Disciplining standard form contract terms through online information flows: An empirical study

New York University Law Review

Volumn 85, Issue 5, 2010, Pages 1618-1660

  Chari, Nishanth V ^a,b  

^a NEW YORK UNIVERSITY SCHOOL OF LAW   (United States)

^b UNIVERSITY OF CALIFORNIA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 79251644752     PISSN: 00287881     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (259)

1. W. David Slawson, Standard Form Contracts and Democratic Control of Lawmaking Power, 84 HARV. L. REV. 529, 529 (1971). For an encyclopedic review of the scholarship regarding standard form contracts.
2. Clayton P. Gillette, Standard Form Contracts (N.Y. Univ. Sch. of Law Ctr. for Law, Econs. & Org., Working Paper No. 09-18, 2009), available at http://ssrn.com/abstract=1374990.
3. Florencia Marotta-Wurgler, What's in a Standard Form Contract? An Empirical Analysis of Software License Agreements, 4 J. EMPIRICAL LEGAL STUD. 677, 678 (2007).
4. Brower v. Gateway 2000, Inc., 676 N.Y.S.2d 569, 572-73 (App. Div. 1998) (enforcing terms shipped in box with computer equipment).
5. Click wrap agreements are standard terms to which a buyer must click "I agree" prior to purchasing a particular piece of software; browse wrap agreements are agreements not directly presented to a consumer but available for viewing-for example, privacy policies and terms of use on websites.
6. Mark A. Lemley, Terms of Use, 91 MINN. L. REV. 459+459-460 (2006).
7. id. at 459 & n.2 (citing cases and noting that every court scrutinizing clickwrap agreements has found them enforceable);
8. id. at 462-63 & n.7 (citing cases and noting that courts have occasionally rejected browsewrap agreements to protect unwitting consumers but have generally upheld those entered into by sophisticated commercial entities in competition with drafter).
9. Infra Appendix Table 1 for examples of the variation in common contract terms. These types of SFCs are known as rolling contracts.
10. Clayton P. Gillette, Rolling Contracts as an Agency Problem, 2004 WIS. L. REV. 679+681 (discussing contracts presenting terms to buyers concurrently with delivery or first use of goods).
11. Margaret Jane Radin, Boilerplate Today: The Rise of Modularity and the Waning of Consent, 104 MICH. L. REV. 1223+1225 (2006);
12. Friedrich Kessler, Contracts of Adhesion-Some Thoughts About Freedom of Contract, 43 COLUM. L. REV. 629+640 (1943) ("Standard contracts in particular could thus become effective instruments in the hands of powerful industrial and commercial overlords enabling them to impose a new feudal order of their own making upon a host of vassals.");
13. Russell Korobkin, Bounded Rationality, Standard Form Contracts, and Unconscionability, 70 U. CHI. L. REV. 1203+1205-1206 (2003) ("[Sellers] have an incentive to make attributes buyers do not consider ('non-salient' attributes) favorable to themselves....").
14. Infra Part I.A (describing evolution of scholarly approaches ultimately embracing SFCs as market-driven and efficient).
15. Avery Wiener Katz, Standard Form Contracts, in 3 NEW PALGRAVE DICTIONARY OF ECONOMICS AND THE LAW 505 (Peter Newman ed., 1998).
16. Infra Part I.B (describing Law and Economics (L&E) approach to SFCs, which conditions intervention by courts and legislatures on market failure and assumes that SFC terms reflect tradeoffs with other product features).
17. Shmuel I. Becher & Tal Z. Zarsky, E-Contract Doctrine 2.0: Standard Form Contracting in the Age of Online User Participation, 14 MICH. TELECOMM. & TECH. L. REV. 303+ 315 (2008) ("Some aggrieved consumers... might examine the SFC they originally formed with the vendor to acquaint themselves with their rights and obligations.");
18. Infra Part I.C (summarizing Becher and Zarsky's framework and detailing unique benefits and challenges in online context). A recent empirical study confirms that, at the very least, the number of consumers who read SFCs ex post far exceeds the number who read such contracts ex ante.
19. Shmuel I. Becher & Esther Unger-Aviram, The Law of Standard Form Contracts: Misguided Intuitions and Suggestions for Reconstruction 13-14 (Aug. 7, 2009) (unpublished manuscript), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1443908 (presenting survey data reflecting that "a significantly larger proportion of consumers... report intent to read the contract ex post (rather than ex ante)").
20. Yannis Bakos & Chrysanthos Dellarocas, Cooperation Without Enforcement? A Comparative Analysis of Litigation and Online Reputation as Quality Assurance Mechanisms 2 (MIT Sloan Sch. of Mgmt., Working Paper No. 4295-03, 2003), available at http://ssrn.com/abstract=393041 ("Information technology is having dramatic impacts on the cost, scale and performance of reputation mechanisms.").
