Performance Risk, Form Contracts and UCITA
Leo L. Clarke[*]
February 20, 2001
7 Mich. Telecomm. Tech. L. Rev. 1 (2001)
Cite as: Leo L. Clarke, Performance Risk, Form Contracts and UCITA,
7 Mich. Telecomm. Tech. L. Rev. 1 (2001),
available
at <http://www.mttlr.org/volseven/clarke.html>.
Comments about this article should be sent to mttlr@umich.edu
I. INTRODUCTION
A. The Economics of Form Contracts and the Uncertainty of Perfromance Risk
B. The UCITA Approach (or Lack Thereof)
C. A New Rubric: Transactional Risk Rather Than Contract or Tort
1. The Commercial Context
2. Why Not the UCITA Default Rules?
3. Why Not the Restatement's "Reasonable Expectations" Rule?
D. Transactional Risk: Defining the Performance Obligation
1. The Software License
2. The Information Access License
3. Product Definition v. Performance Obligation
E. Transactional Risk: Enforcing the Performance Obligation
F. A Road Map
II. PERFORMANCE PROBLEMS AND FORM-CONTRACT TERMS
A. Form Terms Defined
B. The Scholarly Analysis
1. Llewellyn's Approach
2. More Recent Analyses of Presumed Consent
C. Shrinkwraps and Clickwraps: A Form by Any Other Name?
1. Lower Information Costs
2. Product Differentiation and Complexity
3. Clickwraps and Price
III. UCITA'S APPROACH TO FORM TERMS
A. UCITA and Assent to Form Contracts
1. The Statutory Text and the Official Comments
2. UCITA Propaganda on the Enforcement Issue
B. UCITA's Regulation of Performance Terms
1. Warranties
2. Choice of Law
3. Forum Selection
4. Remedies Limitations
5. UCITA's "Fundamental Public Policy" Exception to Enforcement
C. Where UCITA Went Wrong
1. The Drafters Were Too Certain about Certainty
2. UCITA is too Doctrinal and Not Commercial Enough
3. The Information Age Changed the Paradigm Form Transaction
4. UCITA Overvalues the Pro-Form-Giver Normative Argument
5. Certainty and Remedies Don't Mix
IV. ALLOCATING PERFORMANCE RISK THROUGH LEGISLATION
A. Paramaters for Performance Legislation
1. Transactional Reality v. Contract and Tort Concepts
2. Regulation v. Paternalism
3. Performance Obligation v. Performance Remedy
4. Risk Management v. Loss Allocation
5. Determining Whether the Performance Standard was Met: Risk of Failure and Risk of Loss
6. Insurance v. Collateral Source Rule
7. Commercial Behavior v. Litigation Posturing: Reducing Transaction Costs
B. A Model "Performance Risk" Statute
C. Comparing the Resolution of Hypothetical Disputes Under UCITA and the Model Statute
1. The Software Hypothetical
2. The information access hypothetical
3. A Look in the Rearview Mirror: How the Model Statute Would Have Applied to Y2K.
D. A Semi-Final Word on Performance, Risk, and Innovation
V. CONCLUSION
INTRODUCTION
A. The Economics of Form Contracts and the Uncertainty of Perfromance Risk
Problems in the performance of bargains have always plagued commerce. Whether it was the miller who did not receive the shaft, the building owner with the wrong pipes, or the bank with "millennial bug"-infested software, the failure of a product to meet a purchaser's expectations has always presented vexing commercial problems. Even when both parties to the transaction intend to honor their undertakings and act in a commercially honorable fashion, disagreements still arise over just what performance was due and how the problem should be solved.
It is not surprising that our legal institutions also have struggled to allocate these economic risks. For generations, performance problems in commercial transactions have been treated primarily as matters of contract law, and courts have looked to the intentions of the parties to allocate risks of performance failure. The contractual approach has particularly failed to provide neat answers when the transactors neither ignored those risks nor allocated them through negotiation, but instead used a standardized document that purported to assign all performance risks to the buyer through warranty disclaimers and limitations on remedies.
In these "form contract" cases, courts cannot confidently point to parties' actual, manifest, or mutual intentions, or blithely fill in the gaps with default rules on the presumption that the parties would have arrived at the same result if they had not ignored the issues. Nevertheless, for the most part, the black letter doctrine has enforced the form-giver's whim on the basis that the form-taker should have read the contract and demanded a change or refused to deal if she did not like the terms.[1]
The problem with this approach, at least with respect to complex transactions, is that courts of justice must turn a blind eye to a three-fold truth:
1. The form-giver uses the form to avoid the transaction costs of separate negotiation;
2. No rational form-taker would incur the costs necessary to price and negotiate the form performance terms; and
3. No rational form-giver would expect a rational customer to do so.
In other words, there is no tangible mutual assent and no factual basis, such as the form-giver's reasonable reliance, from which to imply assent worthy of public enforcement. When it comes to allocating performance risk, form contracts should not be worth the paper on which they are printed.
Scholars as far back as Kessler and Llewellyn have recognized the doctrinal and policy problems with the assent fiction.[2] Llewellyn also recognized the tendency of courts to use covert reasoning to avoid enforcing form contracts in particularly troublesome cases. No scholarly commentator has suggested that the form contract rules provide a satisfactory answer to the commercial problem of performance risk. So, one might think that the dawn of the "information economy" would be a propitious time to implement a new doctrinal approach. Apparently not: the National Conference of Commissioners on Uniform State Laws (the "Conference") has promulgated a comprehensive commercial statute that fails to remedy or even modify the law of form contracts in purely commercial transactions.
The Uniform Computer Information Transactions Act ("UCITA") - drafted to provide the background law for many of the most significant transactions in the information age[3] - simply accepts as holy writ "the duty to read" tradition that has been handed down by our forefathers. Specifically, § 113 of UCITA provides that the Act's default rules can be varied by agreement and §§ 112 and 208 provide that manifested assent to a standard form (by, for example, use of the product) constitutes assent to each term of that form.[4]
The Conference proclaims that the Act is "a statute for our time;" setting "forth uniform legal principles applicable to computer information transactions."[5] This article considers whether the several States should adopt the Act as drafted, when it continues a fiction that condones economic inefficiency and unfair resolution of disputes.[6] Concluding that UCITA cannot profess to be "a coherent contract law framework for analyzing a license . . . the dominant contractual framework for commerce in computer information"[7] without a more uniform treatment of performance problems in form transactions, I propose an alternative treatment based on efficient risk management principles.
B. The UCITA Approach (or Lack Thereof)
The Conference has promulgated UCITA to cover "a variety of transactions" such as "access by Fortune 500 businesses to sophisticated databases as well as distribution of software to the general public; it also covers custom software development and the acquisition of various rights in multimedia projects."[8] Indeed, UCITA is a marvel of legal engineering, the result of a four-year drafting (and lobbying) process that addresses a myriad of enigmatic issues created by the collision of contract law, commercial law, intellectual property law and the ever-accelerating development of the Internet and other electronic computer, telecommunications and multi-media technologies.[9]
Nevertheless, for all its achievements, UCITA is fatally flawed when it comes to the enforcement of standardized performance terms in commercial licensing transactions.[10] It is the premise of this Article that the Conference has tripped over its own bandwidth in promulgating UCITA and thereby missed a golden opportunity to bring the legal framework into alignment with the realities of those transactions.
It is ironic that the Conference, in purporting to follow the lead of U.C.C. progenitor Karl Llewellyn,[11] missed his real message. Llewellyn, a father of "legal realism," believed that the law must reflect, not dictate, the substance of commerce.[12] Llewellyn's "reflection" was not an end in itself, but a means of achieving an efficient and dynamic economy. Enforcement of standardized performance terms will not achieve that lofty goal: the transactional costs of investigating, evaluating, and negotiating performance risk allocations simply preclude the bargaining necessary for presumed efficiency.[13]
UCITA's archaic notions of "freedom of contract" and presumed assent,[14] flowing from a 19th Century "duty to read,"[15] ignore decades of prominent scholarship that demonstrated the failure of contract doctrine to address adequately the phenomenon of standardized contracts.[16] As a result, the Act fails to provide a commercially or legally realistic allocation of risk when products fail to perform in transactions documented by boilerplate terms.
The failure of UCITA is particularly surprising because the Conference is so acutely aware of the significance of the information economy, the benefits of information technology, and the need to reflect the realities of commercial practice.[17] In light of that awareness, however, the Conference apparently has approached the problem using an obsolete "image."[18] UCITA's image is that commerce will depend on efficient "default rules" and the freedom of parties to negotiate out of those default rules. In fact, the reality of the new economy, as it was in the last stages of the old, is that the low per transaction costs of drafting and imposing standardized contract terms makes default rules largely irrelevant except perhaps as backdrops for a status quo bias when parties do negotiate terms.[19]
The new image that UCITA should have addressed is a transaction where the licensee "accepts" boilerplate terms because she is both trusting and sophisticated enough to know that (1) she does not have, and cannot economically attain, sufficient information about a proprietary product[20] to evaluate accurately the performance risk presented by the terms, (2) the licensor neither expects nor desires negotiation of those terms and (3) the term will be employed only as a last resort measure in the unlikely event that business considerations fail to resolve an unanticipated performance breakdown. Insofar as UCITA nowhere deals with this image, it is destined to fail in its quest to make commercial licensing transactions more certain and efficient.
C. A New Rubric: Transactional Risk Rather Than Contract or Tort
1. The commercial context
Instead of focusing solely on a passé image of arm's length contracts, the new commercial statute should address the reality of bargaining failure. That reality is the failure of both parties to address an issue as part of their "dickered terms" and/or their failure to provide a commercially fair and just resolution after the problem occurs. For it is a business reality, one that Karl Llewellyn would recognize if he were alive today, that business people - as opposed to lawyers or drafters - do not expect boilerplate provisions to resolve issues, and do not expect them to govern unanticipated problems.[21]
The business context is quickly sketched. Large and small businesses enter into innumerable UCITA transactions in which they do not consciously allocate the risks surrounding the "product" if it fails to perform to their expectations. That they act this way, even in a culture dominated by designed obsolescence, frightful waste, and rampant refusal to accept responsibility, should not be surprising. Commercial actors ignore risks because conscious allocation is too time-consuming for simple transactions or beyond the transactor's cognitive limits in more complex deals.
Even sophisticated businesses that regularly acquire "computer information" products cannot fully evaluate the performance characteristics of increasingly complex technology, the risks of performance failure, or the probable resulting monetary losses. It is possible to construct performance risk forecasts, but the models require so many assumptions that only the naïve believe that the resulting projections are little more than guesses. In short, the transactional costs of investigating performance risks and negotiating contractual terms are seldom worth the candle. And when they are, we can expect the parties to negotiate, not to rely on forms. In this world, when the parties resort to form contracts, allocating performance risk through notions of presumed assent and manufactured intention amounts to sheer delusion.
Thus, contrary to the repeated assertion of the Conference, the issue is not "contract" and the need is not for a "contract statute." Instead, we are in a no-man's land between the regimes of contract and tort. Just as community life goes on regardless of where governments draw political boundaries, commerce continues regardless of where courts and legislatures draw their conceptual lines. The commercial need for legislation, if there is one as the Conference claims, is for rules to be applied when (1) there is a claim that a product has not performed; and (2) there has been no actual (as opposed to constructive) meeting of the minds on what performance obligation the licensor undertook or what remedies the licensee should have to compensate for any failure of performance.
I do not address here the issue whether legislation is the best means of dealing with this issue or whether uniform legislation is appropriate at this time. Rather, I limit my conclusion to the following: if legislation is enacted, it should cast aside the historic legal pigeonholes of contract and tort, and instead should create obligations and remedies based on the characteristics of the transaction, except to the extent that the parties actually negotiate to the contrary.[22] The two key determinants of risk allocation should be (1) that a particular risk of performance should fall on the transactor that can best bear or spread the risk of loss; and (2) that the protected transactor should have an economical remedy for fairly enforcing that risk allocation.
2. Why not the UCITA default rules?
Why draft a whole new statute to address performance risk? UCITA includes carefully drafted majoritarian default rules which we are assured represent the prevailing commercial practices in UCITA transactions. The obvious solution, then, is to declare that form contracts are unenforceable and that the UCITA gap-fillers apply. Four closely related reasons why I reject that solution follow:[23]
1. UCITA incorporates too much from Article 2 of the Uniform Commercial Code. Those rules cannot shake their old-economy genetic code. They are based too much on common-law, formalistic judicial notions inconsistent with modern risk allocation principles.
2. UCITA's default rules are not sufficiently explicit regarding the role of risk management in assigning liability and determining damages. The explicit freedom of contract philosophy of UCITA will sustain a status quo bias that will lead old school judges and lawyers to under-appreciate the new practicalities of the information economy. They will tend to continue to translate new problems to old images.[24]
3. UCITA's warranty terms are too lax. Its minimal warranty standards, which are admittedly rooted in past law, do not provide a sufficient threat of liability to induce producers to efficiently incorporate product failure liabilities into pricing. Therefore, individual end-users are left to absorb randomly occurring losses from performance failures. That result is simply not efficient.
4. UCITA does not include remedial provisions specially designed to deal with the thorny issues of product performance failures presented in UCITA transactions, especially given the increasing complexity of computer networks and information databases. Explicit damage rules are necessary to provide the incentives to licensors and licensees to be commercially reasonable in minimizing total transactional risk rather than their own risk.
3. Why not the Restatement's "reasonable expectations" rule?
A second natural alternative to the enforcement of form contracts would be to adopt the "reasonable expectations" test of § 211(3) of the Restatement (Second) of Contracts.[25] Section 211 of the Restatement explicitly binds a form-taker whether or not she reads or knows of the terms.[26] Subsection (3) knocks out any term, however, that the form-giver should expect that the form-taker would reject if the form-taker were aware of it.[27] This standard is called the "reasonable expectations" rule, although the section invites confusion as to whose expectations are relevant.[28] In fact, one could argue that there are three relevant "expecters:" the actual form-taker, the hypothetical reasonable form-taker and the form-giver.[29]
I reject § 211(3) as a replacement for UCITA's formula for three simple reasons. First, as White demonstrates, its history does little to commend it.[30] Section 211(3) has not been employed extensively in commercial transactions. Instead, it has been used most often in consumer insurance cases and most often by an activist appellate judge in Arizona.[31] Second, the test merely invites covert action by courts since it is prone to ex post rationalization in light of the actual turn of events regardless of the ex ante probability of the event.[32] Third and most fundamentally, the rule is under-inclusive with respect to the factors that support non-enforcement of the terms. It is not just the presence of the term that creates the inefficient result. Indeed, a lack of consent and inefficiency is much more likely to be the result of imperfect information, cognitive limitations and/or transactional costs. In short, recourse to "reasonable expectations" would create more problems than it would solve.
D. Transactional Risk: Defining the Performance Obligation
The transactional risk management approach suggested here requires an explication of the concepts of "performance obligation" and "performance remedy." Before proceeding to that discussion, however, the reader who is not familiar with UCITA transactions - most commonly software licenses and information access contracts - may wish to consider the following two hypothetical cases, which should put the issues in context.
1. The software license
The success and growth of Holly, Orbison & Ely ("HOE"), a civil engineering firm in Lubbock, Texas led it to upgrade its separate time-keeping and billing software programs that could only handle 20 users. Jennings, HOE's office manager, purchased from Minit-Paid Corp. for $9,000 a 25-year, 30-user license for time and expense accounting and billing software. The magazine ads for the software stated the program was "Windows 98 Compatible," the "most advanced time management software available for mid-sized professional firms" and "top-rated" by ProTech Review. Moreover, the sales representative of the program reseller told Jennings that other accounting, legal and engineering firms in town had installed the program without a hitch in less than an hour and that they were running it successfully with no substantive problems.
HOE's computer network, however, crashed when its in-house computer specialist attempted to install the program - making the network inoperable for six hours. When that problem was solved, HOE began using the program. At the end of the first month, however, the program was unable to produce consolidated billing records for those matters on which more than one engineer worked. This glitch required accounting personnel to work sixteen hours of overtime to create HOE's invoices for those matters and to properly credit the engineers for time worked.
Jennings, embarrassed over the debacle, considers what remedy HOE might have for the substantial out-of pocket losses in installation costs and processing timesheets and invoices. He would also like to recoup some of the lost income from the network crash and the late billings. When he reads, for the first time, the "clickwrap" license agreement that appeared as the first screen after the computer specialist commenced the installation process, he notes the following terms:
-There are no express warranties, except that the CD is free from defects in workmanship and materials;
-All implied warranties are disclaimed;
-The licensor's sole liability to the licensee, whether in tort or in breach of contract, is limited to replacement of the CD within five days after return by the licensee of the original CD;
-The licensor is not liable for consequential or incidental damages;
No written or oral representations or statements made prior to the purchase are part of the agreement between the parties;
-The licensee agrees not to issue any review of the product, including adverse comments to other actual or potential licensees, without the licensor's consent; and
-Any claims against the licensor are subject to arbitration under the rules of the International Chamber of Commerce, must be commenced in Massachusetts and are governed by Massachusetts law. Moreover, the only issue before the arbitrators shall be whether the licensor complied with the limited replacement remedy.
Naturally, the specialist had clicked on the "I agree" dialog box without reading the program so that she could complete the installation before lunch. The program would not have permitted the installation process to continue if she had not. In light of these provisions, Jennings concludes that the cost of investigating the cause of the problems and retaining an attorney to contest the license agreement far exceeds the cost of the software, the lost time and the lost income.
