Summary Contents
Has this happened to you? You plunk down a pretty penny for the latest and greatest software, speed back to your computer, tear open the box, shove the CDROM into the computer, click on “install” and, after scrolling past a license agreement which would take at least fifteen minutes to read, find yourself staring at the following dialog box: “I agree.” Do you click on the box? You probably do not agree in your heart of hearts, but you click anyway, not about to let some pesky legalese delay the moment for which you’ve been waiting. Is that “clickwrap” license agreement enforceable?
Standardized end user and consumer license agreements have a bad reputation. Professor Margaret Jane Radin associates them with “democratic degradation.”Footnote 1 Chief Justice John Roberts has admitted that he does not read them (see Section 17.3, Note 2). But, for better or worse, these maligned instruments have become a part of US law that is not likely to disappear entirely in the foreseeable future. As a result, it is worth spending some time to understand the contours and ramifications of these ubiquitous contractual documents.
This chapter reviews the development of consumer license agreements through their three principal phases of development: the paper “shrinkwrap” agreements that accompanied packaged software, the electronic “clickwrap” or “click-through” agreements that emerged with the popularization of the Internet, and the even more amorphous “browsewrap” agreements that seemingly bind users to website terms and other contractual commitments without any affirmative indication of assent. The principal issue in the discussion and cases that follow is contract formation – is a valid and enforceable contract formed under the various circumstances that are described? Once a contract is found to exist, then routine principles of contract interpretation that are described elsewhere in this book apply.
17.1 Shrinkwrap Licenses
Beginning in the 1980s, the computer software industry had to contend with a question that had not previously been asked: how to license valuable intellectual property (IP) to thousands, if not millions, of consumer software users in an efficient and effective manner. Clearly, it would not be possible to execute a signed license agreement with every consumer who purchased a diskette containing a computer game or utility. Nor did the industry want to rely solely on copyright law to protect their products, as the book and magazine publishing industries had done for centuries. Though a consumer who purchased a copy of a software program on magnetic tape, a diskette or a hard drive might “own” that copy, it should obtain no rights, by implication or otherwise, to exercise rights in the manufacturer’s copyright. While the copyright laws prevented purchasers of books from illegally photocopying and distributing them, photocopying a book took a lot of effort – more than the average consumer would be willing to expend for a relatively modest payoff. Computer software, on the other hand, could be copied and redistributed with the click of a button. In the view of the industry, more robust protection than the law provided was needed (more on this below).
The answer that the industry arrived at was the “shrinkwrap” license agreement, a paper license agreement affixed to the package in which a software program was sold, visible through the clear plastic shrinkwrap surrounding the package. The consumer’s assent to the terms of the license was evidenced by her tearing open the package and using the software within. One of the first legal tests of this licensing structure came in the now-seminal ProCD case.
86 F.3d 1447 (7th Cir. 1996)
EASTERBROOK, CIRCUIT JUDGE
Must buyers of computer software obey the terms of shrinkwrap licenses? The district court held not, for two reasons: first, they are not contracts because the licenses are inside the box rather than printed on the outside; second, federal law forbids enforcement even if the licenses are contracts. The parties and numerous amici curiae have briefed many other issues, but these are the only two that matter – and we disagree with the district judge’s conclusion on each. Shrinkwrap licenses are enforceable unless their terms are objectionable on grounds applicable to contracts in general (for example, if they violate a rule of positive law, or if they are unconscionable).
ProCD, the plaintiff, has compiled information from more than 3,000 telephone directories into a computer database. We may assume that this database cannot be copyrighted, although it is more complex, contains more information (nine-digit zip codes and census industrial codes), is organized differently, and therefore is more original than the single alphabetical directory at issue in Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991). ProCD sells a version of the database, called SelectPhone, on CD-ROM discs. (CD-ROM means “compact disc – read only memory.” The “shrinkwrap license” gets its name from the fact that retail software packages are covered in plastic or cellophane “shrinkwrap,” and some vendors, though not ProCD, have written licenses that become effective as soon as the customer tears the wrapping from the package. Vendors prefer “end user license,” but we use the more common term.) A proprietary method of compressing the data serves as effective encryption too. Customers decrypt and use the data with the aid of an application program that ProCD has written. This program, which is copyrighted, searches the database in response to users’ criteria (such as “find all people named Tatum in Tennessee, plus all firms with ‘Door Systems’ in the corporate name”). The resulting lists (or, as ProCD prefers, “listings”) can be read and manipulated by other software, such as word processing programs.
ProCD’s SelectPhone product (c.1996).
The database in SelectPhone cost more than $10 million to compile and is expensive to keep current. It is much more valuable to some users than to others. The combination of names, addresses, and SIC codes enables manufacturers to compile lists of potential customers. Manufacturers and retailers pay high prices to specialized information intermediaries for such mailing lists; ProCD offers a potentially cheaper alternative. People with nothing to sell could use the database as a substitute for calling long distance information, or as a way to look up old friends who have moved to unknown towns, or just as an electronic substitute for the local phone book. ProCD decided to engage in price discrimination, selling its database to the general public for personal use at a low price (approximately $150 for the set of five discs) while selling information to the trade for a higher price. It has adopted some intermediate strategies too: access to the SelectPhone (trademark) database is available via the America Online service for the price America Online charges to its clients (approximately $3 per hour), but this service has been tailored to be useful only to the general public.
If ProCD had to recover all of its costs and make a profit by charging a single price – that is, if it could not charge more to commercial users than to the general public – it would have to raise the price substantially over $150. The ensuing reduction in sales would harm consumers who value the information at, say, $200. They get consumer surplus of $50 under the current arrangement but would cease to buy if the price rose substantially. If because of high elasticity of demand in the consumer segment of the market the only way to make a profit turned out to be a price attractive to commercial users alone, then all consumers would lose out – and so would the commercial clients, who would have to pay more for the listings because ProCD could not obtain any contribution toward costs from the consumer market.
To make price discrimination work, however, the seller must be able to control arbitrage. An air carrier sells tickets for less to vacationers than to business travelers, using advance purchase and Saturday-night-stay requirements to distinguish the categories. A producer of movies segments the market by time, releasing first to theaters, then to pay-per-view services, next to the videotape and laserdisc market, and finally to cable and commercial tv. Vendors of computer software have a harder task. Anyone can walk into a retail store and buy a box. Customers do not wear tags saying “commercial user” or “consumer user.” Anyway, even a commercial-user-detector at the door would not work, because a consumer could buy the software and resell to a commercial user. That arbitrage would break down the price discrimination and drive up the minimum price at which ProCD would sell to anyone.
Instead of tinkering with the product and letting users sort themselves – for example, furnishing current data at a high price that would be attractive only to commercial customers, and two-year-old data at a low price – ProCD turned to the institution of contract. Every box containing its consumer product declares that the software comes with restrictions stated in an enclosed license. This license, which is encoded on the CD-ROM disks as well as printed in the manual, and which appears on a user’s screen every time the software runs, limits use of the application program and listings to non-commercial purposes.
Matthew Zeidenberg bought a consumer package of SelectPhone in 1994 from a retail outlet in Madison, Wisconsin, but decided to ignore the license. He formed Silken Mountain Web Services, Inc., to resell the information in the SelectPhone database. The corporation makes the database available on the Internet to anyone willing to pay its price – which, needless to say, is less than ProCD charges its commercial customers. Zeidenberg has purchased two additional SelectPhone packages, each with an updated version of the database, and made the latest information available over the World Wide Web, for a price, through his corporation. ProCD filed this suit seeking an injunction against further dissemination that exceeds the rights specified in the licenses (identical in each of the three packages Zeidenberg purchased). The district court held the licenses ineffectual because their terms do not appear on the outside of the packages. The court added that the second and third licenses stand no different from the first, even though they are identical, because they might have been different, and a purchaser does not agree to – and cannot be bound by – terms that were secret at the time of purchase.
Following the district court, we treat the licenses as ordinary contracts accompanying the sale of products, and therefore as governed by the common law of contracts and the Uniform Commercial Code. Whether there are legal differences between “contracts” and “licenses” (which may matter under the copyright doctrine of first sale) is a subject for another day. Zeidenberg does argue, and the district court held, that placing the package of software on the shelf is an “offer,” which the customer “accepts” by paying the asking price and leaving the store with the goods. In Wisconsin, as elsewhere, a contract includes only the terms on which the parties have agreed. One cannot agree to hidden terms, the judge concluded. So far, so good – but one of the terms to which Zeidenberg agreed by purchasing the software is that the transaction was subject to a license. Zeidenberg’s position therefore must be that the printed terms on the outside of a box are the parties’ contract – except for printed terms that refer to or incorporate other terms. But why would Wisconsin fetter the parties’ choice in this way? Vendors can put the entire terms of a contract on the outside of a box only by using microscopic type, removing other information that buyers might find more useful (such as what the software does, and on which computers it works), or both. The “Read Me” file included with most software, describing system requirements and potential incompatibilities, may be equivalent to ten pages of type; warranties and license restrictions take still more space. Notice on the outside, terms on the inside, and a right to return the software for a refund if the terms are unacceptable (a right that the license expressly extends), may be a means of doing business valuable to buyers and sellers alike. Doubtless a state could forbid the use of standard contracts in the software business, but we do not think that Wisconsin has done so.
Transactions in which the exchange of money precedes the communication of detailed terms are common. Consider the purchase of insurance. The buyer goes to an agent, who explains the essentials (amount of coverage, number of years) and remits the premium to the home office, which sends back a policy. On the district judge’s understanding, the terms of the policy are irrelevant because the insured paid before receiving them. Yet the device of payment, often with a “binder” (so that the insurance takes effect immediately even though the home office reserves the right to withdraw coverage later), in advance of the policy, serves buyers’ interests by accelerating effectiveness and reducing transactions costs. Or consider the purchase of an airline ticket. The traveler calls the carrier or an agent, is quoted a price, reserves a seat, pays, and gets a ticket, in that order. The ticket contains elaborate terms, which the traveler can reject by canceling the reservation. To use the ticket is to accept the terms, even terms that in retrospect are disadvantageous. Just so with a ticket to a concert. The back of the ticket states that the patron promises not to record the concert; to attend is to agree. A theater that detects a violation will confiscate the tape and escort the violator to the exit. One could arrange things so that every concertgoer signs this promise before forking over the money, but that cumbersome way of doing things not only would lengthen queues and raise prices but also would scotch the sale of tickets by phone or electronic data service.
