231 See id. § 109(d).

sales in these unfavorable states, thus reducing the risk of being subject to a harsh legal regime. This avoidance imposes costs on consumers, who are unable to buy products that are available elsewhere. These costs, in turn, may translate into political action in states that would otherwise be prone to adopting consumer protection legislation. A potential consumer who sees a computer screen warning that certain software costs more in her state or cannot be downloaded there at all is likely to support reducing regulatory burdens. Accordingly, jurisdictional competition can develop even if states initially attempt to block it by refusing to enforce contractual choice of law. The parties’ inevitable ability to flee or avoid contact with harsh states can mute the impact of inefficient mandatory state laws. This situation, in turn, translates into political action by taxpayers and consumers who lose as firms relocate from and avoid contact with harsh states. In the end, regulatory states may reverse inefficient mandatory legislation or at least may take the intermediate step of enforcing contractual choice of law to encourage firms to sell locally.

This analysis is not intended to demonstrate that mandatory laws and restrictions on contractual choice of law are trivial. If states seek to impose mandatory rules and to protect those laws by barring contractual choice of law, it may take some time for an efficient equilibrium to develop. The most important lesson from this analysis is that the law may develop in unforeseen ways because of factors such as choice of law and jurisdiction and the way software is sold. This possibility counsels against hasty resort to drastic measures such as enactment of federal law that might solve some problems but create many others. Another lesson from this analysis is that the potential for state competition may affect uniform law drafting. As explained earlier, to ensure widespread adoption, uniform lawmakers anticipate possible state competition through choice of law.232 If uniform lawmakers could shut down this competition, they would be more free to adopt provisions favoring interest groups that would otherwise be weeded out through state competition. That uniform lawmakers recognize the potential for state competition reflects the persistence of this competition even in the face of states’ efforts to stop it.

CONCLUSION

The focus of the debate over UCITA and UCC 2B has been almost entirely on the positions the NCCUSL and ALI proposals should take on substantive legal issues. We have shown that this debate has missed some of the more important points concerning the future of software sales law. It is absurd to suppose that NCCUSL or ALI can, or should, ever bring the states together as one body over a matter as controversial as this. Rather, the focus should be on jurisdictional competition and not on uniformity. The efficiency of the law in this as in other areas ultimately should, and

probably will, depend on whether contracting parties themselves can determine the applicable law. Contractual choice of law not only addresses the same problem of legal indeterminacy that uniformity addresses, but it does so with a greater assurance of reaching an efficient equilibrium than under a top-down process of imposed uniformity. Moreover, jurisdictional choice is still feasible even if states refuse to enforce choice of law clauses as long as sellers can find a way to avoid oppressive laws. Thus, contractual choice of law is not only efficient but also inevitable. In the final analysis, how the law is made and who makes it will determine the future of software sales law to a far greater extent than anything NCCUSL and ALI have done so far.