Unknown to Jennings, the compilation function of the program would have worked fine if information had been entered directly into the program. Most of HOE's employees, however, were maintaining their time records on a common (but not market-dominant) spreadsheet program and then "cutting and pasting" the information into the time-keeping program. Internal formatting information from the spreadsheet program that was picked up in the "cutting" step interfered with the time-keeping function's ability to recognize and transfer that data to the composite billing function. The software documentation did not indicate that such a practice would create problems.
2. The information access license
Geodata Systems provides computerized geographic and demographic information to a variety of customers through access to a database of layered graphical maps coded by zip codes or census tracts with the information requested by the customer. For example, the customer can order a map that color codes metropolitan areas according to rankings based on zoning; population characteristics including income, race, ethnic origin and education; transportation and utility information; local real property and tax records; presence of retail chains; number of churches and libraries; and a myriad of other factors. Customers can visit Geodata's website, identify a geographic area and the information they desire regarding that area. Geodata will then create a layered graphical map coded for each type of information requested.
The cost of the access license varies according to projected usage and the number of component databases that the user may access. Geodata's standard license agreement describes the product offered as "geographic information systems derived from sources deemed reliable by Geodata." The agreement disclaims any responsibility for the accuracy of any data obtained from third parties or the accuracy of the maps themselves. It also includes a standard integration clause, a disclaimer of all express and implied warranties, an exclusive remedy of an extension of access for six months, and a denial of liability for incidental and consequential damages. No information is provided on the site about the details of the sources of the underlying data.
Spacechain, an Internet service provider, has been using Geodata information for the last year to plan its mail distribution of disks offering its service. Spacechain has developed its own criteria for ranking markets based on various economic and social factors and uses Geodata to match its criteria to markets around the country. Spacechain maintains records on the response rate of its mailings according to its internal market rankings. Generally, response rates correlate very highly with rankings. However, Spacechain has just noticed that the correlation between ranking and response has been markedly different for the State of Michigan. After considerable investigation, Spacechain has learned that Geodata stopped adding new information to the database from state and local agencies in Michigan after prices for that information were dramatically increased two years ago.
Spacechain has paid Geodata $65,000 for Michigan maps and has spent $350,000 in mailings to Michigan over the last two years. It estimates that the lower response rate compared to other states has cost it $6.6 million in lost revenue. Spacechain signed Geodata's standard license agreement after negotiating the price, the number of simultaneous users and number of databases to which it would have access. It did not negotiate any other terms of the transaction after being told that the sales staff did not have the authority to change the terms. Geodata's representative did tell Spacechain's buyer, "We've never let this stuff interfere with a good relationship, and we've never enforced any of this boilerplate."
Spacechain has demanded that Geodata credit Spacechain for the amount spent on Michigan maps, reimburse it for its other losses, and warrant that the rest of the database to which Spacechain has access contains the most recently available data. Geodata has cited the disclaimer in the license agreement and the provision denying any liability for incidental and consequential damages. Spacechain responds that the disclaimer cannot be deemed to permit it to offer obviously stale information. But Geodata's rejoinder is simple, "Our researchers concluded that Michigan's demographics have not changed substantially enough to make the information unreliable for all purposes, and our agreement disclaims any warranty of fitness for a particular purpose."
3. Product definition v. performance obligation
The two hypotheticals demonstrate the nature of the performance risk problem. In each, the licensee has not gained the benefits expected, perhaps reasonably, from the license.[33] In neither of the cases did the parties knowingly allocate the risk of the particular performance failure involved. Can we conclude, therefore, that the licensee therefore accepted all risks not expressly or customarily accepted by the licensor? UCITA would say so. But is risk allocation a question of contract alone, or is there something else at work here?
One way to resolve this risk allocation issue would be to look first to what is being licensed. As UCITA's Prefatory Note states, a significant attribute of products in UCITA transactions is that the product is often defined solely by reference to the licensing agreement.[34] A toaster can be described separately from an agreement to sell it, but a license to use a software program cannot be defined simply by reference to a computer disk or the contents of the disk. Nevertheless, many aspects of the definition of the licensed product (number of users, location of use, term, rights to reverse engineer or modify) are unrelated to how the product performs. In other words, the seeming specificity of licensing agreements as to product description (hence "specifications") may shed little light on the performance obligation.
For example, the Minit-Paid transaction seems to have an explicit description of the product being licensed - a CD, which a computer with described specifications could read, that contains a program that performs some sort of time-keeping and invoice creation functions that can be used by up to 30 users for a 25 year term. The very nature of transaction, however, demonstrates that precise contours of the product's performance are not established in the agreement. It is far from self-evident that customers would pay $9,000 to purchase what the form literally provides: a series of options to acquire CDs until one is sufficiently non-defective to allow the computer to read and perform some sort of time-keeping and billing functions. The form description does not answer the question of how what is being licensed will accomplish the purposes for which Minit-Paid designed the program and for which HOE licensed it. We know that a toaster is expected to produce edible toast in a reasonably consistent and timely fashion. But what, precisely, is the software to do? Is it software that will install without causing the network to crash and handle time-keeping for the maximum number of licensees without glitch or is it just a program that gets a licensee most of the way there most of the time?[35]
Similarly, in the Geodata case, the form undoubtedly specifies what is being licensed, but there is substantially less information as to how the product is to perform. Was the mapping component the primary feature (value) or is the underlying data just as important? Does the license just allow access to whatever happens to be "in" the database at any given time? Or does the licensor have some inherent obligation to make the information both accessible to its customers and reliable for their uses?[36]
This distinction between product definition and performance obligation reflects the differing interests of the form-giver licensor and the form-taker licensee. The form-giver licensor is very interested in making sure that the form does not provide any more rights than the licensee is paying for. On the other hand, it has little interest in delineating the performance requirements of that product since setting a performance standard can create a potential liability that its warranty disclaimers and other exculpatory provisions would otherwise avoid. Moreover, because the licensee has no input into the terms, missing from the form's description of the product is any reflection of the form-taker's expectations regarding precise performance characteristics.[37]
The one-sided terms of form performance standards should come as no surprise given the parties' relative access to relevant performance data. These information disparities are of two types. The first, but less significant, is that the parties do not have equal access at equal cost to information about how the form document purports to allocate risks and the effectiveness of those allocations. The second disparity is that the parties do not have equal access at equal cost to information about the actual value of those risks (probability times damage) and as to the cost of risk avoidance or transfer.
For example, licensors Minit-Paid and Geodata have the following term-related information advantages: (a) they already know what the form contract says; (b) they know whether, in the event of a performance problem, they will accommodate the licensee's complaint or rely on the form; and (c) they have in all likelihood obtained legal advice as to the probable judicial treatment of the terms. They also enjoy decided advantages as to the performance characteristics of their products, especially if they keep records of beta test results and customer complaints. Finally, their control over product design, production and distribution allows them to spread costs over a larger number of events and thereby provides substantial advantages in reducing the risks of performance problems, resolving problems and spreading the losses that do occur.
These disparities mean that licensees HOE and Spacechain would have to incur substantial transactional costs to reach informational equity with their licensors so that they could price the form terms. Unless the price and risk reductions that could be achieved by acquiring that information exceed the costs of obtaining that information (or confirms that the terms of the transaction are otherwise favorable), neither the licensees nor society can be confident that bargaining will reach efficient results.
In sum, absent an ex ante statement of performance obligation, we are left with determining ex post which transactor was in a better position to bear the particular loss actually experienced. Only then can we back into the judgment that an efficient (bargained) exchange would have included that performance definition or obligation.
E. Transactional Risk: Enforcing the Performance Obligation
Allocation of the performance risk, however, is only one part of the conundrum presented by the above hypotheticals. The other part is how to enforce an allocation of risk to the licensor when the loss from that risk is deemed to fall on the licensee. In the illustrated scenarios, enforcing the form terms would mean that the licensee bears the risk of the licensors' failure to meet any performance obligations they had assumed. Moreover, the in terrorem intent of the form greatly reduces the likelihood that form-takers will even try to enforce a performance obligation.
There is nothing in the statement of either hypothetical that makes it self-evident, from the viewpoint of the parties or of society, that a licensee should be barred from enforcing an otherwise optimal allocation of risk. Certainly, there is no evidence of actual bargaining as to remedies, so the "contract" language is only one piece of the puzzle. If we conclude that a statute should allocate to the licensor responsibility for the performance breakdowns in the hypotheticals, it certainly does not follow that the licensor should be able contractually to preclude the licensee from enforcing that allocation, at least in the absence of real consent. To the contrary, it would seem to follow that un-bargained-for limitations on remedies should not eviscerate the statutory risk allocations. To the extent that UCITA adopts such an outcome, one would expect the Conference to offer a cogent explanation of its reasons. Unfortunately, it does not.
F. A Road Map
This Article is limited to UCITA's treatment of performance problems in the context of form contracts used in transactions between commercial (i.e. non-consumer) parties. Part II will describe the scholarship regarding form contracts in particular-scholarship that apparently has been rejected or ignored by the Conference. Part II also analyzes whether shrinkwraps and clickwraps, as prototypical UCITA contracts, present any particular basis for rejecting scholarly analysis.
Part III examines in detail the most relevant provisions of UCITA and the Official Comments supporting those provisions. True to its word, UCITA's formalistic freedom of contract approach would enforce standardized contractual terms regarding performance obligations and remedies, despite any empirical basis for finding a meeting of the minds or any persuasive demonstration that presumed assent is consistent with public norms.
Part IV explains an approach that focuses on risk bearing and spreading and demonstrates why that approach is more efficient and fair. I present a model statute that would implement such an approach. The statute breaks down the problem into five parts: definitions, specification of product performance standards, tests for determining whether those standards were met, remedies for performance failure, and a deferred effective date. I then apply UCITA and the model statute to the foregoing hypothetical cases to analyze the difference in outcomes and incentives. I conclude Part IV with a consideration of how the model statute would have applied if Y2K had lived up to its billings, and how it might affect innovation in information technology.
II. PERFORMANCE PROBLEMS AND FORM-CONTRACT TERMS
A. Form Terms Defined
The threshold issue is defining "form terms." I will use the term here to describe a commonplace but very specific transaction:
-a set of standard "contractual" terms
-in written or electronic format
-prepared by one party (the form-giver)
-for use as the standard terms in commercial transactions as to which it is a party
-presented to the other party (the form-taker) in circumstances where the parties do not in fact negotiate the terms, and
-where the form-giver has no reasonable expectation that the typical form-taker would incur the costs necessary to evaluate the terms and their value/costs to the form-taker.
Thus, the paradigm transaction is one where the form-giver cannot reasonably expect an economically rational form-taker to obtain and evaluate all the information necessary (1) to calculate the legal, technical and economic risks of performance; (2) value accurately the prices of those performance terms; and (3) negotiate substitute contractual terms to allocate those risks.[38]
Crucial to this analysis, it is worth restatement that a "form contract" means a non-negotiated set of "contract terms" where the reasonable form-giver would expect that the form-taker would be required to incur irrational transactional costs to make an informed (efficient) consent to the terms. I do not include within the scope of the concept any term that the parties actually negotiate. For example, if the parties negotiate the terms of a warranty, the warranty provisions and the remedies for breach of warranty would not be included within the concept of "form terms."
UCITA itself has a slightly different definition:
"Standard form" means a record or a group of related records containing terms prepared for repeated use in transactions and so used in a transaction in which there was no negotiated change of terms by individuals except to set the price, quantity, method of payment, selection among standard options, or time or method of delivery.[39]
This definition differs from mine in two primary ways. First, UCITA permits negotiation only on specified terms: if the parties go beyond those terms, for example to negotiate a longer notice period before acceptance or waiver of rights, then the entire contract falls out of the definition.[40] My definition focuses on the non-negotiated nature of specific terms or issues because the fact that the parties negotiate a matter of immediate concern with a high probability of occurrence does not mean that they will or should rationally negotiate provisions on more contingent events such as performance breakdowns. This conclusion follows from the fact that, in UCITA transactions, material product performance disputes are relatively rare.
Second, and more importantly, the UCITA definition does not include my last element, that the form-giver must not have a reasonable expectation that the form-taker can evaluate the term accurately.[41] This element is crucial to properly limit special treatment of form terms to those transactions where the information costs and obstacles are sufficient to create a high probability of an inefficient bargain. Because UCITA approaches the whole problem from a different perspective - caveat form-taker - it does not include that element.
Note that my definition and UCITA's do not incorporate many of the factors cited in discussions of standardized contracts. Neither definition turns on the relative size of the parties, their relative bargaining power, the size of the print, or the complexity of the transaction or the form language. It doesn't matter whether the form-taker actually read the form or whether the form-giver made the terms at issue conspicuous or pointed them out to the form-taker, or even whether there was a "take-it-or-leave-it" aspect to the transaction.[42] These are all aspects of what are commonly called contracts of adhesion. The image of such contracts generally relates to consumer transactions with fine print forms and subjects like insurance, mortgages and durable goods purchases. Such concepts add little to the debate about the enforceability of forms in the commercial context, and I eschew them here.[43] As Towle makes clear, in the information economy even consumers have, or will soon have, the ability to be form-givers, and the investigation into the relative bargaining power of a dot.com e-commerce site and a licensee of free software would be fruitless.[44] Thus, the primary determinant is the bar that information costs create to informed assent.
Neither UCITA nor I find any pejorative connotation to follow as a matter of course from form terms. They are a rational product of any business which produce substantial benefits for all parties. Thus, form terms may include provisions that are solely in the form-taker's favor (e.g. a money back guarantee) as well as those solely in the form-giver's favor (a disclaimer of warranty). They may or may not have been taken into account by the form-giver in setting product price or features. They do not imply any overreaching, fraudulent, deceptive or unfair trade practices. They are neither necessarily inefficient nor unfair. When any of these unfavorable elements is present, doctrines of unconscionability or conflict with public policy may come into play, but I will assume in this article that both parties are acting as rational economic players, just as society would want them to act. In short, it is the not the existence of form performance terms, but their random impact that creates the problem.
B. The Scholarly Analysis
Now we can turn to the question of whether form terms should be enforced. There is little new about the doctrinal issues presented by form contracts. The crux of the matter is the fact that the parties' transaction includes as one aspect a set of contractual terms presented by one party as its "standard" terms of doing business and the other party does not contest the inclusion (imposition) of those terms. The issues created by this situation were extensively considered by Llewellyn[45] and have been thoroughly analyzed by Slawson in 1971, Rakoff in 1983, and Eisenberg in 1982 and 1995, to name but a few.[46]
1. Llewellyn's approach
Llewellyn acknowledged that the notion of mutual assent to form terms was a fiction, but he emphasized the necessity of enforcing the dickered bargain.[47] While this portion of his analysis was undoubtedly correct, Rakoff has shown that Llewellyn implicitly recognized that the case for enforcing all terms - and especially "contingent" or "invisible" terms such as those dealing with performance and remedies - was not nearly as strong.[48] In other words, while Llewellyn could be understood to have adopted the position that a form-taker "must take the good with the bad,"[49] the result could not be justified under generally accepted notions of efficiency or justice. Thus, even UCC Article 2, clearly Llewellyn's brainchild, does not impose a pure duty to read/enforce all forms, but adopts a layered approach to contracting among merchants based on reasonable presumptions of assent.[50]
2. More recent analyses of presumed consent
More recently, scholars such as Slawson, Rakoff and Eisenberg have developed a more consistent analysis demonstrating why a doctrine of presumed intent should not be used to enforce form contracts. An amalgam of the various analyses would proceed along the following lines:[51]
-The basic premise of contract law is that the public justice system should enforce bargains only if (and because) the defendant has consented to the obligation/forbearance the plaintiff seeks to enforce.[52] Thus, it has been widely acknowledged both by traditional doctrinalists[53] and by law-and economics-theorists[54] that there is no just basis for the application of contract law absent actual or presumed assent.
-In a partially completed transaction (i.e., the licensee has received the product or access to it) the fact that a party has not negotiated specific terms or even read the contract does not preclude recognition that a contract exists.[55] Form contracts promote efficiency by reducing transaction costs that would otherwise be incurred to negotiate and document routine (at least from the viewpoint of the form-giver) transactions. They are also supported by notions of justice in that the form-giver has parted with value in the transaction, and her reliance should be protected.
-Recognition, however, of an enforceable agreement's existence, to the extent of the dickered bargain, does not necessarily mean that each term of the form contract should be enforced.[56] The traditional duty to read the contract was merely a control principle to preclude the form-taker from avoiding bargains that had appeared to be advantageous ex ante but proved ex post not to be so great. No public policy supports that duty when the form-giver could not reasonably expect that the form terms were read, understood or assented to.[57]
-In short, it should be irrelevant whether a form-taker actually reads, understands or fails to negotiate a form term, because the essential purpose of the form is to avoid the necessity of negotiation and documentation.[58] To impose a duty to read a contract that the form-giver does not want to negotiate, probably has not authorized its employees to negotiate, and has not relied upon, would unjustifiably extend the duty to read.[59]
-Indeed, lawyers and economists argue: (1) only a contract that involves an actual meeting of the minds will create a value-maximizing exchange[60] and (2) contracts will produce efficient results only if the form-taker has the information necessary to price each contract term by comparing the discounted present value of the future "service stream" it will receive from the product and the discounted present value of the sum of outlays over time for purchase, maintenance, repair, and rights enforcement.[61]
-Thus, when the issue is limited to non-negotiated terms of form contracts where there is no basis for a finding of actual assent to the particular term (as opposed to the transaction in its entirety), enforceability of that term must turn on whether assent to the specific term can be inferred or implied.[62] The only basis for such a factual inference would be when the term is one that a rational person would have agreed to if she (a) had considered it and (b) had the information to properly evaluate both the meaning of the term and the present value or cost of the term if it were enforced.[63] In that case, enforcement would be efficient and consistent with concepts of both consumer sovereignty and social welfare maximization.[64]
-So far, the analysis would conclude that the common law should enforce form terms against a form-taker when the term would have been favorable or neutral ex ante. Thus, a form-taker cannot take advantage of the lack of actual assent to "second guess" a bargain in light of ex post outcomes.[65]
-On the other hand, it follows that there is no reason to enforce terms that could only benefit form-givers ex ante. Such terms are inefficient (would not be agreed to by a rational person with full knowledge of the term and its present value) and assent therefore cannot be implied.[66] The remaining uncertainty is how to handle terms where the ex ante risk is unknowable by the form-taker because information, other transaction costs or cognitive limitations make knowledge uneconomical.