Consumer goods work the same way. Someone who wants to buy a radio set visits a store, pays, and walks out with a box. Inside the box is a leaflet containing some terms, the most important of which usually is the warranty, read for the first time in the comfort of home. By Zeidenberg’s lights, the warranty in the box is irrelevant; every consumer gets the standard warranty implied by the UCC in the event the contract is silent; yet so far as we are aware no state disregards warranties furnished with consumer products. Drugs come with a list of ingredients on the outside and an elaborate package insert on the inside. The package insert describes drug interactions, contraindications, and other vital information – but, if Zeidenberg is right, the purchaser need not read the package insert, because it is not part of the contract.
Next consider the software industry itself. Only a minority of sales take place over the counter, where there are boxes to peruse. A customer may place an order by phone in response to a line item in a catalog or a review in a magazine. Much software is ordered over the Internet by purchasers who have never seen a box. Increasingly software arrives by wire. There is no box; there is only a stream of electrons, a collection of information that includes data, an application program, instructions, many limitations (”MegaPixel 3.14159 cannot be used with Byte-Pusher 2.718”), and the terms of sale. The user purchases a serial number, which activates the software’s features. On Zeidenberg’s arguments, these unboxed sales are unfettered by terms – so the seller has made a broad warranty and must pay consequential damages for any shortfalls in performance, two “promises” that if taken seriously would drive prices through the ceiling or return transactions to the horse-and-buggy age.
According to the district court, the UCC does not countenance the sequence of money now, terms later. One of the court’s reasons – that by proposing as part of the draft Article 2B a new UCC sec. 2-2203 that would explicitly validate standard-form user licenses, the American Law Institute and the National Conference of Commissioners on Uniform Laws have conceded the invalidity of shrinkwrap licenses under current law, depends on a faulty inference. To propose a change in a law’s text is not necessarily to propose a change in the law’s effect. New words may be designed to fortify the current rule with a more precise text that curtails uncertainty.
What then does the current version of the UCC have to say? We think that the place to start is sec. 2-204(1): “A contract for sale of goods may be made in any manner sufficient to show agreement, including conduct by both parties which recognizes the existence of such a contract.” A vendor, as master of the offer, may invite acceptance by conduct, and may propose limitations on the kind of conduct that constitutes acceptance. A buyer may accept by performing the acts the vendor proposes to treat as acceptance. And that is what happened. ProCD proposed a contract that a buyer would accept by using the software after having an opportunity to read the license at leisure. This Zeidenberg did. He had no choice, because the software splashed the license on the screen and would not let him proceed without indicating acceptance. So although the district judge was right to say that a contract can be, and often is, formed simply by paying the price and walking out of the store, the UCC permits contracts to be formed in other ways. ProCD proposed such a different way, and without protest Zeidenberg agreed. Ours is not a case in which a consumer opens a package to find an insert saying “you owe us an extra $10,000” and the seller files suit to collect. Any buyer finding such a demand can prevent formation of the contract by returning the package, as can any consumer who concludes that the terms of the license make the software worth less than the purchase price. Nothing in the UCC requires a seller to maximize the buyer’s net gains.
Section 2-606, which defines “acceptance of goods”, reinforces this understanding. A buyer accepts goods under sec. 2-606(1)(b) when, after an opportunity to inspect, he fails to make an effective rejection under sec. 2-602(1). ProCD extended an opportunity to reject if a buyer should find the license terms unsatisfactory; Zeidenberg inspected the package, tried out the software, learned of the license, and did not reject the goods. We refer to sec. 2-606 only to show that the opportunity to return goods can be important; acceptance of an offer differs from acceptance of goods after delivery; but the UCC consistently permits the parties to structure their relations so that the buyer has a chance to make a final decision after a detailed review.
Some portions of the UCC impose additional requirements on the way parties agree on terms. A disclaimer of the implied warranty of merchantability must be “conspicuous.” UCC sec. 2-316(2), incorporating UCC sec. 1-201(10). Promises to make firm offers, or to negate oral modifications, must be “separately signed.” UCC secs. 2-205, 2-209(2). These special provisos reinforce the impression that, so far as the UCC is concerned, other terms may be as inconspicuous as the forum-selection clause on the back of the cruise ship ticket in Carnival Lines. Zeidenberg has not located any Wisconsin case – for that matter, any case in any state – holding that under the UCC the ordinary terms found in shrinkwrap licenses require any special prominence, or otherwise are to be undercut rather than enforced. In the end, the terms of the license are conceptually identical to the contents of the package. Just as no court would dream of saying that SelectPhone (trademark) must contain 3,100 phone books rather than 3,000, or must have data no more than 30 days old, or must sell for $100 rather than $150 – although any of these changes would be welcomed by the customer, if all other things were held constant – so, we believe, Wisconsin would not let the buyer pick and choose among terms. Terms of use are no less a part of “the product” than are the size of the database and the speed with which the software compiles listings. Competition among vendors, not judicial revision of a package’s contents, is how consumers are protected in a market economy. ProCD has rivals, which may elect to compete by offering superior software, monthly updates, improved terms of use, lower price, or a better compromise among these elements. As we stressed above, adjusting terms in buyers’ favor might help Matthew Zeidenberg today (he already has the software) but would lead to a response, such as a higher price, that might make consumers as a whole worse off.
The district court held that, even if Wisconsin treats shrinkwrap licenses as contracts, § 301(a) of the Copyright Act prevents their enforcement. The relevant part of § 301(a) preempts any “legal or equitable rights [under state law] that are equivalent to any of the exclusive rights within the general scope of copyright as specified by section 106 in works of authorship that are fixed in a tangible medium of expression and come within the subject matter of copyright as specified by sections 102 and 103.” ProCD’s software and data are “fixed in a tangible medium of expression,” and the district judge held that they are “within the subject matter of copyright.” The latter conclusion is plainly right for the copyrighted application program, and the judge thought that the data likewise are “within the subject matter of copyright” even if, after Feist, they are not sufficiently original to be copyrighted. One function of § 301(a) is to prevent states from giving special protection to works of authorship that Congress has decided should be in the public domain, which it can accomplish only if “subject matter of copyright” includes all works of a type covered by sections 102 and 103, even if federal law does not afford protection to them.
But are rights created by contract “equivalent to any of the exclusive rights within the general scope of copyright”? Three courts of appeals have answered “no.” The district court disagreed with these decisions, but we think them sound. Rights “equivalent to any of the exclusive rights within the general scope of copyright” are rights established by law – rights that restrict the options of persons who are strangers to the author. Copyright law forbids duplication, public performance, and so on, unless the person wishing to copy or perform the work gets permission; silence means a ban on copying. A copyright is a right against the world. Contracts, by contrast, generally affect only their parties; strangers may do as they please, so contracts do not create “exclusive rights.” Someone who found a copy of SelectPhone on the street would not be affected by the shrinkwrap license – though the federal copyright laws of their own force would limit the finder’s ability to copy or transmit the application program.
Think for a moment about trade secrets. One common trade secret is a customer list. After Feist, a simple alphabetical list of a firm’s customers, with address and telephone numbers, could not be protected by copyright. Yet Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470 (1974), holds that contracts about trade secrets may be enforced – precisely because they do not affect strangers’ ability to discover and use the information independently. If the amendment of § 301(a) in 1976 overruled Kewanee and abolished consensual protection of those trade secrets that cannot be copyrighted, no one has noticed – though abolition is a logical consequence of the district court’s approach. Think, too, about everyday transactions in intellectual property. A customer visits a video store and rents a copy of Night of the Lepus. The customer’s contract with the store limits use of the tape to home viewing and requires its return in two days. May the customer keep the tape, on the ground that § 301(a) makes the promise unenforceable?
A law student uses the LEXIS database, containing public-domain documents, under a contract limiting the results to educational endeavors; may the student resell his access to this database to a law firm from which LEXIS seeks to collect a much higher hourly rate? Suppose ProCD hires a firm to scour the nation for telephone directories, promising to pay $100 for each that ProCD does not already have. The firm locates 100 new directories, which it sends to ProCD with an invoice for $10,000. ProCD incorporates the directories into its database; does it have to pay the bill? Surely yes; Aronson v. Quick Point Pencil Co., 440 U.S. 257 (1979), holds that promises to pay for intellectual property may be enforced even though federal law (in Aronson, the patent law) offers no protection against third-party uses of that property.Footnote 2 But these illustrations are what our case is about. ProCD offers software and data for two prices: one for personal use, a higher price for commercial use. Zeidenberg wants to use the data without paying the seller’s price; if the law student and Quick Point Pencil Co. could not do that, neither can Zeidenberg.
Although Congress possesses power to preempt even the enforcement of contracts about intellectual property … courts usually read preemption clauses to leave private contracts unaffected. American Airlines, Inc. v. Wolens, 513 U.S. 219 (1995), provides a nice illustration. A federal statute preempts any state “law, rule, regulation, standard, or other provision … relating to rates, routes, or services of any air carrier.” Does such a law preempt the law of contracts – so that, for example, an air carrier need not honor a quoted price (or a contract to reduce the price by the value of frequent flyer miles)? The Court allowed that it is possible to read the statute that broadly but thought such an interpretation would make little sense. Terms and conditions offered by contract reflect private ordering, essential to the efficient functioning of markets. Although some principles that carry the name of contract law are designed to defeat rather than implement consensual transactions, the rules that respect private choice are not preempted by a clause such as § 1305(a)(1). Section 301(a) plays a role similar to § 1301(a)(1): it prevents states from substituting their own regulatory systems for those of the national government. Just as § 301(a) does not itself interfere with private transactions in intellectual property, so it does not prevent states from respecting those transactions. Like the Supreme Court in Wolens, we think it prudent to refrain from adopting a rule that anything with the label “contract” is necessarily outside the preemption clause: the variations and possibilities are too numerous to foresee.
Aronson emphasized that enforcement of the contract between Aronson and Quick Point Pencil Company would not withdraw any information from the public domain. That is equally true of the contract between ProCD and Zeidenberg. Everyone remains free to copy and disseminate all 3,000 telephone books that have been incorporated into ProCD’s database. Anyone can add SIC codes and zip codes. ProCD’s rivals have done so. Enforcement of the shrinkwrap license may even make information more readily available, by reducing the price ProCD charges to consumer buyers. To the extent licenses facilitate distribution of object code while concealing the source code (the point of a clause forbidding disassembly), they serve the same procompetitive functions as does the law of trade secrets. Licenses may have other benefits for consumers: many licenses permit users to make extra copies, to use the software on multiple computers, even to incorporate the software into the user’s products. But whether a particular license is generous or restrictive, a simple two-party contract is not “equivalent to any of the exclusive rights within the general scope of copyright” and therefore may be enforced.
REVERSED AND REMANDED.
Notes and Questions
1. Shrinkwrap and assent. The original “shrinkwrap” software licenses were visible in their entirety through the clear plastic packaging of the software box or diskette. By the time of ProCD, however, software vendors, not wishing to detract from the visual appeal of their packaging, included only a sticker indicating that licensing terms could be found inside the box, on the theory that if a consumer opened the box, then read the terms and was dissatisfied, he or she could return the product for a refund. What practical difficulties arise from this theory? How many consumers to you think requested such refunds? What risks exist for the software vendor in this scenario?
2. ProCD and the rise of the EULA. The ProCD case stands for two important principles of law. First, as discussed above, shrinkwrap license agreements can be enforceable contracts. But the second principle established in ProCD is equally important: the Copyright Act does not preempt state contract law when it seeks to cover material protected (or not protected) by copyright. As the court in ProCD notes, it was not the first court to rule in this manner on preemption, and the Supreme Court’s decision in Aronson laid the groundwork for ProCD, though in the area of patents rather than copyrights. But ProCD opened the door to consumer software contracts (end user license agreements or “EULAs”) that grew in length and contained an increasing number of legal terms that went well beyond the restrictions imposed by the Copyright Act. Just a few of the terms included in typical EULAs are:
limitations on the number of users/devices;
restrictions on uses (noncommercial, educational, no spam);
prohibitions on rental, resale, reverse engineering and transfer;
limitations and exclusions of warranty and damages;
consent to use of personal data; and
disputes will be resolved by arbitration in a designated locale.
Are EULA terms like this reasonable? How many consumers to you think are aware of the EULA limitations on the hundreds of different software programs that they use on a daily basis? This issue is discussed in greater detail in Section 17.3.
3. Preemption and the “extra element.” Most courts that have reviewed application of Section 301 of the Copyright Act to state law claims adopt what has been described as the “extra element” test. Under this approach, a state law claim is not preempted if it requires proof of a qualitatively extra or different element from that required to prove infringement. ProCD, and a number of other decisions, stand for the proposition that a contract claim involves that extra element. How would you describe the “extra element” that is involved? What types of claims might be subject to preemption? Does Judge Easterbrook suggest any of these in ProCD?
4. Reverse engineering. As in ProCD, in the absence of misuse or overreaching, courts have enforced standard-form contracts even if the contract terms give an IP holder rights beyond those afforded by copyright, patent or other applicable laws. For example, in Bowers v. Baystate Technologies, Inc., 320 F.3d 1317 (Fed. Cir. 2003), the Federal Circuit held that a shrinkwrap license agreement prohibiting reverse engineering of software was not preempted by the copyright law, even though reverse engineering would likely have been permissible as fair use under copyright law. The Bowers decision was criticized by Judge Dyk, who dissented. In his view, such contractual clauses had the potential to displace the protections of federal law in a manner that would not have been permissible had they been enacted by a state legislature. Judge Dyk acknowledges that parties may in “freely negotiated” agreements give up rights like fair use that are otherwise available under the law, but doing so under a contract of adhesion, which effectively gives the user no alternative, should not be permitted.Footnote 3 Which of these positions do you find more persuasive?
998 P.2d 305 (Wash. 2000)
JOHNSON, JUSTICE
Mortenson is a nationwide construction contractor. Respondent Timberline is a software developer located in Beaverton, Oregon. Respondent Softworks, an authorized dealer for Timberline, is located in Kirkland, Washington and provides computer-related services to contractors such as Mortenson.
Since at least 1990, Mortenson has used Timberline’s Bid Analysis software to assist with its preparation of bids. Mortenson had used Medallion, an earlier version of Bid Analysis, at its Minnesota headquarters and its regional offices. In early 1993, Mortenson installed a new computer network operating system at its Bellevue office and contacted Mark Reich (Reich), president of Softworks, to reinstall Medallion. Reich discovered, however, that the Medallion software was incompatible with Mortenson’s new operating system. Reich informed Mortenson that Precision, a newer version of Bid Analysis, was compatible with its new operating system.
Mortenson wanted multiple copies of the new software for its offices, including copies for its corporate headquarters in Minnesota and its northwest regional office in Bellevue. Reich informed Mortenson he would place an order with Timberline and would deliver eight copies of the Precision software to the Bellevue office, after which Mortenson could distribute the copies among its offices.
After Reich provided Mortenson with a price quote, Mortenson issued a purchase order dated July 12, 1993, confirming the agreed upon purchase price, set up fee, delivery charges, and sales tax for eight copies of the software. The purchase order indicated that Softworks, on behalf of Timberline, would “[f]urnish current versions of Timberline Precision Bid Analysis Program Software and Keys” and “[p]rovide assistance in installation and system configuration for Mortenson’s Bellevue Office.” The purchase order also contained the following notations:
Provide software support in converting Mortenson’s existing Bid Day Master Files to a format accepted by the newly purchased Bid Day software. This work shall be accomplished on a time and material basis of $85.00 per hour. Format information of conversion of existing D-Base Files to be shared to assist Mortenson Mid-West programmers in file conversion.
— System software support and upgrades to be available from Timberline for newly purchased versions of Bid Day Multi-User.
— At some future date should Timberline upgrade “Bid Day” to a windows version, M.A. Mortenson would be able to upgrade to this system with Timberline crediting existing software purchase toward that upgrade on a pro-rated basis to be determined later.
Below the signature line the following was stated: “ADVISE PURCHASING PROMPTLY IF UNABLE TO SHIP AS REQUIRED. EACH SHIPMENT MUST INCLUDE A PACKING LIST. SUBSTITUTIONS OF GOODS OR CHANGES IN COSTS REQUIRE OUR PRIOR APPROVAL.” The purchase order did not contain an integration clause.
Reich signed the purchase order and ordered the requested software from Timberline. When Reich received the software, he opened the three large shipping boxes and checked the contents against the packing invoice. Contained inside the shipping boxes were several smaller boxes, containing program diskettes in plastic pouches, installation instructions, and user manuals. One of the larger boxes also contained the sealed protection devices for the software.
All Timberline software is distributed to its users under license. Both Medallion and Precision Bid Analysis are licensed Timberline products. In the case of the Mortenson shipment, the full text of Timberline’s license agreement was set forth on the outside of each diskette pouch and the inside cover of the instruction manuals. The first screen that appears each time the program is used also references the license and states, “[t]his software is licensed for exclusive use by: Timberline Use Only.” Further, a license to use the protection device was wrapped around each of the devices shipped to Mortenson. The following warning preceded the terms of the license agreement:
CAREFULLY READ THE FOLLOWING TERMS AND CONDITIONS BEFORE USING THE PROGRAMS. USE OF THE PROGRAMS INDICATES YOUR ACKNOWLEDGEMENT THAT YOU HAVE READ THIS LICENSE, UNDERSTAND IT, AND AGREE TO BE BOUND BY ITS TERMS AND CONDITIONS. IF YOU DO NOT AGREE TO THESE TERMS AND CONDITIONS, PROMPTLY RETURN THE PROGRAMS AND USER MANUALS TO THE PLACE OF PURCHASE AND YOUR PURCHASE PRICE WILL BE REFUNDED. YOU AGREE THAT YOUR USE OF THE PROGRAM ACKNOWLEDGES THAT YOU HAVE READ THIS LICENSE, UNDERSTAND IT, AND AGREE TO BE BOUND BY ITS TERMS AND CONDITIONS.
Under a separate subheading, the license agreement limited Mortenson’s remedies and provided:
LIMITATION OF REMEDIES AND LIABILITY
NEITHER TIMBERLINE NOR ANYONE ELSE WHO HAS BEEN INVOLVED IN THE CREATION, PRODUCTION OR DELIVERY OF THE PROGRAMS OR USER MANUALS SHALL BE LIABLE TO YOU FOR ANY DAMAGES OF ANY TYPE, INCLUDING BUT NOT LIMITED TO, ANY LOST PROFITS, LOST SAVINGS, LOSS OF ANTICIPATED BENEFITS, OR OTHER INCIDENTAL, OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE SUCH PROGRAMS, WHETHER ARISING OUT OF CONTRACT, NEGLIGENCE, STRICT TORT, OR UNDER ANY WARRANTY, OR OTHERWISE, EVEN IF TIMBERLINE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES OR FOR ANY OTHER CLAIM BY ANY OTHER PARTY. TIMBERLINE’S LIABILITY FOR DAMAGES IN NO EVENT SHALL EXCEED THE LICENSE FEE PAID FOR THE RIGHT TO USE THE PROGRAMS.