-The form term is no more than the biased risk allocation by the form-giver and is entitled to no presumption of enforceability. Since the question of contract vel non is determined by the partially executed nature of the transaction, the statute should enforce the contract according to the dickered terms along with "default rules" that reflect the legislature's views as to generally efficient outcomes.[67] This is preferable to an open-ended default rule that would delegate absolute discretion to the court in individual cases because the legislature is in a better position than a court to determine efficiency.[68]
-An argument could be made that the form terms should be enforced if the form-giver demonstrates that the term is consistent with trade custom in negotiated transactions, because trade customs may be presumed to reflect efficiency.[69] Even terms generally accepted in the trade, however, may be inefficient. Terms can be presumed efficient only if the standard assumptions of the modern free-market economics model are satisfied.[70] Those assumptions include (1) that all parties have complete information regarding all relevant facts (risks); and (2) transaction costs are minimal. When the form term is one that is beyond the form-taker's cognitive capability or rational boundary,[71] these assumptions cannot be satisfied, and there is no reason to give weight to the customary term.[72] Such cognitive problems are especially significant in complex transactions where forms include many terms, where the transaction is not routine for the form-taker and where the terms deal with future contingencies.[73] These issues suggest that performance terms based on "uninformed" trade usage may not provide efficient results in negotiated transactions, and certainly will not do so in form transactions.[74]
In sum, synthesis of scholarly comment reveals that form terms should not be enforced because there is no basis for concluding that they are the product of informed consent (a meeting of the minds) or cognitive risk-taking.
A responsive argument could be made that the form-taker's failure to negotiate the terms demonstrates a willingness to gamble on whatever risk allocation is proposed in the form. This seems to be the perspective of the duty to read adherents. But the law should not infer that a form-taker is gambling because there is another, more logical explanation: far from gambling, the form-taker is acting rationally within the limits of cognition. [75]Therefore, no form-giver would be justified in relying on an assumption of risk acceptance because the form-taker's conduct is entirely consistent with her desire to avoid transactional costs - costs which would only reduce the net value ultimately transferred to the form-taker in the vast majority of cases where performance is not a material problem.[76] From society's viewpoint, there appears to be no justification for encouraging one party to incur irrational transaction costs just because the other party has sought to impose inefficient terms.[77] The essence of the analysis that form terms should not be enforced under a contract theory, then, is that any presumed assent by a form-giver is not sufficiently related to economic efficiency or justice considerations to justify public enforcement.[78]
C. Shrinkwraps and Clickwraps: A Form by Any Other Name?
Shrinkwraps and clickwraps are but a special species of form agreements. It is somewhat surprising that the question whether they should be enforced according to their terms should have generated such a substantial amount of commentary.[79] This is especially so given the limited number of cases that have addressed the issue.[80] Much of this commentary, however, can be explained by the novelty of the specific terms or by the fact that the context for the discussion of enforceability was the analysis of copyright and other intellectual property law issues. Enforceability of contract terms is ancillary to the crux of those problems. That is, the issues are made more complex by the form terms, but they are not dependent on those terms.[81] Even the relatively few articles that discuss form terms in the context of performance problems that might arise in UCITA-type transactions generally leave the topic without resolving the ultimate enforceability issue.[82]
Nevertheless, because the primary problems with form terms arise from information disparities, the continued development of information technology may suggest that shrinkwraps and clickwraps should be enforced for reasons that do not apply to traditional forms. It is therefore worth addressing in some depth whether shrinkwraps and clickwraps do present any unique issues of enforceability under the scholarly analysis.[83]
The issue is easily resolved with respect to shrinkwraps. Their distinctions from the run of the mill agreement are the timing of supposed assent and their relationship to the licensing of intellectual property. Shrinkwraps in and of themselves, however, present no unique issues that favor enforcement. To the contrary, shrinkwraps present more obstacles to enforcement under the basic analysis than traditional forms since they are "reverse unilateral contracts"[84] at best and require an expansive view of "layered contracting" in order to be enforceable at all.[85]
Clickwraps, however, present stronger arguments for enforcement because their on-line nature allows both the form-giver and the form-taker to take advantage of developing information technology to overcome some of the information obstacles to presumed assent. There are at least three factors that should be taken into account in analyzing this issue.
1. Lower information costs
Examples of form contracts include parking lot tickets, insurance contracts received after the premium is paid, and car rental agreements shoved across the counter at the weary traveler. While there are also harried parties to license transactions, clickwraps do not present the same image of the time-constrained form-taker. For example, in In re RealNetworks, the court enforced arbitration and choice of forum clauses in a clickwrap in part because licensees could view the terms and print them at leisure in the peace and quiet of their own homes. [86]
Information technology also has the potential for allowing form-givers to increase dramatically disclosures to their form-takers. For example, a clickwrap could include links to annotated forms that provide supplemental explanations and pros and cons regarding particular clauses, including even links to objective third party sites like licensees' benevolent associations. Such developments could dramatically reduce the current contractual information imbalance between the parties.
This is not to say, however, that the mere increased disclosure regarding contract terms should not make clickwraps enforceable as a matter of course. Commentators have rightly pointed out the prohibitive cost imbalance between the parties in obtaining legal advice to analyze form terms,[87] but understanding the legal effect of the terms is just one transactional cost. In a sense, it is the least important because most commercial form-takers will understand the general effect even of common non-dickered terms such as warranty disclaimers and remedies limitations.
More significant are the costs of obtaining other information necessary to "price" the term (i.e. information about other performance disputes concerning the product). Even though internet and database technologies should dramatically reduce those costs, we can expect that form-givers, if provided with an enforcement incentive, would provide such additional information themselves.[88] For example, a form-giver might include with its clickwrap a link to information on costs of arbitration and even results in arbitrations involving the terms at issue. The benefits of technology may be so great as to permit the growth of competition in performance terms through sites that compare the forms of different providers as well as product performance characteristics.[89]
These developments would be positive, but for three reasons they are unlikely to solve the enforcement problem identified by scholars. The first is that providing more information and disclosures does not eliminate transaction costs but only substitutes some costs (hopefully lower) for others. Even relatively small information costs may preclude efficient decision-making if they are still significant in comparison to the product cost. Second, there is such a thing as too much information. Form-givers may be in a Catch-22: form terms will not be enforced if there is too little information available to the form-taker, and they will not be enforced if the information provided would exceed form-takers' cognitive capacities by requiring study and comprehension beyond the constraints of rationality.[90]
Finally, advancements in information technology will not affect form-givers' desire to maintain the confidentiality of proprietary information. In the Minit-Paid hypothetical, the office manager could efficiently evaluate whether to accept the performance terms only if he knew the probability that an installation problem or multiple timekeeper problem would occur. It is highly unlikely that the licensor would willingly publish such competitor-sensitive information. Yet without such information, there is little reason to believe that transactions on form terms will be efficient.[91]
In sum, while information technology promises many advancements that will make clickwraps more user friendly than our previous images of form contracts, technology itself is unlikely to remove the basic objections to enforcement of form terms.
2. Product differentiation and complexity
A pillar of the scholarly objection to form terms is that an unrealistic presumption of assent is inconsistent with the goal of economic efficiency. In opposition, critics note that part of the "freedom of contract" mantra is that form-givers' behavior in imposing unfair terms will be restrained by competition.[92] To some extent then, the enforceability of clickwraps should turn on whether competition among form-givers will create efficient performance terms.
A crucial assumption of the economic model is that products are not differentiated, so that one producer's butter is just as acceptable to consumers at a given price as another producer's butter.[93] To the extent that there are not perfect substitutes for the form-giver's product, competition will be less than perfect. The vast majority of transactions within the scope of UCITA will involve products that are protected by patent, copyright, trademark, trade secret or other proprietary restrictions. The fact that the parties are transacting in such differentiated products means it will be much less likely that people will be able to rely on competition to make performance terms efficient. [94]
For example, it is unlikely, thanks to intellectual property laws, that Minit-Paid's program will have the same bundle of features as its competitors' products. While competing programs may be functional equivalents, their very complexity may make it impossible or impracticable for customers to quantify accurately the risks of non-performance. The risk of customer confusion will be especially severe when licensors promote the patented or copyrighted nature of their products.
This issue is not merely theoretical. Holly Towle correctly points out that a major reason virtually all licensors disclaim warranties of merchantability is that they are not comfortable with the concept of a software program's "ordinary purposes."[95] This practice does not merely reflect a distrust of the specificity of language or the competency of judges and juries. To be sure, just imagine the transaction costs of valuing warranty exposure with a complex product like a software program consisting of a million lines of code and licensed to perform functions that have not been performed by software before. Further, one must take into account valuation of the same risk for competitors' products. In short, the differentiated and complex nature of most UCITA products make it extremely unlikely that even the most methodical form-taker will be able to compute performance risk accurately.[96]
3. Clickwraps and price
A third characteristic of the typical UCITA clickwrap transaction that does not bode well for competitive performance terms is the cost structure of the typical software or information product. Designing the software program and writing software code account for a large portion of the expenses incurred in software manufacture and distribution. Licensing and distributing copies of software on disk typically represent only a few dollars of the licensing fee. Similarly, designing and populating an information database requires substantial capital, but granting access to additional users costs very little. This cost structure means that a licensor's average costs per unit will fall over almost all of its product's life. As a result, marginal cost pricing will not maximize profits.[97]
In those circumstances, there is no assurance that pricing will accurately reflect marginal costs or result in an optimal allocation of resources. This situation differs, therefore, from that assumed by those who suggest that competition will create efficient performance terms.[98] In short, those even though software and information markets may be extremely competitive in vying for investor and customer dollars, they are not sufficiently competitive in the economics sense such that performance terms will be efficient.
In sum, an argument could be advanced that general objections to form terms should not apply to clickwraps because technology allows form-givers to overcome the obstacles to efficient decision-making by form-takers. Although it is true that innovative use of linking to information could make it easier to evaluate the content and legal effect of clickwraps, it does not appear that those innovations would sufficiently eliminate the information disparities about the product itself or the cognitive limits that preclude efficient decisions. Therefore, disincentives to informed assent would continue to preclude a well-founded presumed consent doctrine, and public enforcement of form performance terms will still not be justified.
III. UCITA'S APPROACH TO FORM TERMS
There is persuasive scholarship, employing both law and economics and conceptual approaches, holding that form terms in shrinkwraps and clickwraps should not be enforced. UCITA takes the opposite approach. This Part considers UCITA's stated policy objectives and its statutory provisions to determine if it adequately responds to the objections of scholars.
A. UCITA and Assent to Form Contracts
An extensive exegesis of UCITA's treatment of performance terms is beyond the scope of this Article. It is also beside the point because form contracts will simply repeat the Act's default rules where they are in the form-taker's favor and supercede them when they are not. Therefore, this section will merely set the context with a brief explanation of UCITA's general approach to enforcement of form terms.
1. The statutory text and the Official Comments
UCITA's approach to form contracts in commercial transactions can be captured in one word (enforced) and in a description of four sections, § 102(4), § 113, § 112(a) and § 208.[99] Section 102(4) defines "agreement" as the "bargain of the parties in fact."[100] Recall that a premise of the scholarly analysis is that form terms should not be enforced because there is no basis for a finding that the parties in fact had a meeting of the minds (bargain) on form terms. A neutral reading of § 102(4) might suggest that, where there is no meeting of minds as a matter of fact, there is no "agreement" that would include the form terms.[101] The Official Comment, however, makes it clear that the definition breaks no new ground, but is merely carried over from the UCC.[102] The real purpose of the definition is to clarify that "agreement" includes not only a writing but also conduct, trade usage and courses of dealing and performance-in short, Llewellyn's realism.[103] Therefore, we should look elsewhere to determine the enforceability of form terms.
The next relevant provision, § 113(a), sets forth the Act's basic assertion of the primacy of "freedom of contract":
The effect of any provision of this [Act], including an allocation of risk or imposition of a burden, may be varied by agreement of the parties.[104]
The remainder of § 113(a) subjects the general rule numerous qualifications.[105] To the extent that those qualifications exist, the Act adopts a regulatory approach. In particular, § 113(a)(3) lists several performance terms whose content is regulated, including choice of law and forum, procedural aspects of disclaimers of implied warranties, liquidated damages, and limitations periods.[106] Neither § 113 nor its Official Comment mention form contracts; thus the search for provisions applicable to form terms continues.
A third possibly relevant section is § 112, which determines when a party manifests assent to a record or term.[107] Although § 112 is long and complex, the essence is set forth in subsection (a): assent is manifest if the party "acting with knowledge of, or after having an opportunity to review the record or term" "authenticates" it "with intent to adopt or accept it" or "intentionally engages in conduct or makes statements with reason to know" that the other party may infer assent from the conduct or statement.[108] Under § 112(d), use of the product would constitute assent.[109]
Official Comment 2 to § 112 states that the term "manifesting assent" comes from Restatement (Second) of Contracts § 19.[110] The Comment makes it clear that an "opportunity to review" the record or term is the crucial prerequisite to assent.[111] Official Comment 8 addresses the "opportunity to review" requirement in more detail, but again the intent seems to be limited to traditional contract doctrinal notions.[112] For example, the Comment states:
Common law does not clearly establish this requirement, but the requirement of an opportunity to review terms reasonably made available reflects simple fairness and establishes concepts that curtail procedural unconscionability. For a person, an opportunity to review requires that a record be made available in a manner that ought to call it to the attention of a reasonable person and permit review. This requirement is met if the person knows of the record or has reason to know that the record or term exists in a form and location that in the circumstances permit review of it or a copy of it.[113]
Most relevant to the form terms issue are Comments 3, b and d.[114] Comment b starts off with the reasonable statement that "[a]ssent occurs if a person acts or fails to act having reason to know its behavior will be viewed by the other party as indicating assent."[115] This statement could be construed as recognizing the last element of my definition of "form terms" by requiring that a form-giver have reason to believe that entering into a transaction necessarily implies consent to the form performance terms. But such an interpretation seems inconsistent with the later statement in the same Comment "[t]hat is the same rule that prevails in all other contract law."[116] The intentional conduct requirement is satisfied if the alternative of refusing to act exists, even if refusing leaves no alternative source from which the licensee may acquire the computer information.
The statements in Comment 3(d) buttress this conservative interpretation of UCITA's intent:
Assent to particular terms. This Act distinguishes between assent to a record and, when required by this Act or other law, assent to a particular term in a record. Assent to a record encompasses all terms of the record. Section 208. Assent to a particular term, if required, requires acts that specifically relate to that term. This is like a requirement that a party "initial" a clause to make it effective. One act, however, may assent to both the record and the term if the circumstances, including the language of the record, clearly indicate that this is true. . . .[117]
In short, § 112 adopts the position that assent is manifest even if the form-giver is not reasonable in inferring assent to a particular term.[118]
This brings us to § 208, the final and most deadly provision. It provides that a party "adopts" the terms of a "standard form" as the terms of the contract if the party manifests assent to the record as a whole.[119] Subsection 3 drives the nail home, stating:
If a party adopts the terms of a record, the terms become part of the contract without regard to the party's knowledge or understanding of individual terms in the record, except for a term that is unenforceable because it fails to satisfy another requirement of this [Act].[120]
Official Comment 2 is explicit about the policy underlying this rule:
Adopting Terms. A party that assents to a record adopts the record as the terms of the contract whether or not the record is a standard form. There is no difference between a customized record or terms of a standard form. Standard forms are common and provide efficiency for both parties; they are used by both licensees and licensors. Treating them as of lesser effect than other records would place commercial contract law in conflict with commercial practice.[121]
Thus, UCITA's Drafters not only reject a new look at form contracts but also recognize that the results of that decision will be more significant in the information age.[122]
It appears that the scholarship discussed in Part II failed to penetrate the drafting process, or perhaps more fairly, to persuade the drafters. [123]No reference is made to the efficiency arguments made by the scholars. We are told only that forms are "common" and "provide efficiencies for both parties."[124] While that may be true for neutral terms, such as trade usage regarding mode or timing of delivery, or for "visible" terms that merely reflect the most commonly dickered deal, it appears that the Drafters did not consider the application of these principles to the particular problem of performance terms.
2. UCITA propaganda on the enforcement issue
It is unfortunate that the Comments only state in conclusory terms why the Drafters chose to follow tradition. However, a Memorandum written by the Chair of the Drafting Committee, Carlyle Ring, and the Reporter, Raymond Nimmer, sheds additional light on the drafters' reasoning.[125] For example:
What if I don't read the contract, am I still bound by it?Yes. UCITA and common law both require only that you have a chance to read the contract, not that you actually read it. Many of us sign and accept contracts without reading them in full. This is not good practice, but it does not mean that this failure by one party can change the terms that the other party is relying on. [126]
Note the assumption in the last sentence, that the form-giver is "relying on" the enforceability of the terms. That seems to be an odd assumption for a Committee drafting a statute for a new economy, under a decade old. As the scholarly analysis proves, the reliance argument is not so cut and dry. The form-giver's position could just as well be described as a "mere hope" or in terrorem gambit. And the form-taker's position could be portrayed as reliance on the form-giver's good faith in standing behind its product, at least to the extent that a rational person would accept risk in an economy that seeks an efficient, welfare maximizing result.[127]
The Memorandum goes even further in discussing the enforceability of shrinkwraps:
But isn't it true that shrinkwrap licenses are unenforceable under current law, and UCITA will enforce them? No. Most courts hold that shrinkwraps are enforceable or simply enforce their terms without any contest of their enforceability. Some courts have invalidated such contracts where the deal was clearly closed before the shrinkwrap was presented and was not an agreed modification. However, shrinkwrap contracting is a standard method of doing business in both software and various types of hard goods. Billions of dollars of commerce rely on it. UCITA adopts, as uniform law, the position of the majority of the cases and adds procedural and substantive protections for the licensee that might be inferred, but are not made explicit in the decisions.[128]
Here, the authors look to the few reported court decisions for authority, rather than commercial practice. In other words, they are fine lawyers, instructing participants to look to the courts when they can. When the courts do not provide guidance, Ring & Nimmer switch to citing unverified and unverifiable "commercial practice."[129]
When the authors finally reach the public policy issue of enforcement, they find the answer "easy," despite the intuitive leanings of business people, judges and "spooked" academics:
Why would law want to allow such contracts [shrinkwraps]? That's easy. This is an efficient means of doing business that benefits both parties and is used in billions of dollars of transactions. A number of leading contract law scholars have written in support of similar practices in criticizing proposals in Article 2 revisions. Professor Randy Barnett, Boston University School of Law, explained why when commenting on a proposal that would have invalidated this distribution method:
Though the idea of consumers paying for goods before they examine all the terms of the agreement has spooked some academics, their concerns [do not result from] any real impairment of contractual consent. [I] speak here not only as a contracts professor who has written extensively on the importance of contractual consent, but also as a frequent consumer of such goods as electronics and software. It is not a bother in the slightest to pay for a good in a store, or on-line, and then examine the terms in the comfort of my own home provided that I can return the good should I reject the terms. To the contrary, I cannot imagine anything other than an aesthetic objection to this practice. True, consumers who dislike a term in the agreement are put to some inconvenience when they must return a good, as they would in returning any good with which they are not completely satisfied upon inspection, though even they benefit from the lower prices and more specifically tailored terms that result from the practice. But this minor inconvenience in no way warrants a frontal attack on this form of contracting on the grounds of lack of assent. There is certainly assent, though it happens after initial payment. There need no be law against that.
Professor Hal S. Scott, Harvard Law School, in a letter to the Article 2 committee, described this type of contracting as an "established retail practice of sending the full legal terms of a purchased product with the shipped product, after payment has already been made. This practice is of great value to both sellers and buyers. [Invalidating or curtailing it would create] a costly and unworkable system of contract administration. These costs will be passed on to consumers in the form of higher product costs."[130]
The answer is "easy" in part because the Drafters switch the focus to the "mass market license," which UCITA subjects to special rules that do not also apply to the mere business-to business-license.[131] The authors' conclusions, however, are subject to more fundamental objections. Note the following express policy justifications for form contracting and the equally facile responses derived from the scholarly analysis:
-Form contracts are commonly used.
-Certainly, because they are efficient for the form-giver to propose and inefficient for the form-taker to oppose.
- The concerns do not result from "an impairment of contractual consent" (apparently) because the form-taker has an unrestricted right of return.
-The return privilege is only required in mass-market transactions, which are excluded from our discussion; and the return privilege will often have lapsed before the performance problem arises.[132] Their point also shifts the issue from term consent to "contractual consent" and ignores both the efficiency and cognitive arguments against enforcement of form terms, and totally ignores the problem of performance risk.
-"There is certainly assent, though it happens after initial payment."
-This ipse dixit is classic formalism that even Williston would wink at. The proponents do not explain how the "assent" is anything more than an empty gesture communicating neither attention nor understanding. It is as though form-giver's offer had consisted of several clearly distinct phrases and several inaudible phrases. Would the form-taker's "I accept" include the latter terms?
-Requiring actual assent would impose a costly and unworkable system of contract administration.
-Not necessarily, if more even-handed risk allocations were supplied by default rules. The role of legislation should be to seek efficient outcomes where markets fail. It is beyond peradventure that form performance terms tend to be inefficient. By assuming that the only allocative system is one of contract, rather than regulation, the Drafters ignore a plausible solution.
-These costs will be passed on to consumers in the form of higher product costs.
-This statement has no empirical support, and is inconsistent even with law-and-economics reasoning. By definition, the terms are not negotiated, so there is no rational reason for the form-giver to take the benefit of risk allocation into price. Even a monopolist will not take non-cognitive (to the buyer) risk allocations into account in determining price.[133] If, as Towle asserts, licensors cannot evaluate the cost of a warranty of merchantability, how can they accurately value the savings from a disclaimer so as to effect an efficient price reduction?
So, what is one to make of UCITA's decision to grant impenetrable enforceability to conscionable form performance terms in commercial settings? Is the Conference just following the ghost of Llewellyn? Is commercial law scholarship worthless unless it validates (even in conclusory fashion) perceptions of commercial practice? Before addressing those issues in Section C of this Part, I will first look at the details of UCITA's regulation of form terms to see if perhaps its Drafters reached the right result through the wrong doors.
B. UCITA's Regulation of Performance Terms
UCITA's Reporter, Professor Nimmer, is correct in citing the importance of images in developing an efficient and just commercial law.[134] The question of the proper image, unfortunately, is far from resolved by UCITA. Karl Llewellyn's "law of the horse" gave us an image where the buyer at the town livery stable inspected the horse and purchased it.[135] In contrast, by 1970, Slawson asserted that in the modern economy customers no longer "contract," they "buy promises."[136] Slawson saw that pervasive merchandising, technological change and the prevalence of complex products with significant intangible attributes meant that customers relied much more on the seller's carefully created product "image."[137] As a result, the price paid by the customer was for the promised image and only that.
Thus, where caveat emptor made sense for the law of the horse, Slawson argued that a warranty of fitness for intended purpose should be implied in all sales involving promises, despite attempted disclaimers in form contracts.[138] Slawson's view was not, of course, ever adopted by the UCITA, whose predominant image was much more that of the toaster.[139] One might ask whether the Drafters, in devising a regime for a UCITA world, should have given more attention to the "buying promises" image. Certainly, the sophistication of electronic and information technology and the "pig-in-a-poke" nature of many UCITA transactions (where the full performance of a product may be received by customers only over significant time periods), suggest that Slawson's image be given some consideration in determining whether UCITA transactions should be treated differently from Article 2 sales of goods. If the more modern image of buying promises were adopted, we would expect to see UCITA sections regulating performance terms with a view toward ensuring that the licensor actually delivered on those promises. Even a superficial look at the Act, however, discloses that UCITA leans sharply toward the "goods" image for product performance, despite Nimmer's arguments that that image is no longer appropriate.
I have defined "performance terms" to mean those contractual provisions that define the promised (or more accurately, disclaimed) performance of the product being licensed and the remedies provided (limited) for policing that performance. Naturally, many transactors do negotiate and define performance criteria in great detail. The terms of the bargain can be reflected in a host of specifications, express and implied warranty terms and liquidated damages and similar remedies provisions. Such transactions are not fodder for this article. Rather, the image here is more akin to my initial hypotheticals where the parties have un-communicated conceptions in mind (use of the information in the Geodata case) or incomplete conceptions (the cutting and pasting problem in the Minit-Paid hypothetical). I will limit the discussion here to warranties and disclaimers and the following limitations on remedies: limitations to exclusive remedies, denial of liability for incidental and consequential damages, choice of law and choice of forum.[140]
Warranties
The discussion on warranties need not detain us long. Although UCITA, like Article 2, includes provisions on both express and implied warranties, it also permits enforceable, facile and complete exclusions of express warranties and disclaimers of implied warranties.[141] Only the form-giver's scrivener with the dullest pen will effectively fail to avoid all warranty obligations.
Moreover, even UCITA's implied warranties appear unsuitable to serve as mandatory default rules. There is no pervasive warranty of merchantability or even minimal functionality, and what does exist is carried over from the Article 2 goods image.[142]
2. Choice of law
The Act's approach to remedies is a little more regulatory, comporting with the Drafters' philosophy, given that remedies are not within the exclusive expertise of the merchant class. The result, however, is no more consistent with efficiency than the approach to contract enforceability. Unfortunately, the Official Comments are not any more persuasive.
To begin with choice of law provisions, boilerplate choice of law clauses are enforceable without qualification.[143] Official Comment 2 to § 109 explains:
Contract terms that select the law applicable to the contract are routine in commercial agreements. The information economy accentuates their importance because it allows remote parties to enter and perform contracts spanning multiple jurisdictions and in circumstances that do not depend on physical location of either party or the information. Subsection (a) enables small companies to actively engage in multinational business; if the agreement could not designate applicable law, even the smallest business would be subject to the law of all fifty States and all countries in the world. That would impose large costs and uncertainty on an otherwise efficient system of commerce; it would create barriers to entry.[144]
The first and second sentences of the Comment are irrelevant truisms. Clearly, some law is going to apply where the transaction is multi-jurisdictional, and UCITA transactions differ from non-UCITA transactions in this regard, if at all, only in that the location of the form-taker might be unknown to the form-giver. As with any form term, there are two choices: (1) The form-giver should be able to pick unilaterally the jurisdiction whose law favors it; or (2) the form-giver should bear the transactional costs of identifying the relevant legal issues, determining the risks of enforcing the form contract under the law of various jurisdictions in which the product will be offered and reflecting the increased risks if any in the product's price.[145] It is certainly not self-evident that the former is a more efficient or commercially desirable result than the latter. Yet UCITA's drafters clearly prefer the former, even urging that it should be a uniform rule.[146]
Nor is it clear that choice of law is a significant issue, even for small businesses, in deciding whether to do business either across the country or across the world.[147] It is difficult to accept that UCITA's Drafters really believe that small businesses use their precious capital to pay lawyers to evaluate legal contingencies abroad. Even if firms do so, the reduction in information costs and the exploding availability of information on foreign law suggests that UCITA vastly overstates the costs of determining choice of law risk.[148] On the other hand, if UCITA is correct and the costs are high for form-givers, form-takers would incur much higher aggregate transaction costs to evaluate the same risk. If anything about this topic is clear, it is that society will waste fewer resources if licensors investigate the risk of the application of various jurisdictions' laws to any single form contract than if their hundred or thousands of customers have to do the same. In short, UCITA's choice of law rule may make fine ideology, but it is neither fair nor efficient. [149]
3. Forum selection
UCITA's approach to forum selection clauses suffers from the same impairments. By adopting the rationale of U.S. Supreme Court admiralty cases enforcing such clauses, the Drafters again missed an opportunity to consider the proper role of a legislature in dealing with commercial efficiencies.[150] Two examples should suffice. First, § 110 states that forum clauses are enforceable unless unreasonable and unjust.[151] The only gloss on this qualification is in Official Comment 3 to § 110:
The agreed term is unenforceable if it has no valid commercial purpose and has severe and unfair impact on the other party. This may preclude enforcement of forum agreements that choose an unreasonable forum solely to defeat the other party's ability to contest disputes. Terms may be unreasonable in that they have no commercial purpose or justification and their impact may be unjust if the term unfairly harms the other party. On the other hand, an agreed choice of forum based on a valid commercial purpose is not invalid simply because it adversely affects one party, even if bargaining power was unequal. . . .
Agreed choices of forum are important in electronic commerce. Court decisions on jurisdiction in the Internet demonstrate the uncertainty about when merely doing business on the Internet exposes a party to jurisdiction in all states and all countries. That uncertainty affects all businesses, but it has greatest impact on small enterprises. Choice of forum agreements thus serve a significant commercial purpose by allowing parties to control the uncertainty and the cost that it creates.[152]
What would the Drafters have said had they directly considered the enforceability of forum clauses in form contracts? Here is one possible translation if the Drafters viewed the issue consistently with the scholarship highlighted above:
As a general rule, the states have adopted long-arm statutes that subject a contracting party to the jurisdiction of the courts in the other party's home state. This Act would override that legislative judgment at the option of a form-giver. The form-giver may substitute the courts of any other jurisdiction unless the form-taker can prove that the form-giver's choice is both unreasonable and unjust. The form-giver's decision to shift to the form-taker the cost of litigating in an inconvenient or less-advantageous forum should be enforced by the courts even though no rational form-taker would incur the costs necessary to "price" the clause and even though no form-taker would be reasonable in inferring that the form-giver's purchase of a product actually evidenced any agreement to such a clause.
Agreed forum clauses are common in transactions covered by the Act, and therefore should be enforced without consideration of the net impact on litigation outcomes or product performance. They could have three primary effects with respect to performance-related claims arising from form-based transactions: 1. The form-taker could incur the irrational cost of investigating and pricing the additional risk and the cost of pursuing its claim in the stated forum over litigating in either of the parties' home states. This approach would be inefficient from a societal perspective, and we hope it won't happen. 2. The form-taker could ignore the term until a claim arose, and then factor the additional cost into a decision whether to pursue the claim. That calculus will result in a substantial shift of resolution losses to form-takers when the forum choice increases costs or reduces potential gains from pursuit of the claim to the extent that claims are foregone rather than pursued. It therefore reduces the likelihood that form-givers will comply with performance obligations. (If no claim arose, the form-taker would reap the benefit of any price savings from such a clause. We have, however, no generally applicable empirical evidence of, or theoretical basis to predict, such pricing impacts). 3. To the extent that this state has claimant-friendly substantive or procedural law other than this Act or judges or juries who tend to favor claimants, this Section will tend to deprive the state's citizens of the benefit of those advantages.
This explanation might cause legislators to view § 110 differently from the "Official Comment." By portraying forum clauses as the status quo and products of a quid pro quo, the Comment prejudices any meaningful discussion of the merits of the issue.
The irony is that, as the Comment recognizes, forum clauses are chosen to reflect the form-giver's preference for litigation results (including transaction costs) in one jurisdiction over others. As Rakoff asks, why should that choice be enforced just because the form-giver thought of it?[153] For example, it is understandable that a form-giver located in New York or Kansas and doing business with form-takers in plaintiff-friendly states like Mississippi or Alabama would insert a forum clause. Why would a legislature in the latter states adopt § 110? Citizens of those states, which are likely to have fewer form-givers and more form-takers, would tend to suffer a net loss. Even where the differences in outcomes between the chosen forum and alternative fora are not so dramatic, at the margin the choice will result in performance defects being un-remedied. The Drafters do not suggest why this is in society's interests.
In sum, while it is one thing for the Drafters to "explain the intent" of uniform law provisions through Official Comments, it is quite another to overstate the reasons for adoption. Such failings only support claims that the drafting process is biased in favor of form-givers.[154]
4. Remedies limitations
UCITA § 803 generally follow the UCC in enforcing a form-giver's choice of or limitation on remedies.[155] On the surface, this is not surprising given the Act's stated acceptance of freedom of contract. On closer inspection, however, UCITA denotes perhaps the Drafters' greatest missed opportunity. Perhaps no area of contract law doctrine is more clearly dependent on public policy considerations than remedies. Official Comment 2 to Section 803 states that:
Parties may by agreement fit their remedies to their particular deal. This is fundamental to contract practice and defines the cost of a transaction. A party that agrees to accept all liability for breach will charge more than a party that contractually limits liability. Similarly, a party may not be willing to acquire a product unless it obtains particular remedies and recourse. How parties order these choices depends on the agreement, but no principle of commercial contract law suggests that parties' ability to control these issues should be precluded.[156]
The Drafters ability to write hyperbole in the emphasized language defies present explication. Suffice it to say, this statement clearly cannot withstand attack if the scholarly analysis is accepted as even marginally credible.
To price a remedies limitation efficiently requires detailed information about the frequency and severity of performance breakdowns, the transactional costs associated with repairing or replacing the product, and so on. Regardless of those concerns, the Comment points to a direct and universal connection between performance terms and pricing (a party will "charge more" for a product with judicial remedies for performance problems, and a product with limited remedies will cost less).[157] This is pure ideology. Even if it is assumed that remedies are factored into pricing, the minimum question is whether the price impact is efficient. There is no reason to presume that it will be efficient since the factoring is done solely by the form-giver.[158]
Form-takers, especially with respect to complex technology products, are unlikely to have the requisite technical information. The costs of obtaining such performance information, even if not proprietary, are substantial. Therefore, even negotiated contracts are likely to be inefficient in that regard.[159] In these circumstances, a primary public policy justification for judicial enforcement of private agreements dissipates.[160] If we don't have any confidence that a "contract" will result in a socially desirable outcome, why should the public's resources be used to enforce the contract?