Reich personally delivered the software to Mortenson’s Bellevue office, and was asked to return at a later date for installation. The parties dispute what happened next. According to Neal Ruud (Ruud), Mortenson’s chief estimator at its Bellevue office, when Reich arrived to install the software Reich personally opened the smaller product boxes contained within the large shipping boxes and also opened the diskette packaging. Reich inserted the diskettes into the computer, initiated the program, contacted Timberline to receive the activation codes, and wrote down the codes for Mortenson. Reich then started the programs and determined to the best of his knowledge they were operating properly. Ruud states that Mortenson never saw any of the licensing information described above, or any of the manuals that accompanied the software. Ruud adds that copies of the programs purchased for other Mortenson offices were forwarded to those offices.
Reich claims when he arrived at Mortenson’s Bellevue office he noticed the software had been opened and had been placed on a desk, along with a manual and a protection device. Reich states he told Mortenson he would install the program at a single workstation and “then they would do the rest.” Reich proceeded to install the software and a Mortenson employee attached the protection device. Reich claims he initiated and ran the program, and then observed as a Mortenson employee repeated the installation process on a second computer. An employee then told Reich that Mortenson would install the software at the remaining stations.
In December 1993, Mortenson utilized the Precision Bid Analysis software to prepare a bid for a project at Harborview Medical Center in Seattle. On the day of the bid, the software allegedly malfunctioned multiple times and gave the following message: “Abort: Cannot find alternate.” Clerk’s Papers at 60. Mortenson received this message 19 times that day. Nevertheless, Mortenson submitted a bid generated by the software. After Mortenson was awarded the Harborview Medical Center project, it learned its bid was approximately $1.95 million lower than intended.
Mortenson filed an action in King County Superior Court against Timberline and Softworks alleging breach of express and implied warranties. Timberline moved for summary judgment of dismissal in July 1997, arguing the limitation on consequential damages in the licensing agreement barred Mortenson’s recovery. Mortenson countered that its entire contract with Timberline consisted of the purchase order and it never saw or agreed to the provisions in the licensing agreement.
Mortenson used Timberline’s Precision Bid Analysis software to prepare a bid for a project at Harborview Medical Center in Seattle. The software malfunctioned.
Analysis
Terms of the Contract
Mortenson [argues that] even if the purchase order was not an integrated contract, Timberline’s delivery of the license terms merely constituted a request to add additional or different terms, which were never agreed upon by the parties. Mortenson claims under RCW 62A.2-207 the additional terms did not become part of the contract because they were material alterations. Timberline responds that the terms of the license were not a request to add additional terms, but part of the contract between the parties. Timberline further argues that so-called “shrinkwrap” software licenses have been found enforceable by other courts, and that both trade usage and course of dealing support enforcement in the present case. For its section 2-207 analysis, Mortenson relies on Step-Saver Data Sys., Inc. v. Wyse Tech., 939 F.2d 91 (3d Cir.1991). Mortenson claims Step-Saver is controlling, as “virtually every element of the transaction in the present case is mirrored in Step-Saver.” We disagree.
First, Step-Saver did not involve the enforceability of a standard license agreement against an end user of the software, but instead involved its applicability to a value added retailer who simply included the software in an integrated system sold to the end user. In fact, in Step-Saver the party contesting applicability of the licensing agreement had been assured the license did not apply to it at all. Such is not the case here, as Mortenson was the end user of the Bid Analysis software and was never told the license agreement did not apply.
Further, in Step-Saver the seller of the program twice asked the buyer to sign an agreement comparable to their disputed license agreement. Both times the buyer refused, but the seller continued to make the software available. In contrast, Mortenson and Timberline had utilized a license agreement throughout Mortenson’s use of the Medallion and Precision Bid Analysis software. Given these distinctions, we find Step-Saver to be inapplicable to the present case. We conclude this is a case about contract formation, not contract alteration. As such, RCW 62A.2-204, and not RCW 62A.2-207, provides the proper framework for our analysis.
RCW 62A.2-204 states:
(1) A contract for sale of goods may be made in any manner sufficient to show agreement, including conduct by both parties which recognizes the existence of such a contract.
(2) An agreement sufficient to constitute a contract for sale may be found even though the moment of its making is undetermined.
(3) Even though one or more terms are left open a contract for sale does not fail for indefiniteness if the parties have intended to make a contract and there is a reasonably certain basis for giving an appropriate remedy.
Although no Washington case specifically addresses the type of contract formation at issue in this case, a series of recent cases from other jurisdictions have analyzed shrinkwrap licenses under analogous statutes.
In ProCD, which involved a retail purchase of software, the Seventh Circuit held software shrinkwrap license agreements are a valid form of contracting under Wisconsin’s version of U.C.C. section 2-204, and such agreements are enforceable unless objectionable under general contract law such as the law of unconscionability. The court stated, “[n]otice on the outside, terms on the inside, and a right to return the software for a refund if the terms are unacceptable (a right that the license expressly extends), may be a means of doing business valuable to buyers and sellers alike.”
In Hill, the customer ordered a computer over the telephone and received the computer in the mail, accompanied by a list of terms to govern if the customer did not return the product within 30 days. Relying in part on ProCD, the court held the terms of the “accept-or-return” agreement were effective, stating, “[c]ompetent adults are bound by such documents, read or unread.” Elaborating on its holding in ProCD, the court continued:
The question in ProCD was not whether terms were added to a contract after its formation, but how and when the contract was formed – in particular, whether a vendor may propose that a contract of sale be formed, not in the store (or over the phone) with the payment of money or a general “send me the product,” but after the customer has had a chance to inspect both the item and the terms. ProCD answers “yes,” for merchants and consumers alike.
Interpreting the same licensing agreement at issue in Hill, the New York Supreme Court, Appellate Division concluded shrinkwrap license terms delivered following a mail order purchase were not proposed additions to the contract, but part of the original agreement between the parties. The court held U.C.C. section 2-207 did not apply because the contract was not formed until after the period to return the merchandise.
We find the approach of the ProCD, Hill, and Brower courts persuasive and adopt it to guide our analysis under RCW 62A.2-204. We conclude because RCW 62A.2-204 allows a contract to be formed “in any manner sufficient to show agreement … even though the moment of its making is undetermined,” it allows the formation of “layered contracts” similar to those envisioned by ProCD, Hill and Brower. We, therefore, hold under RCW 62A.2-204 the terms of the license were part of the contract between Mortenson and Timberline, and Mortenson’s use of the software constituted its assent to the agreement, including the license terms.
The terms of Timberline’s license were either set forth explicitly or referenced in numerous locations. The terms were included within the shrinkwrap packaging of each copy of Precision Bid Analysis; they were present in the manuals accompanying the software; they were included with the protection devices for the software, without which the software could not be used. The fact the software was licensed was also noted on the introductory screen each time the software was used. Even accepting Mortenson’s contention it never saw the terms of the license, it was not necessary for Mortenson to actually read the agreement in order to be bound by it.
Furthermore, the U.C.C. defines an “agreement” as “the bargain of the parties in fact as found in their language or by implication from other circumstances including course of dealing or usage of trade or course of performance”. Mortenson and Timberline had a course of dealing; Mortenson had purchased licensed software from Timberline for years prior to its upgrade to Precision Bid Analysis. All Timberline software, including the prior version of Bid Analysis used by Mortenson since at least 1990, is distributed under license. Moreover, extensive testimony and exhibits before the trial court demonstrate an unquestioned use of such license agreements throughout the software industry. Although Mortenson questioned the relevance of this evidence, there is no evidence in the record to contradict it. While trade usage is a question of fact, undisputed evidence of trade usage may be considered on summary judgment.
As the license was part of the contract between Mortenson and Timberline, its terms are enforceable unless “objectionable on grounds applicable to contracts in general.”
We affirm the Court of Appeals, upholding the trial court’s order of summary judgment of dismissal and denial of the motions to vacate and amend.
Notes and Questions
1. Offer and counteroffer. Though cases like ProCD and Mortenson generally established that shrinkwrap agreements are enforceable contracts, some courts have declined to enforce them under certain circumstances. One such case is Step-Saver Data Sys., Inc. v. Wyse Technology, 939 F.2d 91, 102–03 (3d Cir. 1991), which the court in Mortenson distinguished. In that case, Step-Saver was a value-added reseller (VAR) that obtained computer terminals from Wyse and software operating systems from TSL. Step-Saver combined these components with IBM computers and sold them as a package to consumers. When consumers complained that the system did not work, Step-Saver brought a warranty claim against TSL. TSL argued, in response, that a box-top disclaimer accompanying its operating system software negated any warranties to Step-Saver or its customers. The court disagreed, holding that in order for the warranty disclaimer to form part of the contract between TSL and Step-Saver, it must constitute a formal counteroffer by TSL, without which TSL was unwilling to complete the transaction.
On its face, the box-top license states that TSL will refund the purchase price if the purchaser does not agree to the terms of the license. Even with such a refund term, however, [TSL] may be relying on the purchaser’s investment in time and energy in reaching this point in the transaction to prevent the purchaser from returning the item. Because a purchaser has made a decision to buy a particular product and has actually obtained the product, the purchaser may use it despite the refund offer, regardless of the additional terms specified after the contract formed. But we need not decide whether such a refund offer could ever amount to a conditional acceptance; the undisputed evidence in this case demonstrates that the terms of the license were not sufficiently important that TSL would forego its sales to Step-Saver if TSL could not obtain Step-Saver’s consent to those terms. [A Step-Saver employee] testified that TSL assured him that the box-top license did not apply to Step-Saver, as Step-Saver was not the end user of the Multilink Advanced program. Supporting this testimony, TSL on two occasions asked Step-Saver to sign agreements that … contained warranty disclaimer and limitation of remedy terms similar to those contained in the box-top license. Step-Saver refused to sign the agreements; nevertheless, TSL continued to sell copies of Multilink Advanced to Step-Saver.
With this reasoning in mind, what should a licensor like TSL do in order to limit its liability to resellers like Step-Saver?
A 1980s-era computer terminal of the type at issue in Step-Saver.
2. Invoicing and terms beyond software. Software vendors are not alone in their use of standardized consumer contracts. For example, in Greenfield v. Twin Vision Graphics, Inc., 268 F. Supp. 2d 358 (D.N.J. 2003), a court enforced a restriction on the use of a commercial photograph that was contained in a photographer’s invoice. The court distinguished Step-Saver, reasoning that TSL continued to provide the Multilink software even after Step-Saver refused to enter into a contract with the requested disclaimer. The photographer, in contrast, did not provide additional services after sending the invoice, nor did the customer reject the terms of the invoice. Puget Sound Financial, LLC v. Unisearch, Inc., 47 P.3d 940 (Wash. 2002) also involved a term contained in an invoice – a damages limitation relating to services provided by a public records search firm. The court held that the damages limitation became part of the contract whether or not the customer expressly agreed to the invoice containing the terms because the limitation was common in the trade and routinely used in such contracts. Given these decisions, what limits exist on the ability of product and service vendors to impose contractual restrictions and terms on customers through invoices and other post-transaction documentation?
3. Seed tag licenses. For years, Monsanto and other sellers of genetically modified seeds and other proprietary agricultural products have affixed legal terms to the bags and packaging of these products. Known as “seed tag” licenses, these terms generally prohibit the purchaser from selling new seeds that result from the planting of the original seeds (see discussion of Bowman v. Monsanto in Section 23.4, Note 8). Recently, however, this practice has invaded the consumer market. In 2020, a surprised Twitter user circulated a photo of a EULA on a plastic bag of Carnival brand seedless grapes. It read:
The recipient of the produce contained in this package agrees not to propagate or reproduce any portion of this produce, including “but not limited to” seeds, stems, tissue, and fruit.
Is there any product the use of which is not susceptible to limitation by contract of adhesion?
4. Unconscionability. Despite general skepticism toward standard-form contracts, US law has generally accepted that they are enforceable, so long as they do not rise (or fall) to the level of unconscionability. While there is no clear definition of unconscionability, the Restatement (Second) of Contracts § 208 (Comment), explains, somewhat unhelpfully:
The determination that a contract or term is or is not unconscionable is made in the light of its setting, purpose and effect. Relevant factors include weaknesses in the contracting process like those involved in more specific rules as to contractual capacity, fraud, and other invalidating causes; the policy also overlaps with rules which render particular bargains or terms unenforceable on grounds of public policy.
In AT&T Mobility, LLC v. Concepcion, 563 U.S. 333 (2011), the Supreme Court summarized the California law of unconscionability as requiring “a ‘procedural’ and a ‘substantive’ element, the former focusing on ‘oppression’ or ‘surprise’ due to unequal bargaining power, the latter on ‘overly harsh’ or ‘one-sided’ results.” One common challenge to online agreements arises from arbitration clauses that require consumers to participate in binding arbitration, often in distant cities, in order to resolve disputes under the agreements. In Carey v. Uber Technologies, Inc., Case No. 1:16-cv-1058 (N.D. Ohio Mar. 27, 2017), a district court applied the procedural unconscionability test to uphold a contractual clause that delegated the question of whether an issue was arbitrable to the arbitrator. The court found that the user’s assent to the clickwrap terms and the opportunity to opt out of the arbitration provision defeated any finding of procedural unconscionability. Likewise, in Corwin v. NYC Bike Share, LLC, 238 F. Supp. 3d 475 (S.D.N.Y. 2017), the court held that an injury release in a clickwrap agreement was not unconscionable because the full text of the release agreement was embedded within the registration page, the user could not continue to register before manifesting assent and the terms were in plain view. Given these precedents, what kind of shrinkwrap license terms, if any, might a court find unconscionable?
Vendors have sought to control the use of products, including seedless grapes, through “shrinkwrap” agreements.
The doctrine of unconscionability under US law today is distinctly anemic. Amit Elazari Bar On has recently proposed that unconscionability be reinvigorated along the lines used in Israel. There, according to Elazari, “if a term in a standard form contract meets the criteria of unconscionability presumptions, the burden of proof is borne by the drafter, who must prove that, in view of the contract as a whole and its particular circumstances, the condition in question is justified and reasonable.”Footnote 4 How might Elazari’s proposal change the frequency of unconscionability findings in the United States? If you were advising a software or online services provider, would you recommend that they support or oppose such a change to the law? Why?
17.2 Clickwrap and Browsewrap Licenses
In the mid-1990s, firms began to distribute consumer software and services via the Internet. In doing so, they had to contend with means for imposing binding contractual terms on users. Without a physical software disc or CD, they could not rely on the tried-and-true shrinkwrap method, but they quickly adapted the principles of ProCD to the online realm. What emerged were contractual terms presented to the user in electronic, on-screen form only, in which the user’s assent was manifested through clicking an “I ACCEPT” or similar graphic button. Initially, courts struggled with the ability of parties to form contracts electronically through the click of a button on a computer screen. However, they soon accepted the notion that such “clickwrap” or “click-through” agreements offered sufficient evidence of assent to form binding contracts if the user’s notice and assent were clear.
A committee of the American Bar Association, writing early in the evolution of clickwrap agreements, proposed a series best practices for the creation of legally binding clickwrap agreements.Footnote 5 The principles articulated by the ABA Committee recommended that online terms be clearly displayed, and users be given an opportunity to review them, that users have the opportunity to manifest their acceptance or rejection of the terms, that an opportunity to correct errors be provided, and that records be maintained to prove assent. Over the years, online agreements meeting these guidelines have generally been found to be enforceable.
Some software vendors, however, sought to streamline the contracting process further by eliminating the consumer’s click – after all, traditional shrinkwrap agreements simply required the user to open a package and begin to use a software program in order to be bound. A new generation of agreements thus emerged that sought to bind the user simply by virtue of his or her use of the licensed software, website or service. These agreements became known as “browsewrap” agreements.Footnote 6 The following case is one of the first to deal thoroughly with the issues that they raised.
306 F.3d 17 (2d Cir. 2002)
SOTOMAYOR, CIRCUIT JUDGE
This is an appeal from a judgment of the Southern District of New York denying a motion by defendants-appellants Netscape Communications Corporation and its corporate parent, America Online, Inc. (collectively, “defendants” or “Netscape”), to compel arbitration and to stay court proceedings. In order to resolve the central question of arbitrability presented here, we must address issues of contract formation in cyberspace. Principally, we are asked to determine whether plaintiffs-appellees (“plaintiffs”), by acting upon defendants’ invitation to download free software made available on defendants’ webpage, agreed to be bound by the software’s license terms (which included the arbitration clause at issue), even though plaintiffs could not have learned of the existence of those terms unless, prior to executing the download, they had scrolled down the webpage to a screen located below the download button. We agree with the district court that a reasonably prudent Internet user in circumstances such as these would not have known or learned of the existence of the license terms before responding to defendants’ invitation to download the free software, and that defendants therefore did not provide reasonable notice of the license terms. In consequence, plaintiffs’ bare act of downloading the software did not unambiguously manifest assent to the arbitration provision contained in the license terms.
We also agree with the district court that plaintiffs’ claims relating to the software at issue – a “plug-in” program entitled SmartDownload (“SmartDownload” or “the plug-in program”), offered by Netscape to enhance the functioning of the separate browser program called Netscape Communicator (“Communicator” or “the browser program”) – are not subject to an arbitration agreement contained in the license terms governing the use of Communicator … We therefore affirm the district court’s denial of defendants’ motion to compel arbitration and to stay court proceedings.
Background
In three related putative class actions, plaintiffs alleged that, unknown to them, their use of SmartDownload transmitted to defendants private information about plaintiffs’ downloading of files from the Internet, thereby effecting an electronic surveillance of their online activities in violation of two federal statutes, the Electronic Communications Privacy Act, 18 U.S.C. §§ 2510 et seq., and the Computer Fraud and Abuse Act, 18 U.S.C. § 1030.
In the time period relevant to this litigation, Netscape offered on its website various software programs, including Communicator and SmartDownload, which visitors to the site were invited to obtain free of charge. It is undisputed that [plaintiffs] downloaded Communicator from the Netscape website. These plaintiffs acknowledge that when they proceeded to initiate installation of Communicator, they were automatically shown a scrollable text of that program’s license agreement and were not permitted to complete the installation until they had clicked on a “Yes” button to indicate that they accepted all the license terms. If a user attempted to install Communicator without clicking “Yes,” the installation would be aborted. All five named user plaintiffs expressly agreed to Communicator’s license terms by clicking “Yes.” The Communicator license agreement that these plaintiffs saw made no mention of SmartDownload or other plug-in programs, and stated that “[t]hese terms apply to Netscape Communicator and Netscape Navigator” and that “all disputes relating to this Agreement (excepting any dispute relating to intellectual property rights)” are subject to “binding arbitration in Santa Clara County, California.”
Although Communicator could be obtained independently of SmartDownload, all the named user plaintiffs, except Fagan, downloaded and installed Communicator in connection with downloading SmartDownload. Each of these plaintiffs allegedly arrived at a Netscape webpage captioned “SmartDownload Communicator” that urged them to “Download With Confidence Using SmartDownload!” At or near the bottom of the screen facing plaintiffs was the prompt “Start Download” and a tinted button labeled “Download.” By clicking on the button, plaintiffs initiated the download of SmartDownload. Once that process was complete, SmartDownload, as its first plug-in task, permitted plaintiffs to proceed with downloading and installing Communicator, an operation that was accompanied by the clickwrap display of Communicator’s license terms described above.
Netscape Navigator was the most popular early web browser.
The signal difference between downloading Communicator and downloading SmartDownload was that no clickwrap presentation accompanied the latter operation. Instead, once plaintiffs Gibson, Gruber, Kelly, and Weindorf had clicked on the “Download” button located at or near the bottom of their screen, and the downloading of SmartDownload was complete, these plaintiffs encountered no further information about the plug-in program or the existence of license terms governing its use. The sole reference to SmartDownload’s license terms on the “SmartDownload Communicator” webpage was located in text that would have become visible to plaintiffs only if they had scrolled down to the next screen.