Alternatively, even in a forms-based UCITA transaction, it might be rational for a legislature to conclude that form-giver-chosen remedies will be efficient on a second or third best basis, perhaps with a limiting principle on freedom of contract. [161]The UCITA/UCC "failure of its essential purpose" concept could be seen as such a limiting principle, but the Official Comment's explanation strongly implies that the purpose of the remedy must be determined on a "four corners" basis.[162] Since the form-giver's intent is the only reference point in a remedy analysis, it can be expected that the essential purpose test will not serve a limiting function.[163]
Another limiting principle might be a requirement that the form provide at least a fair quantum of remedy.[164] The Official Comment states, however, that courts should not recognize such a principle to invalidate remedy limitations clauses under § 803's essential purpose test. Instead, the Drafters state:
However, the essence of any contract is that parties accept the legal consequences of their deal and that there be at least a fair quantum of remedy in the event of breach. Contracts that do not do so may fail for lack of consideration or mutuality. This does not mean that a court can rewrite the agreement or the agreed remedies. If a remedy is provided and is made exclusive, the fact that it does not fully compensate the aggrieved party is not a reason to allow that party to avoid the consequences of its agreement. Remedy terms are agreed allocations of risks. For example, a contract that limits recovery for software defects used in a satellite system t o the price of the software (e.g., $100,000) is not unenforceable because the defect caused loss of a $1 million satellite. A decision to set a limit affects pricing and risk and cannot be set aside because the loss eventually fell on one party. On the other hand, a contract that states "licensee will have no responsibility for any harm to licensor caused by licensee's intentional breach of any aspect of the agreement" may lack mutuality to establish a contract.[165]
We see once again the Drafters' unfounded and dogmatic reliance on the invisible hand of pricing. The Drafters do not even blush at the fact that a remedies limitation may effectively operate as an utter denial of contractual undertaking. As the last sentence of the quoted Comment says, even proof of a declination of any responsibility for an intentional breach only gets a "may" on the enforceability scale. Form-givers who do not stand behind their products and instead hide behind boilerplate limitations can take great comfort in a Drafting Committee so blind to the problems of form contracts and the intractability of performance obligations. What happened to commercial reasonableness?
5. UCITA's "fundamental public policy" exception to enforcement
No discussion of UCITA's treatment of performance terms in commercial form contracts would be complete without mention § 105's general "out" for contract terms that violate a "fundamental public policy."[166]
The primary focus of this provision is the Drafters' recognition of the significance of First Amendment and intellectual property issues in UCITA transactions.[167] However, the text is not so limited, and the Drafters have addressed the application of § 105(b) to contract terms in Official Comment 3. While most of the focus falls on restricting information use,[168] the Comment makes clear that the Drafters intended that no public policy override will apply in the context of performance terms:
With reference to contract law policies that regulate the bargain of the parties, this Act makes express public policy choices. Contract law issues such as contract formation, creation and disclaimer of warranties, measuring and limiting damages, basic contractual obligations, contractual background rules, the effect of contractual choice, risk of loss, and the like, including the right of parties to alter the effect of the terms of this Act by their agreement should not be invalidated under subsection (b) of this section. This subsection deals with policies that implicate the broader public interest and the balance between enforcing private transactions and the need to protect the public domain of information.[169]
The Comments explicitly disclose once again the Drafters' disregard of the form transaction. The Drafters therefore, miss another opportunity to bring the Act to bear on that type of transaction most requiring legislative oversight.[170]
C. Where UCITA Went Wrong
No law - much less a uniform law on a new and still developing area of commerce - can hope to avoid criticism for its drafters' choices. Other commentators have suggested that UCITA's scope is perhaps too aggressive.[171] In trying to prescribe a legal regime for the most rapidly developing economy in history, the drafters took on a daunting task. Certainly they are to be congratulated on making the drafting process as public as practicable.[172] There is little to be gained at this point by quibbling about particular language choices or by nibbling at the edges of fundamental debates.
Nor is it the intent of the foregoing discussion to urge particular drafting changes. Instead, my purpose is to demonstrate the extent to which the Drafters' freedom of contract ideology interfered with a sound analysis that could have indicated more appropriate ways to achieve efficient results in transactions that are consensual in form but not contractual in essence. An understanding of the weakness(es) of UCITA's approach might help us identify how a legislature should deal with performance risk in form contracts. For that purpose, I believe that there are five primary reasons UCITA fails to address this issue in the proper fashion.
1. The Drafters were too certain about certainty
Perhaps the problem lay in the Drafters' rush to provide "certainty" of contract enforceability in the belief that reduction of contracting risk was necessary for market development. The explosive growth of the information economy during the drafting and legislative process has demonstrated that the need for certainty was highly overrated, and it is highly unlikely that contractual certainty was ever very high on the wish lists of CEOs (as opposed to General Counsel). After all, in markets where the value of information and technology has a half-life of months or days, and where the disputes rage over the very existence of property rights in the subjects of commerce, judicial enforcement of contracts is of subsidiary interest. Moreover, enforceability of the full range of contract law issues, as attempted by UCITA, was unnecessary to the attainment of the core interests of UCITA proponents. Whether form shrinkwrap and clickwrap agreements are enforceable with respect to all terms was not an issue that really needed to be addressed. Nor should it have been addressed, unless it would have received careful analysis. As the foregoing discussion demonstrates, the issues of form contracts and performance problems are complex and cannot be resolved with a mere nod of the head to "duty to read." The Drafters should not have taken on the chore unless they were going to do it justice.
2. UCITA is too doctrinal and not commercial enough
Despite their just admiration for Article 2 and Lex Llewellyn, it appears that the Drafters may have ended up paying mostly lip service to Llewellyn's conviction that commercial legislation must start with commerce and not judicial doctrine. Obviously speculation, but where the Drafters may have gotten off the wagon is in their failing to appreciate the significance a crucial difference between their drafting assignment and Llewellyn's. The law of the sales of goods had been long and hotly litigated for decades. The task of Article 2's drafters was more to achieve uniformity through codification - selecting the rules from precedents more consistent with then current commercial practices - than it was to create a law that would become uniform.[173] The focus on judicial resolution of litigated disputes was therefore understandable. In a sense, the task and process was not that much different from that facing the American Law Institute in promulgating a Restatement of Contracts.
In contrast, UCITA deals with transactions of very recent vintage and with types of products that were not even in existence when the drafting process started.[174] The dearth of case decisions demonstrates that the problem is one of Statement, not Restatement. The Drafters could have started with paradigm and process, ignoring judicial contract doctrine.[175] They could have drafted a law that built on the structure of current commercial transactions, a statute that prescribed how to make those transactions (more) efficient and fair. Instead, they appear to have interpreted too literally Gilmore's admonition "Be accurate rather than logical" to mean that they should be neither bold nor proactive when it came to performance risk. As a result, UCITA's structure at least in regard to performance risk appears to be derived more from judicial constructs than from commercial realities.
To phrase the point somewhat differently, the Drafters became too conservative in replicating the Article 2 approach across all issues. By adopting the "duty to read" myth that predated even the formalists, the Drafters missed an opportunity to solve the conundrum of a rational contracting process (form contracts unread and resultingly unappreciated by form-takers) that produces irrational economic results (risks borne by those who cannot efficiently bear them). They properly recognized that the old Article 2 paradigm did not work for some issues, but they did not identify all the paradigms that needed to be addressed. Negotiated transactions and undocumented transactions, while important, do not present the same issues as form transactions; defining the appropriate scope for licenses and reconciling tensions between federal intellectual property law and state contract law do not present the same issues as allocating performance risk.
3. The information age changed the paradigm form transaction
To over-generalize: in the Article 2 era commercial practices were such that the paradigm transactions were un- or under-documented. Agreements were made through phone calls and telegrams or through the now infamous battles of forms. Gap fillers were important in that context to deal with subjects the parties did not address or as to which their forms conflicted. Shrinkwraps and clickwraps differ in that there is seldom a gap in the comprehensiveness of the terms provided to the form-taker, no real issue of disclosure in terms of time or format, and no responsive form used by the form-taker.[176] UCITA form transactions tend to be truly unilateral contracts accepted by payment or use. The real issue is not whether to provide gap-fillers in terms when neither party addressed an issue, but whether to enforce terms that were apparently undisputed ex ante.
The central form/performance risk problem has never been one of "manifest assent" based on "opportunity to review," but of the form-taker's inability to price the performance terms. That distinction becomes increasingly significant as we move from the Article 2 image of mass-produced fungible goods and established distribution in a mature "production" economy[177] to the UCITA image of highly differentiated, intangible products and services (in which the contract may be the only method of defining the product) with revolutionary distribution channels and an information economy.[178]
Nonetheless, the Drafters still portray the issue as one of disclosure and assent. By looking at an economically irrelevant issue, rather than the increasingly significant one of risk allocation, the Drafters failed to mind their own counsel of keeping attuned to the significance of change.
4. UCITA overvalues the pro-form-giver normative argument
One cannot read the Official Comments quoted above without sensing the Drafters' acceptance of the merits of a caveat emptor approach to assent, not just to the transaction but to each purported term of the transaction. It can be inferred that the Drafters accept the conventional wisdom that (1) forms should bind form-takers just because they can read them; and (2) enforcing forms is efficient and causes the economy to grow in desirable ways. I have already cited scholarship contesting the validity of that wisdom from the perspective of efficiency. In fairness to the Drafters though, the issue in their eyes might not be efficiency alone: their concern could also be one of personal responsibility and fairness. Commercial customers know the risk of dealing with form contracts. They know the forms favor (even unreasonably) the form-giver. No one forced parties to transact, and therefore they shouldn't get a second bite - and if they do, the costs will be shifted back to "innocent" customers, shareholders and employees.[179]
The weakness in such a statement of the issue is that it rests on the notion that the form-giver is in fact relying on the enforceability of the terms in setting price. That reliance cannot be established simply by the presence of a term in the form, because any capable drafter would include one-sided terms even if they were unlikely to be enforced. Inclusion of such terms has long been standard practice, either as bargaining chips with sophisticated form-takers or in terrorem with unsophisticated parties. How else can one explain the presence in standard forms of terms that are clearly unenforceable under controlling law?[180]
Even if we assume that the form-giver has relied on the form-taker's participation in the transaction as demonstrating assent, it is difficult to portray that reliance as reasonable, and therefore, justified in the sight of a legislature. For, as discussed above, the form-giver must be charged with knowledge that the risks allocated by the term cannot be rationally priced by the form-taker. Therefore, the only one who can determine if the risk was shifted for value (taken into account in price) is the form-giver. Can we say that a form-giver is justifiably relying on an assumption of assent just because it sees customers accepting, without negotiation or question, a deal too good to the form-giver to be true? I think not. There is a difference between justified reliance and wishful thinking.
The most fundamental reason for rejecting the caveat emptor approach is that it can apply only where the background law creates the possibility of reliance. Assume that form-givers have priced their products on the basis of form risk allocations condoned by current law. As a result, under the common agreement exemplified in the Minit-Paid hypothetical, the form-giver has assumed no performance risk, and the price should include no amount to compensate Minit-Paid for any risk. After all, a form-giver could reasonably conclude that under UCITA it would be liable only for replacement of a CD and that if the form-taker wanted a greater remedy, it would have to arbitrate in the form-giver's chosen state before the general counsel of the Society for the Advancement of Licensors.
The wonder of legislation, however, is that it changes the law, and therefore the result, prospectively. If UCITA had adopted a mandatory scheme for resolving performance problems in form transactions, the reliance argument would have disappeared. Once form-givers knew that forms would not be enforced, they would change their pricing and/or other terms.[181] Since the legislation would apply to all market participants, fairness, certainty and efficiency would be assured.[182] UCITA's failure, then, is in judging form-takers too harshly for failing to comply with a normative obligation that UCITA could change without harm to the market or the parties. Whether such a mandatory scheme could be devised is the subject of Part IV.
5. Certainty and remedies don't mix
UCITA rightfully prizes certainty as a transactional value. The goal of certainty, however, can only be accorded to certain aspects of performance. UCITA can achieve certainty by holding that a non-performing party always wins if it has incorporated the right exculpatory terms into its form. I have parted ways with UCITA on that score. But once the need for remedy arises, UCITA and I walk the same path. Valuation of contract rights is amazingly speculative because so many different types of transactions fall within UCITA's scope and because those transactions involve new and developing technologies, marketing concepts and markets. To suggest, as UCITA does in § 809, that damages can be measured through applications of statutory formula is exceedingly optimistic.[183] The sheer number of indeterminate concepts employed in the formulas makes it wholly impractical to price the option (or effect) of breach in advance. The Drafters wisely anticipated this in providing in § 809(a)(1)(C), a catchall that damages may be calculated "in any reasonable manner."[184]
In short, UCITA could achieve certainty for all aspects of a transaction only if form terms deprive non-defaulting parties of remedies. If remedies are available, neither transactor can determine, prior to the transaction, the actual cost of non-performance with any more certainty than currently exists (a "reasonable" amount of damages based on "my" view of value based on a complex scenario of possible defaults). This is not to criticize UCITA's remedial provisions. Rather, the point reinforces my earlier conclusion that UCITA's goal of certainty of obligation through the enforcement of form terms will not achieve efficiency when the valuation of remedies is so uncertain.
IV. ALLOCATING PERFORMANCE RISK THROUGH LEGISLATION
If I have made any point by now, it is that freedom of contract is not the answer to the performance risk problem in UCITA transactions. Now comes the hard part: If not freedom of contract, then what? After all, the appeal of "freedom of contract" as a basis for drafting a commercial law statute is that it allows the drafter to avoid the hard questions. The drafters could say that "the parties will know more about that than we do and they're in the best position to deal with the specifics."[185] The trouble with form transactions, however, is that the parties, at least one of them anyway, do not know the true facts and will not seek them out.
Earlier I chided UCITA for paying insufficient attention to preeminent scholarship. In fairness perhaps, the reason that scholarship was not accorded more consideration is that the scholars failed to propose alternative solutions that were consistent with the overall UCC/UCITA approach. For example, Slawson, who looks at the problem using an administrative law analog, argues in general terms that adhesion contracts should be enforced only to the extent that the terms are in the public interest.[186] Rakoff more specifically argues that courts should resort to gap fillers unless the form-giver can demonstrate that the form term is commercially reasonable.[187] Eisenberg is less expansive regarding the implications of his analysis. It appears that he is primarily concerned with the form-giver's right to enforce the defective "bargain" and would limit recovery to restitution. With respect to the rights of the form-taker, he is less explicit and would perhaps resort to default rules.[188]
It is not surprising that the proposals of these critics are so general - they did not limit their scope to a specific type of transaction, particular contractual term or transaction risk. As the complexity of UCITA (necessary as it may be) demonstrates, it is simply not feasible to address a full range of potential positive law rules in a single article. Thus, I am not surprised that it appears that no one has yet developed a coherent, complete and commercially practical solution to the allocation of performance risk. This section of the article lays out one possible solution to the performance risk problem.
A. Paramaters for Performance Legislation
Part IV.B sets forth a statutory proposal, complete with Official Comment. Certainly, it would be unfair to criticize UCITA's Drafters for failing to do what I cannot do myself. Before proceeding to attack the problem, however, it will be worthwhile to identify some parameters of a successful solution. There are at least seven:
1. Transactional reality v. contract and tort concepts
The transactions at issue are voluntary but not contractual. Since the premise of my analysis is the absence of efficient assent, any solution should not import contract doctrine without careful consideration of its applicability to the commercial context. Thus, there is no reason to begin (or end) an analysis of who should bear the risk of consequential damages with the centuries old rule of Hadley v. Baxendale or with the variations on its theme wrought by hundreds of subsequent decisions.
Similarly, it is not helpful to incorporate concepts or terms from tort law. Even though courts may divide the commercial world into property, tort and contracts fiefdoms, there is no tenet of logic or policies that require legislatures to do the same. As with precepts of contract law and terminology, words and rubrics familiar in tort law may provide comfort and efficiency of transmission, but only at an undue cost in qualification and confusion. For example, the "economic loss doctrine" holds that a claimant should be limited to a contract claim (in lieu of a tort cause of action) unless the defendant's wrong has caused bodily injury or property damage to property other than the subject of the transaction.[189] That doctrine, however, is rooted in attempts to separate tort law from contract law for reasons that make little doctrinal and only chimerical practical sense.[190]
In short, the analysis should look through contract and tort labels to the nature of the transaction and the parties' respective positions. A legislative solution should recognize that the distribution of performance risk should be imposed not on the basis of consent but on the following societal judgments:
1. Performance risk should not fall randomly on licensees.
2. Producers are in the best position to avoid or spread the risk that their products will not perform as expected.
3. Licensees cannot efficiently evaluate performance risk except (possibly) through price.
4. Producers should therefore to the extent possible price their products to reflect and spread that risk. [191]
5. To the extent that markets are not sufficiently competitive to cause price to reflect performance risk, the law should impose performance loss on the party that can best manage that risk through avoidance or spreading.
6. Because 1 through 5 lead to outcomes that are preferable to those unilaterally dictated by forms, product markets will tend to be more efficient.
Thus, the guiding principle should not be what the parties would have done if they had thought about it, or what most parties would do if left free to bargain. Neither of those principles is helpful because they require a fiction that the parties could or ever would so think or bargain. Because we do not live in a world of fairly perfect information or minimal transaction costs, the legislative solution should be the one that would be chosen if the transaction were telescoped into a single party decision.[192]
If the two commercial transactors (licensor and licensee) were one (i.e. both the producer and user, so that the producer were developing a product for its own internal use), how would it manage the risk of performance? It would invest in product design and testing until the marginal return from that investment exceeded the marginal out-of-pocket and consequential damages from performance failures. We might expect in such a case that the licensor would design terms to avoid problems that occurred often enough or were serious enough that the anticipated (or experienced) aggregate cost of failures during performance were greater than the marginal design costs. A statutory solution should seek the same objective.