Had plaintiffs scrolled down instead of acting on defendants’ invitation to click on the “Download” button, they would have encountered the following invitation: “Please review and agree to the terms of the Netscape SmartDownload software license agreement before downloading and using the software.” Plaintiffs Gibson, Gruber, Kelly, and Weindorf averred in their affidavits that they never saw this reference to the SmartDownload license agreement when they clicked on the “Download” button. They also testified during depositions that they saw no reference to license terms when they clicked to download SmartDownload, although under questioning by defendants’ counsel, some plaintiffs added that they could not “remember” or be “sure” whether the screen shots of the SmartDownload page attached to their affidavits reflected precisely what they had seen on their computer screens when they downloaded SmartDownload.
In sum, plaintiffs Gibson, Gruber, Kelly, and Weindorf allege that the process of obtaining SmartDownload contrasted sharply with that of obtaining Communicator. Having selected SmartDownload, they were required neither to express unambiguous assent to that program’s license agreement nor even to view the license terms or become aware of their existence before proceeding with the invited download of the free plug-in program. Moreover, once these plaintiffs had initiated the download, the existence of SmartDownload’s license terms was not mentioned while the software was running or at any later point in plaintiffs’ experience of the product.
Even for a user who, unlike plaintiffs, did happen to scroll down past the download button, SmartDownload’s license terms would not have been immediately displayed in the manner of Communicator’s clickwrapped terms. Instead, if such a user had seen the notice of SmartDownload’s terms and then clicked on the underlined invitation to review and agree to the terms, a hypertext link would have taken the user to a separate webpage entitled “License & Support Agreements.” The first paragraph on this page read, in pertinent part:
The use of each Netscape software product is governed by a license agreement. You must read and agree to the license agreement terms BEFORE acquiring a product. Please click on the appropriate link below to review the current license agreement for the product of interest to you before acquisition. For products available for download, you must read and agree to the license agreement terms BEFORE you install the software. If you do not agree to the license terms, do not download, install or use the software.
Below this paragraph appeared a list of license agreements, the first of which was “License Agreement for Netscape Navigator and Netscape Communicator Product Family (Netscape Navigator, Netscape Communicator and Netscape SmartDownload).” If the user clicked on that link, he or she would be taken to yet another webpage that contained the full text of a license agreement that was identical in every respect to the Communicator license agreement except that it stated that its “terms apply to Netscape Communicator, Netscape Navigator, and Netscape SmartDownload.” The license agreement granted the user a nonexclusive license to use and reproduce the software, subject to certain terms:
BY CLICKING THE ACCEPTANCE BUTTON OR INSTALLING OR USING NETSCAPE COMMUNICATOR, NETSCAPE NAVIGATOR, OR NETSCAPE SMARTDOWNLOAD SOFTWARE (THE “PRODUCT”), THE INDIVIDUAL OR ENTITY LICENSING THE PRODUCT (“LICENSEE”) IS CONSENTING TO BE BOUND BY AND IS BECOMING A PARTY TO THIS AGREEMENT. IF LICENSEE DOES NOT AGREE TO ALL OF THE TERMS OF THIS AGREEMENT, THE BUTTON INDICATING NON-ACCEPTANCE MUST BE SELECTED, AND LICENSEE MUST NOT INSTALL OR USE THE SOFTWARE.
Among the license terms was a provision requiring virtually all disputes relating to the agreement to be submitted to arbitration:
Unless otherwise agreed in writing, all disputes relating to this Agreement (excepting any dispute relating to intellectual property rights) shall be subject to final and binding arbitration in Santa Clara County, California, under the auspices of JAMS/EndDispute, with the losing party paying all costs of arbitration.
Unlike the four named user plaintiffs who downloaded SmartDownload from the Netscape website, the fifth named plaintiff, Michael Fagan, claims to have downloaded the plug-in program from a “shareware” website operated by ZDNet, an entity unrelated to Netscape. Shareware sites are websites, maintained by companies or individuals, that contain libraries of free, publicly available software. The pages that a user would have seen while downloading SmartDownload from ZDNet differed from those that he or she would have encountered while downloading SmartDownload from the Netscape website. Notably, instead of any kind of notice of the SmartDownload license agreement, the ZDNet pages offered only a hypertext link to “more information” about SmartDownload, which, if clicked on, took the user to a Netscape webpage that, in turn, contained a link to the license agreement. Thus, a visitor to the ZDNet website could have obtained SmartDownload, as Fagan avers he did, without ever seeing a reference to that program’s license terms, even if he or she had scrolled through all of ZDNet’s webpages.
Discussion
Whether governed by the common law or by Article 2 of the Uniform Commercial Code (“UCC”), a transaction, in order to be a contract, requires a manifestation of agreement between the parties. Mutual manifestation of assent, whether by written or spoken word or by conduct, is the touchstone of contract. Although an onlooker observing the disputed transactions in this case would have seen each of the user plaintiffs click on the SmartDownload “Download” button, a consumer’s clicking on a download button does not communicate assent to contractual terms if the offer did not make clear to the consumer that clicking on the download button would signify assent to those terms. California’s common law is clear that “an offeree, regardless of apparent manifestation of his consent, is not bound by inconspicuous contractual provisions of which he is unaware, contained in a document whose contractual nature is not obvious.”
Arbitration agreements are no exception to the requirement of manifestation of assent. “This principle of knowing consent applies with particular force to provisions for arbitration.” Clarity and conspicuousness of arbitration terms are important in securing informed assent. “If a party wishes to bind in writing another to an agreement to arbitrate future disputes, such purpose should be accomplished in a way that each party to the arrangement will fully and clearly comprehend that the agreement to arbitrate exists and binds the parties thereto.” Thus, California contract law measures assent by an objective standard that takes into account both what the offeree said, wrote, or did and the transactional context in which the offeree verbalized or acted.
The Reasonably Prudent Offeree of Downloadable Software
Defendants argue that plaintiffs must be held to a standard of reasonable prudence and that, because notice of the existence of SmartDownload license terms was on the next scrollable screen, plaintiffs were on “inquiry notice” of those terms. We disagree with the proposition that a reasonably prudent offeree in plaintiffs’ position would necessarily have known or learned of the existence of the SmartDownload license agreement prior to acting, so that plaintiffs may be held to have assented to that agreement with constructive notice of its terms. See Cal. Civ.Code § 1589 (“A voluntary acceptance of the benefit of a transaction is equivalent to a consent to all the obligations arising from it, so far as the facts are known, or ought to be known, to the person accepting.”). It is true that “[a] party cannot avoid the terms of a contract on the ground that he or she failed to read it before signing.” Marin Storage & Trucking, 89 Cal.App.4th at 1049. But courts are quick to add: “An exception to this general rule exists when the writing does not appear to be a contract and the terms are not called to the attention of the recipient. In such a case, no contract is formed with respect to the undisclosed term.”
Most of the cases cited by defendants in support of their inquiry-notice argument are drawn from the world of paper contracting. As [these] cases suggest, receipt of a physical document containing contract terms or notice thereof is frequently deemed, in the world of paper transactions, a sufficient circumstance to place the offeree on inquiry notice of those terms. “Every person who has actual notice of circumstances sufficient to put a prudent man upon inquiry as to a particular fact, has constructive notice of the fact itself in all cases in which, by prosecuting such inquiry, he might have learned such fact.” Cal. Civ.Code § 19. These principles apply equally to the emergent world of online product delivery, pop-up screens, hyperlinked pages, clickwrap licensing, scrollable documents, and urgent admonitions to “Download Now!” What plaintiffs saw when they were being invited by defendants to download this fast, free plug-in called SmartDownload was a screen containing praise for the product and, at the very bottom of the screen, a “Download” button. Defendants argue that under the principles set forth in the cases cited above, a “fair and prudent person using ordinary care” would have been on inquiry notice of SmartDownload’s license terms.
We are not persuaded that a reasonably prudent offeree in these circumstances would have known of the existence of license terms. Plaintiffs were responding to an offer that did not carry an immediately visible notice of the existence of license terms or require unambiguous manifestation of assent to those terms. Thus, plaintiffs’ “apparent manifestation of … consent” was to terms contained in a document whose contractual nature [was] not obvious. Moreover, the fact that, given the position of the scroll bar on their computer screens, plaintiffs may have been aware that an unexplored portion of the Netscape webpage remained below the download button does not mean that they reasonably should have concluded that this portion contained a notice of license terms. In their deposition testimony, plaintiffs variously stated that they used the scroll bar “[o]nly if there is something that I feel I need to see that is on – that is off the page,” or that the elevated position of the scroll bar suggested the presence of “mere formalities, standard lower banner links” or “that the page is bigger than what I can see.” Plaintiffs testified, and defendants did not refute, that plaintiffs were in fact unaware that defendants intended to attach license terms to the use of SmartDownload.
We conclude that in circumstances such as these, where consumers are urged to download free software at the immediate click of a button, a reference to the existence of license terms on a submerged screen is not sufficient to place consumers on inquiry or constructive notice of those terms. The SmartDownload webpage screen was “printed in such a manner that it tended to conceal the fact that it was an express acceptance of [Netscape’s] rules and regulations.” Internet users may have, as defendants put it, “as much time as they need” to scroll through multiple screens on a webpage, but there is no reason to assume that viewers will scroll down to subsequent screens simply because screens are there. When products are “free” and users are invited to download them in the absence of reasonably conspicuous notice that they are about to bind themselves to contract terms, the transactional circumstances cannot be fully analogized to those in the paper world of arm’s-length bargaining. In the next two sections, we discuss case law and other legal authorities that have addressed the circumstances of computer sales, software licensing, and online transacting. Those authorities tend strongly to support our conclusion that plaintiffs did not manifest assent to SmartDownload’s license terms.
Shrinkwrap Licensing and Related Practices
Defendants cite certain well-known cases involving shrinkwrap licensing and related commercial practices in support of their contention that plaintiffs became bound by the SmartDownload license terms by virtue of inquiry notice. For example, in Hill v. Gateway 2000, Inc., 105 F.3d 1147 (7th Cir.1997), the Seventh Circuit held that where a purchaser had ordered a computer over the telephone, received the order in a shipped box containing the computer along with printed contract terms, and did not return the computer within the thirty days required by the terms, the purchaser was bound by the contract. In ProCD, Inc. v. Zeidenberg, the same court held that where an individual purchased software in a box containing license terms which were displayed on the computer screen every time the user executed the software program, the user had sufficient opportunity to review the terms and to return the software, and so was contractually bound after retaining the product.
These cases do not help defendants. To the extent that they hold that the purchaser of a computer or tangible software is contractually bound after failing to object to printed license terms provided with the product, Hill and Brower do not differ markedly from the cases involving traditional paper contracting discussed in the previous section. Insofar as the purchaser in ProCD was confronted with conspicuous, mandatory license terms every time he ran the software on his computer, that case actually undermines defendants’ contention that downloading in the absence of conspicuous terms is an act that binds plaintiffs to those terms. In Mortenson, the full text of license terms was printed on each sealed diskette envelope inside the software box, printed again on the inside cover of the user manual, and notice of the terms appeared on the computer screen every time the purchaser executed the program. In sum, the foregoing cases are clearly distinguishable from the facts of the present action.
Online Transactions
Cases in which courts have found contracts arising from Internet use do not assist defendants, because in those circumstances there was much clearer notice than in the present case that a user’s act would manifest assent to contract terms …
After reviewing the California common law and other relevant legal authority, we conclude that under the circumstances here, plaintiffs’ downloading of SmartDownload did not constitute acceptance of defendants’ license terms. Reasonably conspicuous notice of the existence of contract terms and unambiguous manifestation of assent to those terms by consumers are essential if electronic bargaining is to have integrity and credibility. We hold that a reasonably prudent offeree in plaintiffs’ position would not have known or learned, prior to acting on the invitation to download, of the reference to SmartDownload’s license terms hidden below the “Download” button on the next screen. We affirm the district court’s conclusion that the user plaintiffs, including Fagan, are not bound by the arbitration clause contained in those terms.
For the foregoing reasons, we affirm the district court’s denial of defendants’ motion to compel arbitration and to stay court proceedings.
Notes and Questions
1. A victory for browsewrap. Despite the holding in Specht, two years later the Second Circuit upheld the enforceability of a browsewrap agreement. In Register.com v. Verio, Inc., 356 F.3d 393 (2d Cir. 2004), the court found that a user’s downloading of factual data from a website was sufficient to indicate its assent to the site’s online terms of use. The Register site enabled users to access the Internet’s centralized “WHOIS” database, which contains information relating to the identity of Internet domain name registrants. A user making a WHOIS query through The Register site would receive a reply furnishing the requested WHOIS information, accompanied by a legend stating that: “By submitting a WHOIS query, you agree that you will use this data only for lawful purposes and that under no circumstances will you use this data to … support the transmission of mass unsolicited, commercial advertising or solicitation via email” (i.e., The Register sought to prohibit the use of WHOIS data to fuel email spam). Importantly, this notice arrived after the user submitted its WHOIS request. The court analyzed The Register’s online agreement as follows:
Verio contends that in no instance did it receive legally enforceable notice of the conditions Register intended to impose. If Verio had submitted only one query, or even if it had submitted only a few sporadic queries, that would give considerable force to its contention that it obtained the WHOIS data without being conscious that Register intended to impose conditions, and without being deemed to have accepted Register’s conditions. But Verio was daily submitting numerous queries, each of which resulted in its receiving notice of the terms Register exacted. Furthermore, Verio admits that it knew perfectly well what terms Register demanded. Verio’s argument fails. The situation might be compared to one in which plaintiff P maintains a roadside fruit stand displaying bins of apples. A visitor, defendant D, takes an apple and bites into it. As D turns to leave, D sees a sign, visible only as one turns to exit, which says “Apples – 50 cents apiece.”
D does not pay for the apple. Thereafter, each day, several times a day, D revisits the stand, takes an apple, and eats it. D never leaves money. In our view, however, D cannot continue on a daily basis to take apples for free, knowing full well that P is offering them only in exchange for 50 cents in compensation. Verio’s circumstance is effectively the same. Each day Verio repeatedly enters Register’s computers and takes that day’s new WHOIS data. Each day upon receiving the requested data, Verio receives Register’s notice of the terms on which it makes the data available – that the data not be used for mass solicitation via direct mail, email, or telephone. Verio acknowledges that it continued drawing the data from Register’s computers with full knowledge that Register offered access subject to these restrictions. Verio is no more free to take Register’s data without being bound by the terms on which Register offers it, than D was free, in the example, once he became aware of the terms of P’s offer. We recognize that contract offers on the Internet often require the offeree to click on an “I agree” icon. And no doubt, in many circumstances, such a statement of agreement by the offeree is essential to the formation of a contract. But not in all circumstances. [It] is standard contract doctrine that when a benefit is offered subject to stated conditions, and the offeree makes a decision to take the benefit with knowledge of the terms of the offer, the taking constitutes an acceptance of the terms, which accordingly become binding on the offeree.
Do you agree with the court’s analysis? What do you think of the court’s $0.50 apple analogy? Do you think that Verio had more or less knowledge of the applicable restrictions than an ordinary user of The Register site? How does this case accord with Specht? Could Netscape have prevailed on a similar theory?
2. Browsewrap today. Courts remain divided over the enforceability of browsewrap agreements. Those following Specht have generally found that, even without multiple or repeat transactions, “the enforceability of browsewrap agreements depends upon whether ‘there is evidence that the user has actual or constructive notice of the site’s terms.’” Mohammed v. Uber Techs., Inc., 237 F. Supp. 3d 719, 731 n.8 (N.D. Ill. 2017). Generally, this “actual or constructive notice” should occur before the user begins to use the site in question.
In the end, the assessment of browsewrap agreements often boils down to a question of website design and layout. One court derived the following “general principles” from the growing body of case law on this subject:
First, “terms of use” will not be enforced where there is no evidence that the website users had notice of the agreement
Second, “terms of use” will be enforced when a user is encouraged by the design and content of the website and the agreement’s webpage to examine the terms clearly available through hyperlinkage
Third, “terms of use” will not be enforced where the link to a website’s terms is buried at the bottom of a webpage or tucked away in obscure corners of the website where users are unlikely to see it.Footnote 7
Why are website design and layout so important to the enforceability of online agreements? Do lawyers now need to become familiar with graphical design principles in addition to contract law, or is graphical design now an integral part of contract law?
3. Feels like paper? Some courts continue to analyze electronic contracts as though they were electronic versions of paper contracts. For example, the court in Hubbert v. Dell Corp., 844 N.E.2d 965, 968 (Ill. 2006) considered the enforceability of online terms that included hyperlinks to numerous other documents. It held that hyperlinks “should be treated the same as a multipage written paper contract. The blue hyperlink simply takes a person to another page of the contract, similar to turning the page of a written paper contract.” Do you agree? How often do you click through the linked documents in online terms?
In contrast, the ABA Committee argues that electronic and paper contracts are inherently different:
To equate digital contracts with paper contracts is to ignore the difference that tangibility makes. The recipient of a paper contract is more likely to skim the pages for capitalized and bolded terms than is a recipient of an electronic contract with terms that remain hidden until the hyperlink is clicked.Footnote 8
Which view do you find more persuasive? In the end, does it matter whether electronic contracts can be analogized to their paper counterparts?
4. Automated scraping and acceptance. The legal analysis of shrinkwrap and browsewrap agreements depends on a finding that there was sufficient assent to support contract formation. Yet what happens when both the presentation of the agreement and its “acceptance” are accomplished without human intervention? The Internet today teems with automated programs, agents, bots and spiders that crawl across billions of webpages collecting (“scraping”), compiling and analyzing information for their creators. Can these automated devices “assent” to a website’s terms of use? Many websites contain prohibitions on automated access in their online terms of use, but are these prohibitions enforceable against an automated bot or spider? Some website operators have argued that, even absent contractual assent, unauthorized access to a website may be prohibited by the Computer Fraud and Abuse Act, 18 U.S.C. § 1030, or common law doctrines such as trespass.Footnote 9
5. Contracting authority. One recurrent issue with online agreements is whether the person purporting to agree to the terms of the license has the legal authority to enter into the agreement. For example, a five-year-old can easily click “I ACCEPT” on a computer screen, but lacks the requisite legal capacity to make a contract. Likewise, if a low-level employee of a company purports to bind his or her company pursuant to a clickwrap agreement when downloading a piece of software, is the company legally bound? Does it matter whether the software is a $0.99 app or a $50 million enterprise resource management system? What duty does the licensor have to ensure that the person on the other side of the click actually has authority to bind the licensee?
It may be for this reason that companies like IBM have adopted across-the-board policies prohibiting their employees from clicking to accept any agreement in connection with their work duties. Do you think this approach is effective? What drawbacks might a company implementing such a policy face?
Courts have wrestled with the issue of apparent authority and clickwrap agreements. See National Auto Lenders, Inc. v. SysLOCATE, Inc., 686 F. Supp. 2d 1318 (S.D.Fla. 2010), in which the licensee (NAL) informed a software vendor (SysLOCATE) that only its executives could make decisions on behalf of their company. As a result, when lower-level employees clicked to accept a software agreement proffered on the vendor’s website, the court found that provisions in the software agreement requiring disputes to be resolved through arbitration were not enforceable.
6. Supersedure. If clickwrap, browsewrap and other electronic terms are considered to be binding contracts, then they can supersede prior written agreements, including agreements that were negotiated and signed by the parties. Why might this effect be considered risky by some parties? What might a party do in order to avoid this risk? For a hint, see Section 13.10, dealing with the precedence of agreements.
17.3 The (D)Evolution of Consumer Licenses
Firms and their attorneys do not observe developments in the law of technology licensing passively. Rather, as the below academic study demonstrates, they adapt their agreements to take new legal developments into account.