2. Regulation v. paternalism
The appropriate level of governmental involvement in commercial transactions has been an enduring debate among scholars and law reformers. UCITA's approach is to allow the parties to make their own rules except in rare situations where UCITA regulates certain terms of the transaction.[193] UCITA also recognizes that states have a certain paternalistic interest with respect to consumers, but it defers to other statutes to assert those interests.[194]
While I have rejected the application of the UCITA approach to forms/performance problems, I do not disagree with the Drafters' reluctance to set detailed rules in concrete for commercial transactors. Even form-takers should be able to assume stupid performance risks. On the other hand, as the scholarly analysis demonstrates, both contract doctrine and the public interest in an efficient and fair economy require knowing assent to the imposition of contingent performance terms. The rub, then, is to recognize legitimate autonomy without allowing the economic forces that cause form transactions to overwhelm that autonomy. Form-givers should be able to avoid the mandatory rules only if the form-taker's waiver is both knowing and priced in some fashion.
One approach would be for the rules to be mandatory unless the parties' agreement provided otherwise. To avoid the problem of the "otherwise" being an invisible term in a form, the statute would have to require, at a minimum, that the form-taker separately manifest assent to the waiver of the statutory benefit. Experience with statutory waivers in other commercial forms, such as loan guaranties, demonstrates that even conspicuous language and separate initialing seldom produce knowing assent. This obstacle is even more apparent in the shrinkwrap/clickwrap context, where not even the formality of a signature is required.
Thus, if the mandatory rules are to be generally applicable except in the case of the priced waiver, the burden should be on the form-giver to demonstrate that the waiver was made in a transaction where the public policy for the statute did not apply. To make that showing, the form-giver should be required to show that the transaction was not a form transaction, but commercially negotiated to the extent that the form-giver could reasonably rely on the purported assent to the waiver. The form-giver should be free to make the showing as it sees fit. In addition, to provide some degree of certainty, the statute should include a safe harbor pursuant to which the form-giver would know that the mandatory provisions did not apply. The model statute proposes two such safe harbor provisions: (1) separate pricing for the two options, one on form terms and one on default rule terms;[195] and (2) provision of a disclosure package on performance risks designed to level the playing field.
UCITA fans may argue that this approach is pure paternalism in the commercial context since merchants should have the obligation to know the commercial law, to understand their rights under the mandatory rules and, foolishly, to waive them if they wish. That argument, however, is just a re-cast of the duty to read. The purpose of the statute is simply to deal with the disparity in transactional costs and the information gaps that make form transactions particularly inefficient with respect to performance terms. The proposal may not be a first best solution, but it is designed to provide at least a modicum of fairness in performance terms.
3. Performance obligation v. performance remedy
The doctrinal issue. Contractarians often distinguish between the obligation of a contract and the remedy for breach of that obligation.[196] In commercial law, the formula is generally implied warranty, breach, causation and remedy. Judges and commentators, however, criticize warranty as a useless label; "implied warranty" generally leads to the conclusion that liability should follow.[197] A statutory solution should therefore avoid such common law baggage where possible.
Nevertheless, even though doctrinal distinctions can impede proper analysis, a forceful argument can be made that a fair and efficient performance law must take into account the difference between what a product is supposed to do (the performance standard) and what should happen when it does not do it (the performance remedy). This would seem to be a non-controversial point, except that able commentators have argued that the most basic form of performance obligation, the implied warranty of merchantability, is so vague as to be meaningless.[198] Indeed, the fuzziness of the classic standard of "fit for the ordinary purposes" is cited as a primary reason form-givers invariably disclaim the warranty.[199] In other words, the argument is that the disclaimer is not just the form-giver's ipse dixit, but a commercially reasonable element of risk management in an arm's length deal.[200]
If it were really the case, as it may well be, that a performance standard or obligation cannot be specified ex ante, then perhaps it makes no sense to distinguish performance obligation from performance remedy. The legislative drafter would then be left with three primary options:
(1) Stick with the tried and true but - based on its universal disclaimer, failed - merchantability standard;
(2) Adopt an absolute quality standard that makes the producer virtually strictly liable for performance problems; or
(3) Skip the separate obligation analysis and focus on the remedy question: "which party should bear the loss?"
The choice depends on what is most consistent with commercial reality and therefore with the premise of Lex Llewellyn and UCITA.
Defining a meaningful obligation. As one who has struggled with drafting warranty clauses, either to placate customer demands or in response to vendor reticence, I am doubtful that there is much to be gained from attempting to improve legislatively on the UCC/UCITA formula if one accepts their drafters' premise that the performance obligation can be captured in a sentence or two compiled from judicial decisions.[201] The differences in products and transactions are simply too great to develop a meaningful ex ante standard.[202]
The second option, absolute enterprise liability, presents conceptual and political difficulties. As to the former, Croley and Hanson conclude that the consumer products liability debate, which essentially presents a contest between contract and tort law perspectives, would be resolved best through an absolute enterprise liability rule.[203] Although their reasoning is largely consistent with my approach here, I am not convinced that what makes sense for the Article 2 economy will also make sense for the UCITA economy. First, the potential for end-user error in UCITA transactions seems intuitively higher than in the goods economy, given the relative complexity and rate of change of the technology involved. Second, UCITA products tend to be used as components in large systems (for example, software in networks) to a greater extent than in the old automobile or toaster world, thereby creating a higher probability of causation and remedies issues. Both of these factors suggest that the legislative solution should incorporate a more explicit risk management approach than Croley and Hanson seem to anticipate. In any event, given the point where UCITA currently rests, it seems clear that an absolute liability rule would be politically impracticable, even if it were conceptually preferable.
The third alternative, merging the obligation and loss concepts, is also unsatisfactory because it tends to confuse analysis. The size of the loss influences the performance and causation issues, and skews the parties' perceptions of a "case value" in the event of litigation. The licensor views a small loss as not resulting from a performance problem, and the licensee views a large loss as increasing the likelihood that a fact finder will covertly find a performance flaw where there was none.[204]
As a result, the model statute adopts a fourth approach. Obligation is distinguished from remedy; what constitutes a failure to meet the obligation is also defined, or at least described, in some detail. Moreover, the determination of a failure combines both ex ante and ex post perspectives. I base this approach on my experience in dealing with contracting issues in UCITA transactions. The commercial reality is that even the licensor's business people - as opposed to their lawyer/scriveners - pay little heed to "theoretical" performance ex ante.[205] They know what they intend to sell/license and know much about the actual performance characteristics of their products. These vendors, however, are also honest and sophisticated enough to know that UCITA transactions are as much art as science, and that performance issues are best considered in light of the facts of a particular situation.
Thus, it is not uncommon to hear licensors respond to questions regarding performance obligation (and remedy, for that matter) to the effect that "while we can't guarantee requested performance or minimal risk, we are confident that it won't be a problem. If it is, we'll work with you to eliminate the difficulty."[206] After all, ex ante the parties' cooperative relationship with respect to performance dominates the potential adversarial positions that might develop if performance problems ensue and cannot be satisfactorily resolved.
This dynamic occurs equally in the form, non-negotiated transaction where the parties dicker, if at all, on price, scope of license and other non-contingent terms. There is little reason to believe that business people would address the performance issues any differently in form cases if they were in fact raised.[207] Therefore, legislation that views the nature and scope of the ex ante performance "obligation" through the lens of the actual, ex post dispute will provide a more informed approach. In fact, such an approach may be indistinguishable from "majoritarian" default rule methodology, which seeks to determine what most of the parties in the market would have agreed on if the point had been fully negotiated.[208] Therefore the model statute attempts to balance the flexible nature of commercial practices with a sufficiently clear set of rules to permit reasonable guidance and planning.
Unique remedies for performance problems. It does not necessarily follow that remedial issues must or should be separately analyzed from the same before and after perspectives. The reasonable drafter could conclude that the appropriate remedies for breach in a form-based UCITA transaction should not vary with the nature of the obligation (e.g., delivery date v. product performance): no matter what, the measure of recovery should be the innocent party's benefit of the bargain, less reasonable mitigation, plus any reasonably foreseeable consequential damages. This contractarian view, however, ignores the commercial reality that parties opt out of traditional judicial formulae in negotiated transactions. The fully negotiated commercial approach assigns potential losses to the party that can most cheaply bear or shift the loss, thereby maximizing the total return from the transaction. This is, of course, the same result the law-and-economists believe will produce the most efficient result.[209] Any statutory solution should incorporate those truths by adopting rules that take into account the exceptionally difficult issues of causation and valuation that arise when products do not perform to expectations.
4. Risk management v. loss allocation
The last point made in the discussion of remedies - that commercial practice and economic theory assign the burdens of performance failure to the party who can bear the loss most cheaply - applies generally to performance risk. It therefore deserves further attention. Any substantive rule of performance obligation and remedy must allocate performance risk on the basis, not of ability to bear loss (whether measured in terms of supposed resources or wealth or in terms of "bargaining power"), but on ability to manage risk. [210] The differences between the two concepts are numerous. For example, the former tends to favor result-driven conclusions, while the latter considers what the parties could have done before, during and after the transaction. The former also tends to favor covert judicial decision-making and therefore reduces the likelihood of efficient dispute resolution. Most important, the latter best reflects commercial practices by encouraging each party to forego speculative and opportunistic behavior and to take reasonable measures to reduce both the frequency and the severity of loss.
To be sure, UCITA's performance terms include gap-fillers that reflect this distinction.[211] Many of those rules, however, are derived from common law and tend to distort risk management incentives, and those that can be avoided by the licensor under § 113 and § 208.[212] Legislation eviscerates the salutary nature of a risk management approach if it merely codifies judicial doctrine and enforces form deviations from the gap-filler provisions. A form-giver who denies the form-taker's right to recover consequential damages evades any obligation to minimize the risk of those damages (such as by providing early notice of a potential problem in performance as a result of complaints from other customers) even though the cost of that mitigation would be minimal. Therefore, with respect to each element of a proposed legislative solution, the question persists whether the statute would provide the necessary incentive to each party to minimize overall performance risk and loss.
5. Determining whether the performance standard was met: risk of failure and risk of loss
As mentioned in item IV.A.3 supra, the distinctions between ex post and ex ante perspectives are also relevant in determining whether a product failed to meet a performance standard.[213] Law-and-economists have demonstrated that the failure to appreciate the differences can lead to inefficient outcomes.[214] The problem has significant practical implications. In litigated contract disputes, it is virtually impossible for judge or jury to divorce the analysis of "what could have been" from what actually happened. Empirical research has shown that people overestimate the probability that an event could occur when they are aware of an outcome in which the event did occur.[215] Thus, a statutory solution that ignores the ex post situation invites covert decision-making. For both these conceptual and practical reasons, a drafter must consider how perspective affects evaluations of risk of performance failure, probability of loss and reliability of result.
The challenge to a legislative drafter is in developing an approach that recognizes the ex post perspective but controls it so that it does not distort analysis as to whether the performance obligation was satisfied. This challenge would seem particularly difficult since I have already argued that the only commercially practicable test for determining the content of the performance obligation includes a heavy ex post component.[216]
The present problem is subtly different from that discussed in item 3 above. The issue here is not whether the facts of the particular case can inform the decision on the product's performance standard. It is how to determine whether that standard was met through a proper balance between the licensee's interest in a perfectly operating product and the licensor's interest in producing a commercially viable product at a marketable price. That balance can be achieved only by a solution that takes into account three components:
(1)what the product was expected to do (the performance standard);
(2)the likelihood of a failure to meet that expectation (the performance risk); and
(3)the range of potential injury to the customer (the risk of loss).
Perhaps an illustration drawn from the data entry (cut and paste) aspect of the Minit-Paid hypothetical can make this point more understandable and distinguish it from the previous discussion about the definition of the performance standard. The question discussed under item 3 was whether the particular programming issue - transfer of data from another program - should be taken into account in determining the performance standard to which Minit-Paid will be held. In other words, do we phrase the issue as simply "Did the licensor have an obligation (in contractarian terms, make a warranty) to provide a program that would compile multi-timekeeper billings?" Or is the issue "Did the licensor have an obligation to provide a program that compiled billings when the information was transferred from another program?"[217]
Unless we can understand the "why" of the alleged failure, we cannot provide a commercially reasonable or reliable performance standard. To phrase the question in the more general terms as initially stated is to invite the answer, "Of course," since the most prominent feature of the program was the combined time-keeping and billing function. To phrase the question in the more specific latter terms is to focus the issue on the facts of this case, thereby permitting a focused evaluation of the pros and cons of design and user expectations.
In contrast, the issue under discussion in this item is whether the now-defined performance standard has been breached. This issue considers (1) the likelihood that the failure to design the software to accept information formatted by another program would substantially interfere with the product's performance;[218] and (2) the impact of the problem on the customer's ability to obtain the benefits of the program. Each of these must be evaluated in light of the commercial expectations of the parties regarding the utility of the program and the effects of dysfunction on the licensee's business. The emphasis here must be on objective standards and material harms. Because UCITA transactions often involve relatively low cost products that support costly and revenue-sensitive operations of licensees,[219] we must be careful to limit redress for performance problems to those that really matter. Otherwise, trivial defects could subject the licensor to unduly severe damages.
This approach may seem cumbersome, but reliable results can be achieved only if the solution identifies these discrete components of risk management.
6. Insurance v. collateral source rule
Both tort and contract law pay faint heed to the availability of insurance as a means of risk management.[220] When insurance, however, is a commercially acceptable means of spreading risk, the legislative solution should take insurance into account in fashioning the remedy for performance dysfunction.
The insurance industry has been slow to respond to the changing risk profile of the Information Economy.[221] Traditional liability coverages apply only to claims for bodily injury and tangible property damage.[222] Traditional first party (property and business interruption) insurance generally cover only physical injury to tangible property.[223] Neither type of coverage enables a licensor or licensee to shift most UCITA product performance losses.[224] Even today, there are relatively few broadly available coverages that permit transactors to shift performance risk through insurance.[225] Given the growth of the information economy, insurers will soon be forced to come into step.
Suffice it to say for present purposes that any legislative solution should recognize that generally available insurance is an excellent indicator of the ability of a party to bear performance risk. The transactor who can purchase insurance on the most favorable terms should generally be viewed as the most efficient bearer of performance risk.
7. Commercial behavior v. litigation posturing: reducing transaction costs
Two primary benefits that form terms give form-givers is that they deter the prosecution of claims for defective performance and substantially reduce the net cost of defeating claims that are pursued.[226] Any legislative solution of the forms/performance problem should seek to reduce the transactional costs of resolving disputes while discouraging commercially unreasonable behavior, especially litigation.[227]
Because a legislative solution should not adopt the UCITA freedom of contract ideology, the statute should not be limited to prescribing gap-fillers.[228] Instead, it should favor licensors who provide prompt and competent product support,[229] and transactors who subscribe to commercially reasonable dispute resolution procedures.[230] The goal is to reduce transaction costs for both sides and avoid the current debacle where customer frustration over unreasonable risk allocation leads to wasteful class action suits.
B. A Model "Performance Risk" Statute
What follows is a model statute that could be incorporated into UCITA to address the forms/performance problem. My hubris is not so great that I think this is a match for the years and talent involved in drafting UCITA, but we have to start somewhere. Although many of the terms are vague and imprecise, they are all terms are to be construed in a commercially reasonable manner such that business people, if not lawyers, can give the statute sufficient content to get on with their lives.
THE MODEL PERFORMANCE RISK AMENDMENTS TO UCITA
An Act in Five Parts
The Uniform Computer Information Transactions Act ("UCITA") is hereby amended as follows:
SECTION 1. DEFINITIONS.
(a) The definition of "standard form" in §102(60) of UCITA shall be revised to read:
(60) "Standard form" means a record or a group of related records containing terms prepared by a form-giver for repeated use in transactions and so used in a transaction in which there was no negotiated change of performance terms and in which the form-giver has no reasonable expectation that the form-taker would incur the costs necessary to evaluate the benefits or costs of those performance terms.
(b) The following definitions shall be added to Section 102 of UCITA:
"Form-giver" means a person who tenders a standard form in a transaction subject to this Act.
"Form-taker" means a person to whom a standard form is tendered in a transaction subject to this Act.
"Form transaction" means a transaction to which this Act applies (other than a mass-market transaction or a transaction as to which a consumer is a party) in which performance terms would, but for Sections 619 and 817, have included, pursuant to Section 208, the terms in a standard form. "Form transaction" shall not include any transaction in which (a) the form-giver has provided to the form-taker an opportunity to enter into a transaction for the same product, at a different price, on terms that do not modify or disclaim performance terms that would apply but for §113 of this Act, or (b) the form-giver has provided to the form-taker information sufficient to permit a reasonable form-taker to evaluate the material benefits and costs of the performance terms included in the standard form.
"Performance obligations" are the obligations created by Section 619.
"Performance remedies" are the remedies provided in Section 817.
"Performance terms" means any contract term that purports to establish, modify or disclaim (a) any obligation of a licensor with respect to any characteristic or standard of functionality or performance of the product licensed in the transaction, including without limitation warranties, or disclaimers of warranties or (b) any remedy for breach of any such obligation.
"Product" includes computer information, informational rights in computer information, services, and goods.
OFFICIAL COMMENT.[231]
1. The definition of "standard form" differs from the current definition in UCITA in two primary ways. The first is that UCITA permits negotiation only on specified terms; if the parties go beyond those aspects, for example to negotiate a longer notice period before acceptance or waiver of rights, then the entire contract falls out of the definition. The proposed definition focuses on the non-negotiated nature of specific terms or issues. Second, and more importantly, the UCITA definition does not include that element that the form-giver must not have a reasonable expectation that the form-taker can and will evaluate the term accurately. This element is necessary to properly limit special treatment of form terms to those transactions where the information costs and obstacles are sufficient to create a high probability that there will not be an efficient bargain.