Florencia Marotta-Wurgler and Robert Taylor, 88 NYU L. Rev. 240 (2013)
In this Article, we examine the innovation and evolution of a common type of mass-market consumer standard, End User License Agreements (EULAs). EULAs are an important type of online standard-form contract and have been at the forefront of various regulatory debates. Recently, the American Law Institute approved the Principles of the Law of Software Contracts (Law of Software Contracts), which focuses in large part on mass-market transactions involving EULAs. We use a sample of EULAs from 264 mass-market software firms between 2003 and 2010 to track changes to thirty-two common contractual terms. Our methodology measures the relative buyer-friendliness of each term relative to the default rules of Article 2 of the Uniform Commercial Code (U.C.C.) to examine how the pro-seller bias of EULAs changes over time. Since buyers need to become informed about terms to “shop” around effectively, we measure changes in contract length and readability. We begin exploring the firm, product, and market characteristics that are associated with contract changes. Finally, we record relevant court decisions around the sample period to evaluate whether the sample contracts are sensitive to changes in the enforceability of terms.
There are a number of interesting results. Thirty-nine percent of the sample firms made material changes to their contracts during the seven-year period, despite the fact that the product being licensed was held as constant as possible. While there is no absolute baseline against which to measure contract stickiness, our results contrast with the high degree of standardization and stickiness that has generally been observed in sovereign-bond contracting. In our study, a material change occurs when a EULA changes at least one of the thirty-two terms that we track. The list of terms is fairly comprehensive, as explained in Part II. Contracts have also gotten considerably longer on average but no easier to read; despite being ostensibly written for the consumer, the average license agreement remains, by standard textual analysis criteria, as hard to read as an article in a scientific journal. Increased contract complexity over time is problematic in this context because it increases the cost of becoming informed, which, in the absence of intermediaries who can simplify information, might weaken a market disciplining mechanism.
We find that most of the terms that changed have become more pro-seller relative to the original contract. Most of these changes are driven by firms opting out of U.C.C. Article 2 default rules in favor of relatively more pro-seller terms. Clauses that changed the most (in that they have become relatively more pervasive) are forum-selection and arbitration clauses, restrictions on reverse engineering, and restrictions on transfer. While most terms are likely to change away from the default rules, terms that are more pro-seller relative to the default rules are almost twice as likely to change away from those defaults as terms that benefit buyers, all else being equal. That is, pro-buyer defaults are relatively less sticky than pro-seller defaults. We also document innovations, as new and largely pro-seller terms have been introduced even in the absence of strong property rights. In particular, seven terms that were virtually absent in 2003 emerged by 2010. These relate to remote disablement of software, firms’ ability to collect user information, and terms related to the rights and software of third parties. Most of these new terms allow sellers to increase control over users, which is possible because of technological innovation. What parties are associated with change? We find that younger, growing, and large firms, as well as firms with legal departments, are more likely to innovate. We hypothesize that young and growing firms might be more sophisticated and ambitious, and thus more willing to experiment. We test the hypothesis that contract changes might have been shaped by increased legal certainty on the enforceability of such terms. We find that the terms that have become more enforceable during the sample period were more likely to be used in a pro-seller sense, consistent with this hypothesis.
Number of terms changed, 2003 vs. 2010.
Next, we explore the appearance and adoption of innovative terms. We identified seven terms that were rare or absent at the beginning of the period and fell into the three categories of modification and termination, information collection, and third parties. Terms allowing the drafter to unilaterally modify the agreement are examples of changes borrowed from other areas, such as credit card agreements and online Terms of Use. Terms that define the relationship between the user and third parties are innovations in the narrower sense of the term, as these terms allow software providers to contract out some of the functionalities of their products, arguably to parties who can provide them in a better way at a lower cost. Most of these terms take advantage of technological changes (such as electronic licensing) that allow sellers to exercise more control over buyers’ use of the product. As explained above, we do not mean to imply that the terms that we designate “innovative” are economically efficient or good in any welfare sense. All we can say for sure is that they are novel.
Who are the innovators and who are those who adopt the terms later on? Controlling for contract length, the results show that young and larger companies are more likely to adopt innovative terms. A possible explanation for this finding is that larger firms have more resources and are thus more likely to be aware of technological changes that present opportunities to revise EULAs, or that these firms receive more cutting-edge legal advice. Younger firms might be more sophisticated and also more attuned to technological innovations.
We [also] explore the role of in-house counsel in the evolution of fine print … In both 2003 and 2010, the presence of lawyers is associated with more pro-seller bias. Again, lawyers are associated not with change in terms per se, but with a negative change in bias over the sample period. Of course, firm size and the presence of legal counsel are highly correlated, so it might be hard to identify the contribution of legal counsel to change in terms. We assume that firms with legal departments are likely to assign the job of revising and drafting terms to lawyers.
[L]awyers are also associated with innovation, as firms with lawyers are more likely to adopt innovative terms at the beginning of the sample period. [The data] shows no effect between the presence of lawyers and adoption of the innovative terms at the end of the period. This might be because such firms adopted them earlier. Firms without legal departments might look at the contracts of other firms and copy the innovative terms. This possibility is consistent with accounts of various firms in the sample with whom we communicated. In contrast to previous studies, we find that lawyers (at least those who work in-house) appear to be involved in revising and innovating in mass-market agreements.
Conventional wisdom suggests that standard-form contracts are essentially static given that they are rarely invoked, govern relatively low-price items that are unlikely to be the source of litigation, and are not protected by property rights. This study finds change and innovation in several aspects of common consumer standard-form contracts. Contrary to studies of innovation in law firms, it finds that in-house lawyers are associated with new terms. Almost forty percent of the contracts we examined saw at least one standard term change over the period between 2003 and 2010; some changed more than ten terms. While this number could be perceived as low, especially in an industry as dynamic as software, the results challenge conventional views that a large fraction of consumer fine print is set in stone. We find that contracts have become longer but no simpler to read. On average, EULAs accumulate more terms over time, a process consistent with the observation that the process of contract creation involves the overlaying of terms without much revision. Drafters might be thinking myopically about the effect of the particular term being added as opposed to the meaning of the contract as a whole. The implication of this trend is that, to the extent consumers read terms to comparison shop, the cost of becoming informed about terms has increased. The cost is also higher for would-be intermediaries such as ratings websites and consumer nonprofits. An important implication of this is that proposals for increased contract disclosure are less likely to be effective because what is increasingly costly for consumers is not gaining access to the contract but reading it. Any type of disclosure reform might be more effective if it included directives for plainer and more succinct language. Consumer advocates, who have been lobbying for plain-language laws in consumer agreements for some time, may have picked up this trend.
Notes and Questions
1. Directional evolution. Is it a surprise that EULA terms have steadily grown more pro-seller over the years? Under what circumstances might new pro-consumer terms become ingrained in EULAs?
2. Does anybody read the fine print? What do you think about the usefulness of EULA terms given statements like those of Chief Justice John G. Roberts of the US Supreme Court in response to a question during a 2010 speech:
Roberts admitted he doesn’t usually read the computer jargon that is a condition of accessing websites … “It is a problem,” he added, “because the legal system obviously is to blame for that.” Providing too much information defeats the purpose of disclosure, since no one reads it, he said. “What the answer is,” he said, “I don’t know.”Footnote 10
What is “the answer” in your opinion?
Chief Justice Roberts admits that he doesn’t read the fine print …
3. A thicket of restrictions? Much information that is publicly available on the Web is now used for epidemiological and public health research. Genetic epidemiologists have made important discoveries using automated web crawling techniques (see Section 17.2, Note 4) to gather information from tens of millions of individual genealogy records contained online. Yet, as one recent study has found, many genealogy websites contain terms of use that place numerous restrictions on this publicly accessible data.Footnote 11 Restrictions include:
genealogical use only: limits usage to personal, private or professional genealogical use;
no commercial use: prohibits any commercial use of content;
no downloads: prohibits downloading all or significant portions of other users’ content;
no automated access: prohibits automated scraping, crawling and/or harvesting of content;
no transfer: prohibits unauthorized distribution, reproduction, retransmission, publication, sale, exploitation (commercial or otherwise) or any other form of transfer of any portion of the content;
Most of these restrictions, individually and in combination, appear to prohibit scientific research. Yet, it seems that most researchers are unaware of, or do not understand, such restrictions and, to date, no such restrictions appear to have been enforced against biomedical researchers. Is there an issue here? Would the answer change if more website operators began to enforce their online terms? What do you think about a system in which legally enforceable online terms exist but are widely ignored and seldom enforced?
4. Unilateral modification. One of the contractual terms identified by Marotta-Wurgler and Taylor is the licensor’s unilateral ability to amend the terms of the contract. Think about it. Would you knowingly agree to a contract in which the other party could unilaterally amend the terms simply by notifying you? The very idea sounds absurd, but unilateral amendments to consumer contracts are now pervasive. In a recent article, Shmuel Becher and Uri Benoliel report that of 500 browsewrap agreements used with popular US websites, 81.6 percent could be modified unilaterally by the licensor.Footnote 12 They find that “[c]ommon modifications include, for example, a change in fees, a modification of a dispute resolution clause, or revision of the firm’s privacy policy. In fact, unilateral modifications can address virtually every aspect of a contract.” Does the realization that every aspect of an online contract is malleable give you pause? Does it make such instruments less than contracts? Should it?
Courts appear to be divided over the enforceability of contract terms that are unilaterally amended, as well as the general principle of unilateral amendment. At least one court has held that terms of service that permitted a provider to amend the terms at any time rendered the entire contract illusory.Footnote 13 Do you agree? Why shouldn’t a party be entitled to agree that terms may be amended by the other party in the future?
5. The innovations of lawyers. Marotta-Wurgler and Taylor observe that the “innovation” in online contracting terms has largely been driven not by changes in technology or product offerings, but by lawyers. In your view, has this degree of legal innovation helped or hindered the marketplace?
Problem 17.1
Find an online EULA on your computer, tablet or phone. How long is it? Read it. What terms does it contain that surprise you? Would you have agreed to these terms if you were negotiating the agreement in person, or on behalf of a client?
Now put the shoe on the other foot. If you represented the company that wrote the EULA, what additional terms might be beneficial for your client that you could include in the agreement? How close to the line of unconscionability would you advise your client to venture?
Open access