2. "Form-giver" can refer to either a licensor or licensee. In general, forms used by licensors will seek to minimize product performance obligations, and forms used by licensees will seek to expand those beyond the requirements of background law. The information asymmetries that affect form transactions generally favor licensors and other owners of technology and computer information. However, this Act establishes mandatory obligations and remedies which cannot be changed by either party as long as the transaction is a form transaction. See Section 711 (d).
3. "Form transaction" includes only transactions between commercial parties. The exclusion for mass-market transactions recognizes that the greater or at least special protections granted to commercial licensees in those transactions stem from practical and political concerns. The designation of a transaction as a form transaction does not involve an evaluation of relative bargaining power or sophistication. Even multinational licensees dealing with sole proprietor licensors can enter into form transactions if the requirements of the definition are satisfied.
4. In general, form-givers cannot obtain a waiver of statutory benefits. However, a putative form-giver can always demonstrate that the transaction did not involve a "standard form" as defined in Section 102(60). See Section 711. Moreover, two safe harbors are provided for transactions where the reasons for the statute do not exist. "Form transaction" is defined to exclude those terms. As to the price option, courts should not consider the fairness of the spread between the two options. The provision should be construed liberally since channeling risk into price is recognized as a most efficient means of information about risk. As to the disclosure option, the adequacy of disclosure is a question for the court and is to be evaluated in light of the complexity of the product, the nature of the typical user and value of the transaction compared to the costs of assimilating the information disclosed.
5. "Product" is intended to cover all forms of consideration transferred by the licensor. Although the typical "product" in a UCITA transaction is software or other "information" as defined in § 102(35), the fact that UCITA covers some mixed transactions which include goods and services requires a broad term to give the performance obligation appropriate scope.
SECTION 2. PERFORMANCE OBLIGATIONS.
The following is added to UCITA as Section 619:
SECTION 619. PERFORMANCE OBLIGATIONS IN FORM TRANSACTIONS.
(a) Sections 105 through 113, Parts 2, 3, 4 (except Section 402) of this Act do not apply to performance obligations in form transactions. The provisions of Parts 6, 7 and 8 of this Act apply to performance obligations or performance remedies except to the extent inconsistent with the provisions of this Section or Sections 711 or 817.[232]
(b) No term or condition of any standard form shall be enforceable except to the extent that the standard form provides additional rights or remedies in favor of a person who is not the form-giver, over those otherwise provided in this Act. Nothing in this section shall affect any express warranties made as part of a form transaction.
(c) The licensor in a form transaction has an obligation to the initial end-user of any product that is the subject of the transaction to cause the product to perform in accordance with the following standards:
The product:
Complies, subject to Sections 406 and 407, with the warranties set forth in Sections 401, 403, 404 and 405 to the extent the transaction would fall within the terms of those sections but for subsection (a) of this Section;
Is fit for the ordinary purposes for which products of like kind and quality are used;
Performs the functions which products of like kind and quality perform;
Substantially conforms in all material respects to all marketing materials used by the licensor in distribution of the product; and
Substantially complies with all material statements made in documentation provided as part of transaction.
All material installation, access or use instructions are substantially accurate and reasonably adequate to permit the end-user to obtain the ordinary benefits of the transaction in a timely manner.
OFFICIAL COMMENT.
1. UCITA is a "contract statute," and form transactions are not contracts in the sense of informed meetings of the minds. Therefore, those provisions of UCITA that set forth rules of formation and construction of contracts are generally deemed inapplicable to the performance obligations and performance remedies of form transactions as defined in this statute.
2. For the same reason, the purported "contract" between the parties is deemed ineffective. Instead the statute adopts a mandatory set of performance standards and remedial provisions.
3. The Section's use of the term "obligation" is intended to be neutral between contract and tort principles because neither set of doctrinal rules provides a satisfactory commercial solution to the problem. The obligation runs only from the licensor because it is the licensor, and not intermediary participants in the distribution channel, who is in the best position to control performance risk. The obligation runs only to the initial end-user because it is between the initial end-user and the licensor that the transactional cost and information asymmetries are most significant. This section does not affect the rights and obligations of other distributors and users.
4. The performance obligation established by subsection 2 is defined in part by the "majoritarian" gap-filling provisions of Part 4 of UCITA and partly by the recognition that the majoritarian philosophy, founded as it is on assumptions of informational and transactional cost symmetry, do not apply to form transactions. This amendment departs, therefore, from UCITA's general reluctance to impose on a transactor an obligation to provide a commercially reasonable, or even a minimal, level of performance. The implied warranties therefore represent only one aspect of the obligation imposed in form transactions.
5. Regardless of the type of transaction involved, subsections (c)(1)(ii) and (iii) impose an obligation to conform to the purposes and functions of products of like kind and quality. As usual, price will be a primary indicator of "like kind and quality." Channels of distribution and marketing will also be relevant in determining the relevant set of comparable products. UCITA transactions frequently involve products protected by intellectual property laws and unique products using new technology, so it may be difficult for users to identify close substitutes. This should not be deemed to mean that there is no product of "like kind and quality." The test is "like," not identical. As long as a product can be identified that performs similar enough functions to permit the court to identify a commercially reasonable expectation of performance, the standards of clauses (i) and (ii) of subsection should be applied.
6. Marketing materials, product manuals and other documentation are key sources for a determination of the transactor's own views as to product performance. Specifications and standards included in such materials should generally be included in establishing the standard of performance under this section. On the other hand, disclaimers in documentation should generally not be given effect because the same disclaimers included in a standard form would be unenforceable under subsection (b).
7. Subsection (c)(2) requires that documentation be substantially accurate and reasonably adequate to permit the end-user to enjoy the benefits of the transaction. Installation and integration with other systems and products can cause significant transaction costs. More complete documentation will generally reduce such costs, and the obligation imposed is consistent with the fact that the licensor can bear those educational costs more cheaply than individual users with respect to ordinary uses. The standards are not perfect accuracy and exhaustive detail - the test should be whether the reasonably sophisticated typical user can use the instructions readily to gain the benefits of the transaction. On the other hand, users are in a better position to determine the requirements of unusual or specialized uses, and the subsection leaves the burden of those costs on the end-user.
SECTION 3. BREACH OF PERFORMANCE OBLIGATION.
The following is added to UCITA as Section 711:
SECTION 711. BREACH OF PERFORMANCE OBLIGATION IN FORM TRANSACTIONS.
(a) A licensor in a form transaction shall be liable, to the extent and in the manner provided in Section 817, to the initial end-user of the product for all damages incurred by the end-user as a result of the licensor's failure to perform an obligation under Section 619. The end-user shall have the burden of proving a failure to perform.
(b) Whether a product failed to satisfy the performance standards set forth in Section 619 shall be determined by taking into account all material facts and circumstances involved in design, production, marketing and distribution of the product as well as in the particular end-user's use of the product, the specific causes of the alleged failure and the useful life of the product. Unless the context otherwise requires, the following factors shall be taken into account in determining whether an attribute of a product failed to satisfy a performance standard:
The extent to which the attribute at issue is integral to the product;
The frequency and severity of losses that users could reasonably be expected to incur as a result of the attribute's failure;
The effect on utility and cost of the product if the product had been designed not to include the attribute or to avoid the failure; and
The extent to which an end-user could avoid or mitigate the alleged failure or loss through commercially reasonable maintenance, up-grades or changes in practices or procedures to accommodate the attribute.
A licensor in a form transaction shall not be liable under subsection (a) with respect to the failure of any product to meet a performance standard set forth in Section 619 if:
The licensor proves that the failure was caused by misuse of the product or by the failure of the initial end-user to use reasonable commercial practices in the installation or operation of the product;
The licensor proves that the end-user could have avoided the failure or resulting loss through commercially reasonable maintenance, service or other precautionary measures; or
With respect to a failure of the product to perform in accordance with Section 619(c)(1)(iii):
(i) The licensor proves that the failure did not occur during statistically significant "beta testing" of the product in accordance with reasonable commercial practices in the licensor's trade or industry; or
(ii) The licensor proves, by means of product performance records which it maintains in the ordinary course of its business and which reflect the date, source, and nature of customer complaints regarding the product at issue, that the failure has occurred, as of the date of trial, with respect to the product in less than .01% of the transactions involving the product entered into by the licensor.
(d) The provisions of Sections 619, 711 and 817 cannot be waived by either a licensor or an end-user in a form transaction.
OFFICIAL COMMENT.
1. This Section adopts a risk management approach to determining responsibility for performance failures. Subsection 1 sets forth the general rule that the licensor is liable for all damages incurred. However, the liability extends only to the initial end-user for the reasons stated in connection with Section 619.
2. Subsection (b) sets forth the elements to be considered in evaluating whether a product failed the performance standards of Section 619. The statute places the risk of product performance on the party that can most cheaply identify, evaluate and mitigate that risk. Therefore, both the licensor and the end-user are required to do what is commercially reasonable to avoid loss. However, the primary onus for risk management is on the licensor because the licensor has the most knowledge about the product and is in the best position to spread losses among all users by incorporating the risk into the product's price.
3. Recognizing that the performance obligation under 619, the standard of breach under this section, and the remedy for breach are all more advantageous to users than under existing law, subsection (c) provides several controlling principles to limit the risk that liability will be imposed for defects that have their source in the end-user's practices or in defects that the licensor could not anticipate or avoid through risk management. Clause (3) is intended to recognize the commercial best practices of beta testing and customer complaint tracking. The defense is limited to questions of functionality because other failures (such as accuracy of information in a database) are not as likely to be discoverable through use or testing. The results of beta testing must have been statistically significant or it cannot be inferred that the failure is not inherent in the product. The percentage of consumer complaints that is used in subsection (c)(3)(ii) as a surrogate for an abnormal result reflects a practical resolution rather than a theoretical approach since the premise of informational asymmetries could justify imposing liability for a problem of infinitesimal probability. The limitation to the status of complaints as of time of trial obviously limits the availability of the defense, but is deemed necessary to preclude application to the user who is one of the first to identify the problem.
SECTION 4. REMEDIES
The following is added to UCITA as Section 817:
SECTION 817. REMEDIES FOR BREACH OF PERFORMANCE OBLIGATION IN FORM TRANSACTIONS.
A licensor who is liable to an initial end-user under Section 711 shall compensate the end-user with damages in the following amounts:
(1) "Benefit of the bargain" damages as follows:
(i) If the end-user rejected or revoked acceptance of the product and procured a substitute product, the difference between the cost to the end-user of that substitute product and the product's price, adjusted for differences in the products and the terms of the licenses; or
(ii) If the end-user has retained the right to use the product, the lesser of:
(A) the difference between the price paid by the user for the product and the reasonable value of the product; or
(B) the cost incurred by the end-user to cause the product to comply with the standards of Section 619; or
(C) the cost to the licensor of repairing or replacing the product, but only if the licensor offered that service to its customers and only if the service would have been reasonably provided to the end-user at no cost and without commercially unreasonable conditions; and
(2) The actual marginal expenses of the end-user incurred because employees, equipment or other resources were idled as a result of the licensor's failure to perform its obligation under Section 619; and
(3) The actual marginal costs of the end-user in (i) re-entering into a computer information lost or damaged as a result of the failure and (ii) re-creating products lost or damaged as a result of the failure; provided that costs of re-creating information obtained from third parties shall not be included under this paragraph; and
(4) Any other incidental and consequential damages, including costs of restoration and lost productivity and profits not recoverable under subparagraphs (2) and (3), provided that damages awarded pursuant to this subparagraph shall not exceed ten times the price paid by the end-user for the product.
There shall be offset against the amount due under subsection (a) the proceeds of any amount received by the end-user from any insurer or any other third party with respect to the performance failure.
An action to recover any amount due under this section may be brought in the jurisdiction where the transaction was completed, where the end-user resides or has its principal place of business or where the form-giver has its principal place of business. The forum will apply the law of the state of the end-user. No terms of any standard form used in the transaction will be admissible for any purpose except to establish the definition of the product, the term of any express warranty, the price and effective period of the license and any other term actually negotiated by the parties.
In any action described in subsection (c), the court shall abstain from a decision if an alternative dispute resolution process approved by _________________ [appropriate state agency or judicial body] pursuant to this section is available to the parties.
In any action described in subsection (c), the court shall not permit discovery until the parties, represented by senior executives with authority to resolve the dispute, have submitted the dispute to an independent third party mediator; provided that the court may order limited discovery if necessary to enable the end-user to adequately counter the form-giver's mediation positions.
The provisions of Part 8 (other than Section 807(c) through (e), Section 810 and Section 811) do not apply to claims under this Section.
OFFICIAL COMMENT.
1. This section replaces the more general damages provisions of Section 807 and 809 of the Act with respect to a licensee's damages for performance failures. The rules are neutral with respect to the contractual gap-filling provisions of Parts 6 and 7, which will generally apply because Section 619(2) precludes form terms from varying the provisions of the Act.
2. The rationale for the comprehensive remedies provided in this section is threefold. First, the true societal cost of a product includes both the costs of its production and the costs of the losses that the product causes when used as directed. Second, it is neither efficient nor fair that isolated users absorb randomly occurring losses attributable to the product. Third, those who release a product into the channels of commerce are in the best position to reduce product risk to an efficient level and to spread the remaining risk through product pricing.
3. Nevertheless, the remedies are limited to amounts that the end-user could not avoid through resort to commercially reasonable alternatives. Users should use technical support services and repair and replacement facilities provided by licensors as long as the transactional costs of doing so are not outweighed by continuing loss as a result of failure to cover. Licensors have a reasonable expectation that users will resort to those alternatives but only when the services are provided seasonably and competently in light of the user's circumstances.
4. The products in UCITA transactions are frequently relatively inexpensive compared to the value of the transactions in which they are key elements. Thus the license fee for a specialized medical database with millions of items of information may be less than $50 per access, yet a performance failure might cause a mistreatment that could result in a potential loss of millions of dollars. There has therefore been great concern that imposing consequential damages in UCITA transactions will ruin small competitors, stifle innovation and raise barriers to entry. While those concerns are valid to a limited extent, an absolute bar on risk spreading through liability for consequential damages results in an over-investment in products whose prices do not reflect their true marginal costs and results in an undue disadvantage for higher quality products competing with lower quality products. The approach taken in subsection (a) requires producers to bear those consequential damages that are more attributable to the product failure than to the nature of the user's business.
5. Subsections (a)(2)-(4) modify the traditional Hadley v. Baxendale rule and impose liability for a relatively narrow category of consequential damages. This resolution is a compromise since it is clear, for the reasons described in comment 4, that several types of consequential damages from product failure are reasonably foreseeable. Direct costs, such as downtime and costs of re-entry of lost data, can be recovered without limitation in amount but are limited to actual marginal costs. More remote losses are limited by subsection (a)(4) to ten times product price. The latter limitation is not likely to be unduly burdensome to licensors. The total amount is too small to affect claims regarding low priced, high volume products, and custom products are often not distributed through form transactions. The subsection therefore represents an attempt to reflect the true costs of product performance. The impact of the subsection on licensors is further ameliorated by the fact that any insurance benefits payable to the user are deducted under subsection (b) from the damages recoverable against the licensor. Users will still have an incentive to mitigate consequential damages risk and to obtain business interruption insurance as the insurance markets begin to make insurance available for intangible losses.
6. Subsections (c)-(e) are intended to reduce (1) user transaction costs by establishing a mandatory rule that users can pursue performance claims in their local courts and under their local laws and (2) dispute resolution costs through alternative dispute resolution and discouragement of strike-type suits where costs of discovery distort case value. Leading licensors and licensees have commenced initiatives to provide fair and low cost venues to resolve performance disputes. Subsection (d) is intended to encourage such efforts by providing for governmental evaluation and endorsement.
SECTION 5. EFFECTIVE DATE:
This Act shall become effective ______________ (at least 1 year off).
OFFICIAL COMMENT.
1. This Act substantially changes existing law. To the extent that licensors have relied on the enforceability of standard forms, their reliance may have been incorporated into pricing and product design decisions. The delayed effective date provided in this section will allow time for parties and markets to adjust to the new rules.
* * *
C. Comparing the Resolution of Hypothetical Disputes Under UCITA and the Model Statute
How will cases be resolved under the model statute compared to the resolution under UCITA? Perhaps the best way to examine a statute is to apply it to a fairly typical situation. In this section, I will apply the two statutes to the hypotheticals introduced at the beginning of this article.
1. The software hypothetical.
UCITA. To begin with the time-keeping software hypothetical involving Minit-Paid and HOE Engineers, HOE's claim is that performance defects caused its network to crash on installation and that the program would not compile invoices for multiple timekeeper files. The answer under UCITA is straightforward. The clickwrap is enforced according to its terms since I posited that there is no basis for a finding of unconscionability or conflict with fundamental public policy. Therefore, HOE has no remedy. It does not need a new CD, it has no right to incidental or consequential damages and it has no claim that would suggest that exercising its right to arbitrate would be worthwhile. It cannot even attempt to exert pressure on Minit-Paid by complaining on Internet bulletin boards or in chatrooms because those complaints could amount to a "negative" review of the product and open HOE to breach of contract claims.
The Model statute. The shrinkwrap/clickwrap qualifies as a "standard form" and the grant of the license as a "form transaction." Therefore, Minit-Paid has a performance obligation under § 619, and the shrinkwrap and clickwrap are unenforceable under § 619(b). So far, the analysis seems unproblematic.
The next step is to determine Minit-Paid's performance obligation.[233] There are two alleged defects - the installation problem and the invoice compilation problem. As to the former, HOE would argue that § 619(c)(1)(iii) establishes a standard that any software program must install without causing a system crash. HOE will also claim that the installation standard of § 619(c)(2) requires the documentation be sufficient to permit installation without a crash.
As to the compilation defect, HOE would argue that § 619(c)(1)(ii) sets a standard that the ordinary purpose of multi-user time-keeping and billing software is to compile a unified bill.[234] Second, HOE would argue that the program did not perform the functions similar to other programs of like kind and quality under clause (iii). Third, it might argue that the marketing materials stated that the program would compile the billing statements "automatically", thereby failing the substantial conformance standard of clause (iv). Finally, it might seek to prove that the user's manual disclosed nothing about restrictions on entry of time data, thereby creating a failure of full and accurate documentation under § 619(c)(2).
Once the relevant standards are identified, HOE has the burden under § 711 to prove that Minit-Paid failed to perform its obligation to cause the program to meet those standards. The highest hurdles for HOE on the installation claim will be causation under § 711(a) and whether the software documentation should have included information to enable the installer to avoid the cause of the crash, such as deletion of other software on HOE's servers or other computers. Tucked within that issue will undoubtedly be product design issues (i.e. should Minit-Paid have designed around such potential conflicts?) that will require recourse to the risk management factors of § 711(b).[235]
With respect to the compilation defect, the primary issue will be the application of the risk management factors of § 711(b). Certainly the transfer function between time keeping and invoice creation were integral to the product under subsection (b)(1). Whether the frequency/severity element is satisfied will turn on how reasonable it would be to expect cutting and pasting from other programs. It would seem HOE has an uphill fight on that issue, but discovery might produce examples of similar customer complaints. If HOE demonstrates that the risk of losses was sufficient to require a design change, it will have to prove that the change would not have substantially interfered with the utility and cost of the product. Given the prevalence of the cutting and pasting exercise, it would not seem - to this non-engineer - to have been difficult for Minit-Paid to have programmed around the problem. Finally, clause (4), which looks to the licensee's ability to avoid the risk, would not apply here unless there was some basis to argue that HOE should have tested the program early in its first billing cycle. In sum, HOE would have reasonable arguments that the program failed to meet the performance standards.
The next step is to determine whether the licensor has any defenses under § 711(c). The two relevant clauses, absent the necessary facts to support findings under clause (3), are clauses (1) and (2), which preclude recovery if the licensee does not follow reasonable commercial practices. Clause (1) applies to the installation claim because it requires that the installer take commercially reasonable precautions in installing the program. Clauses (1) and (2) apply to the compilation claim. However, HOE's cutting and pasting technique, while perhaps inefficient, does not seem commercially unreasonable. On the other hand, clause (2) would appear to preclude HOE from recovering for any losses after the problem was first discovered since the glitch is easily solved.
It therefore appears that HOE would qualify to recover its costs under § 817. HOE's expenses for getting the system up after the installation crash would be covered under § 817(a)(3) and its losses arising from its inability to use the system would be covered by § 817(a)(2).
The analysis of damages for the compilation defect is straightforward. The nature of the glitch should not require Minit-Paid to pay "benefit of the bargain" damages under § 817(a)(1)(ii) because the commercially reasonable fix is simply to change the method of input and there is no reason to believe that the new method will be more significantly more expensive than cutting and pasting.[236] Section 817, however, would allow HOE to recoup its costs of re-entering the data for the past bills (§ 817(a)(3)), and to recover the cost of manually compiling the billing statements and lost interest on delayed billings up to ten times the limit on consequential damages (§ 817(a)(4)). This result is consistent with the statute's premise that transactional costs from even relatively minor performance failures should be borne by the licensor as the most efficient risk avoider/spreader.
Of course, the burden of that liability on licensors will be mitigated by the fact that transaction costs of pursuing the claim will generally obviate nuisance claims for small amounts.[237] Even though the statute generally reduces claim resolution costs in comparison to current practice, resolution costs will still be significant in relation to damages except in severe cases, so it is still unlikely that parties will pursue uneconomic litigation.[238] Therefore, the most significant effect of the statute may not be a change in the results or amount of litigation. Instead, the most important impact might be the fact that the statute changes the status quo or background acceptable result.[239] Recognizing an enforceable right for performance defects regardless of form terms will in itself shift the relative ex post bargaining positions of the parties as they work out business accommodations to performance problems.[240] That is a result the UCITA drafters should applaud.
2. The information access hypothetical.
UCITA. Under UCITA jargon, the Geodata-Spacechain license is an "access contract" as defined in § 102(1), and the maps and associated data are "information," "informational content" and "published informational content."[241] Regardless, the form license agreement is enforceable according to its terms, so all implied warranties are disclaimed and parol evidence of the conversations about enforcement cannot be admitted.[242] Therefore, Spacechain has no claim unless Geodata breached the contract specification regarding reliability.
First, assume that Spacechain is confident it could survive summary judgment on the reliability issue. Spacechain's confidence, however, does not matter. The contractual limitation on remedies to a six-month extension of access at no additional cost will be enforced,[243] and its claims for incidental damages and consequential damages fall to the limitations in the form.[244] Spacechain, like HOE, is out of luck under UCITA.
Model statute. Not surprisingly, the issues are different under the model statute. The form contract is vaporized since Geodata could not show that the exceptions apply. The mere fact that Spacechain's representative asked about the boilerplate is not enough to take the transaction out of the statute's grasp. The application of the statute turns on actual negotiation and the form-giver's reasonable expectation that the user will incur the required information and transaction costs. Neither element is satisfied here.
Therefore, Spacechain can recover for the bad information if it can establish a failure to comply with the product standard. The "product" consists of the maps and layered data. The performance failure is that the data is inaccurate because it has been superceded. Geodata knew that the data was superceded, but believed that the changes would be immaterial.
Spacechain's burden is to demonstrate under § 711 that the product failed to meet Geodata's obligation to provide a product whose performance satisfied the standard of § 619. There are several possibilities. The first is an express warranty, available under § 619(b).[245] The specification in the form license that Geodata has deemed information in the database reliable could be construed to constitute an express warranty under UCITA § 402. The UCITA subsection applicable only to published informational content is virtually inscrutable, and Official Comment 8 on the issue is prolix.[246] For present purposes, it suffices that Comment 8 appears to recognize that express contractual commitments will be enforced. Thus, Spacechain would have a reasonable argument that Geodata's use of information that it knew was outdated violated the reliability term of the form. [247]
If that argument is unpersuasive, Spacechain could look to the "implied warranty" standards under § 619(c), of which (c)(1)(ii) looks to be the most promising.[248] Under that standard, Spacechain will have to provide evidence on products of like kind and quality. Following the Official Comment to § 619, Spacechain is not limited to geographic information products but to other products offering similar informational components. The question under (c)(ii) then is whether such dated or superceded information is generally usable.[249] Under clauses (iv) and (v), Spacechain would also want to review marketing materials (whether or not relied on by Spacechain) and product documentation to see if they mention anything about currency of database content.
Section 619(c)(2) might also establish a performance standard in this case. Because reliability caused concern sufficient to warrant its inclusion, a reasonably persuasive argument can be made that Geodata should include information relevant to reliability - even though it were held not to "warrant" reliability - so that users could make their own determinations and use the product appropriately. Here it would not be imposing too great a burden on Geodata to require it to provide reliability data (such as source and date) on a webpage for this purpose. Its failure to do so could be deemed to violate this standard because without that information the use documentation is not reasonably complete.
Assuming that the performance standards of § 619 required either that the database include the most recent Michigan data or that adequate information be provided to allow users to determine reliability, we turn to § 711 to determine if Geodata's product failed those standards. We turn first to the balancing test of § 711(b). I will focus solely on the four "mandatory" elements of the test. Under clause (1), the question is the extent to which the reliability (currency) of the Michigan data was integral. Given the license fee, the nature of the user and the mention of reliability in the form, it would seem that reliability of information is integral if the product were to perform any useful function.
Clauses (2) and (3) require evaluation of the ex ante risks of harm from licensing the database with the outdated data. Clause (2) looks to the risk that significant losses could occur, while clause (3) requires an evaluation of how avoiding the problem would affect cost and utility. On the one hand, we have the apparent testimony of Geodata's staff that they did not think that that the data made much difference. Nothing in the statute, however, indicates that a party's good faith beliefs matter; the tests of the two clauses are objective. On the other hand, we have Spacechain's claim that the difference in success ratios points to the currency problem as a cause. Correlation alone does not establish cause and effect.
Both of these points need to be demonstrated. Certainly the differences in the old and new records might have been insignificant, but it does not appear that either Geodata or its licensees viewed Geodata as an evaluator of information. If they had, then the price would have reflected that value, which would make it more difficult, not less, for Geodata to argue that the product performed as it should have. Thus, we are left with the fact that it is reasonable to forecast significant reliance losses by users from outdated information in a database, and Geodata loses the risk management test. The frequency of the risk (each time the Michigan information is accessed) and the severity (costs incurred in reliance) outweigh the cost of avoidance (either paying the increased fee for the current data and incorporating the cost into price or disclaiming currency).
That takes us to the fourth element - the extent to which users could avoid the losses. Here Geodata is in a much better position to identify currency issues at a lower cost. Individual users would have to otherwise check each underlying source for currency before each use. Since the appeal of a database like Geodata's is to avoid such costs, clause (4) suggests that Geodata failed to meet its obligation.
In sum, it appears that the risk of loss from the performance problem Spacechain experienced could have been more efficiently avoided (or spread among all users) by Geodata. Therefore, Geodata should be held to be liable under § 711.
The next issue under the statute is the measure of damages Spacechain can recover under § 817. This is another case, like the Minit-Paid case, where the damage is done before the user discovers the problem and where the user can avoid further consequential loss by seeking recourse to another source for the defective data. This case differs, however, in that the user cannot fix the performance problem, even though it may desire to keep the license in effect. Here, as in many access contracts, the pricing will provide a ready partial answer: the license with respect to the Michigan database is essentially worthless to Spacechain and the license fee can in effect be so adjusted under § 817(a)(1)(ii)(A).
Those damages do not necessarily compensate Spacechain for all diminution in value since the remaining individual databases are worth less to Spacechain as they cannot be used for Michigan. In other words, if Spacechain computed the value of the license by allocating the costs of general databases on a per-state basis, it does not get the benefit of its bargain. On the other hand, there is no indication that Geodata priced access in that fashion, so it might not be unfair to hold Spacechain to Geodata's pricing. There appears to be no logical way to resolve this inconsistency - but recall that Geodata's marginal cost of permitting Spacechain access to one more database is essentially zero. Since Geodata failed to meet its performance obligation and is not out-of-pocket even if the access price to other databases is reduced pro rata, the statute (§ 817(a)(1)(ii)(A)) provides a remedy that reflects Spacechain's probable pricing decision. This result better achieves the remedial purpose of risk allocation by causing licensors to account for risks in their pricing.
In addition to "benefit of the bargain" damages, Spacechain is entitled to its consequential damages under § 817(a)(3) and (4). Spacechain might seek the costs of creating new market analyses, the cost of new mailings and lost revenue up to the statutory limit. Inclusion of all these elements, however, would constitute double recovery. Spacechain will receive the benefit of its bargain, either by receiving the revenue it would have obtained if the database had met the performance standard (measured by the consequential damages) or by receiving damages equal to its out of pocket costs incurred in making new mailings. It cannot receive damages based on both the failed mailing and a substitute mailing.
3. A look in the rearview mirror: how the model statute would have applied to Y2K
American firms spent billions of dollars "remediating" the Year 2000 dating problem.[250] The situation might be viewed as the largest performance risk in history, given the dire predictions that flooded the media for eighteen months before the dreaded date change. Despite the amounts spent on remediation, litigation over the liabilities of software developers has been minimal.[251] Most observers, other than the lawyers who expected a litigation feeding frenzy, probably prefer to leave the remediation expense where it lies. In any case, it is fair to ask how the model statute would have responded to the situation.
Had the model statute been in effect, the economic loss rule would not have deterred potential claims.[252] Therefore, the initial inclination of counsel to software licensees may have been to assert claims for defective software. That being said, claims under the model statute would run into two significant obstacles for older software. First, if two-digit date coding were state of the art programming at the time of development, the program would have met the performance standards set forth in § 619(c)(1). Second, licensees would have to show under § 711 that licensors could more cheaply manage the risk of extended life of software whose utility spanned decades.
As the price of memory dropped and the century change approached, the decisions of programmers to continue to market software with two-digit dating invited consideration of performance risk. Note that this is not to say that a two-digit product would be per se defective. The licensor developing products during the transition period could address the issue through disclosure that the program had two digit coding and might require revision or replacement before 2000.
The other significant deterrent to performance claims is placement of risk-control responsibility on licensees under § 711(b) and (c). As Y2K gained notoriety, sophisticated licensees also became aware of it. Many of the remediation costs incurred resulted from integration of software into complex systems. Individual licensors are not in a position to control those decisions, although they do have a duty to address common integration issues as the previous hypotheticals have demonstrated.
Thus, the model statute's answer to Y2K is similar to the answer to the forgoing hypotheticals: the answer is not as black-and-white as existing law would have it, but it does offer the benefits of a scalpel to UCITA's bludgeon. Licensors who maintained state of the art practices and provided licensees with notice of the potential problem as the state of the art began to change would be protected. Licensees would have every incentive to stay abreast of design developments that affected complex systems and to remediate those systems before disaster hit. On the other hand, developers who licensed programs with two-digit dating in the mid-1990s might have been liable for not having programmed around the problem or disclosing it. We might have seen more litigation, but litigation is preferable to the current state where licensees have been forced to pass on to their customers Y2K costs that perhaps should have been spread among the shareholders and customers of the software developers.
In sum, the model statute provides a much more detailed analysis of the transactors' rights and remedies. The use of risk management criteria subjects the process and outcome subject to many variables. The generality of the damages provisions - like UCITA's - makes determination of a precise measure of damages less than certain. The squishiness of the statute is, however, one of its virtues. Drafters cannot encapsulate commercial practices and decision-making in formulas, especially given the broad spectrum of UCITA transactions and our lack of experience with market-cum-juridical outcomes. If we value commercial practicality, it must come at the expense of the certainty lawyers desire but business people seldom expect. Llewellyn could expect no more.
D. A Semi-Final Word on Performance, Risk, and Innovation
Technological innovation is to be encouraged, but not at any cost. One only has to consider the long-term costs to the environment that resulted from encouraging innovation in our manufacturing economy from the 1930s to the 1970s without requiring manufacturers to internalize the external costs of their production. By not encouraging producers to build all product costs (including the costs of losses due to performance defects) into pricing, we are only encouraging innovation for the sake of innovation.[253]
As the practice of beta testing software has proven, no one is served by releasing products for distribution before acceptable levels of functionality are established. The model statute is flexible enough to allow each producer to determine when the lure of profit justifies incurring the risk of liability for performance defects. Thus, the model statute can be said to chill innovation only to the extent that responsible risk management is viewed as less than socially optimal behavior. The model statute adopts the philosophy of our tax-and-spend economists:[254] if society desires faster innovation, the more efficient approach would be to tax computer information licensees generally and subsidize innovation, rather than allow the losses of premature products to fall randomly on users.
UCITA does not require licensors to manage, spread or even communicate performance risk. The model statute is designed to provide commercial transactors with a framework to make product choices based on their risk tolerance. By improving the accuracy of those decisions, the model should encourage, not chill, optimal innovation.
V. CONCLUSION
Americans have a love-hate relationship with technology. We become quickly dependent on each new technological advance and grow impatient for the next development. Computers and software now provide us with the ability to access, produce and process information in ways and in timeframes undreamed of just a few years ago. Nothing is more frustrating, however, than a computer crash or the retrieval of a megabyte of irrelevant data by a search engine.
In this context, it is not surprising that licensors seek to rush new products to market and cope as best they can with disappointed customers. In a sense this is a realm where the law should have feared to tread. UCITA's promulgation makes that option as obsolete as a vacuum tube or a punch card. If a statute regulates the full breadth of "computer information transactions," it should provide a workable solution for the form contract performance puzzle. A statute that seals itself with the Llewellyn coat of arms as a true believer in commercial reality should immerse itself in that reality and not flee to formalistic fictions of freedom of contract, assent and duty to read. Scholarship - and not just of the law and economics variety - can shed light on how markets and contractors really function. Clearly, they do not function well in a world of complex information asymmetries and high transaction costs.
This article has followed the scholarship of Slawson, Rakoff, Eisenberg and numerous others in demonstrating that the use of standard forms in UCITA transactions does not lead to efficient or fair transactions, at least insofar as they purport to assign risks of product performance. Enforcement of shrinkwrap and clickwrap agreements premised on the form-taker's failure to comply with the duty to read is nothing more than a fiction masking a subsidy to inefficient form-givers. UCITA is a marvelous work of draftsmanship, but its provisions and comment regarding standard forms and contingent performance terms cannot withstand the strains placed on it by economic theory and commercial reality. To put it bluntly yet with all due respect, the UCITA solution leads to form-giver tyranny rather than bilateral freedom of contract.
Neither traditional contract nor tort principles satisfactorily deal with these issues. Legislation must therefore recognize the necessity of imposing risk management obligations on the transactor who is in the best position to avoid or spread the costs of product failure. Similarly, traditional remedies and resolution procedures must give way to rules that provide transactors the incentive to minimize transaction costs and aggregate losses.
Rather than simply describe what could be done with a different legislative approach, I have burdened the reader with a model statute designed to amend UCITA, complete with ersatz Official Comments. Whether it is a cure worse than the disease I leave to you. But there is little doubt that a uniform and comprehensive law that acquiesces in the random imposition of losses with the excuse that "that's what the parties bargained for" cannot be defended on economic or normative grounds.