21. Becher & Zarsky, supra note 12, at 306 (discussing potential of "blogs, virtual communities, and places of virtual congregation [to] promote efficient and fair... contracting");
22. id. at 351 & n.208 (citing Epinions.com (Epinions) as feedback system that aggregates consumers' numerical rankings "into one overall number (or grade) that accounts for all the users").
23. "Bias" means the extent to which the contract terms favor buyers or sellers.
24. Becher & Zarsky, supra note 12, at 315 (noting that travelers prevented from changing trip dates may complain about airline carrier's "lack of flexibility" when this actually pertains to specific contractual provisions). For example, one reviewer of Alsoft's Disk Warrior on Epinions indicated that she was denied a refund when she discovered that the product would not work on her Mac-a right that most likely was withheld in the SFC.
25. Ouduke, Alsoft Sucks-Alsoft Disk Warrior Full Version for Mac (WDD104), EPINIONS (July 11, 2006), http://www99.epinions.com/review/pr-Alsoft_DiskWarrior_For_Mac/content_2 41507208836 (complaining that company "never warned me that this product doesn't work on Intel Macs, and when I asked for a refund (before they even shipped it!!) I was told no").
26. Of course, in order to discipline sellers, it would also be necessary for ex ante consumers to read and internalize the information presented-an empirical question that is beyond the scope of this.
27. Multiple regression analysis is a statistical tool designed to explain the relationship between a dependent variable and one or more explanatory variables.
28. Daniel L. Rubinfeld, Reference Guide on Multiple Regression, in FEDERAL JUDICIAL CENTER REFERENCE MANUAL ON SCIENTIFIC EVIDENCE 179 (2d ed. 2000).
29. Marotta-Wurgler, supra note 2, at 689-702;
30. Infra Part II.A (providing background on software license agreements and related scholarship);
31. Infra Part II.D.1 (detailing Marotta-Wurgler's contract bias methodology, which compares twenty-three terms to UCC default rules and assigns positive or negative points accordingly).
32. EPINIONS, http://www.epinions.com (last visited Sept. 20, 2010);
33. Infra Part II.B.1 (providing background on Epinions);
34. Infra Part II.D.2 (describing product rating system featured on Epinions).
35. AMAZON, http://www.amazon.com (last visited Sept. 20, 2010);
36. Infra Part II.B.2 (providing background on Amazon);
37. Infra Part II.D.2 (describing product rating system featured on Amazon).
38. The Top 500 List, INTERNET RETAILER, http://www.internetretailer.com/top500/list/ (last visited Sept. 20, 2010).
39. Becher & Zarsky, supra note 12, at 306;
40. Infra Part I.C (outlining Becher and Zarsky's theory).
41. Becher & Zarsky, supra note 12, at 308.
42. RESTATEMENT (SECOND) OF CONTRACTS § 17 (1981) ("[T]he formation of a contract requires a bargain in which there is a manifestation of mutual assent to the exchange and a consideration.").
43. SFCs are drafted in advance by the seller, who is typically the party with greater market power and sophistication.
44. Michael I. Meyerson, The Efficient Consumer Form Contract: Law and Economics Meets the Real World, 24 GA. L. REV. 583, 594-595 (1990) (describing SFCs as adhesion contracts drafted by seller). In this sense, they are rarely the outcome of bargaining.
45. Shmuel I. Becher, A "Fair Contracts" Approval Mechanism: Reconciling Consumer Contracts and Conventional Contract Law, 42 U. MICH. J.L. REFORM 747+748 (2009). SFCs governing online purchases are not presented by an in-person representative of the seller, and even in-person agents of sellers are typically not empowered to make changes to the content of SFCs.
46. Becher & Zarsky, supra note 12, at 308.
47. Jason Scott Johnston, The Return of Bargain: An Economic Theory of How Standard-Form Contracts Enable Cooperative Negotiation Between Businesses and Consumers, 104 MICH. L. REV. 857+865 (2006) (arguing that, while contractual forms are standard, firms give certain employees discretion to depart from SFC terms and to deliver more than promised if in firm's best interest).
48. RESTATEMENT (SECOND) OF CONTRACTS § 211(2) (1981) ("[Standardized agreements are] interpreted wherever reasonable as treating alike all those similarly situated, without regard to their knowledge or understanding of the standard terms of the writing." (emphasis added)).
49. For some SFCs known as "rolling contracts," the terms are presented only after the user has purchased the product, in which case the consumer's choice is to accept the terms or refrain from using the product and return it to the retailer, if permitted. Gillette, supra note 7, at 681.
50. Infra notes 41-42 and accompanying text (describing scholarly consensus that buyers do not read contracts).
51. Lemley, supra note 4, at 465 ("[I]n today's electronic environment, the requirement of assent has withered away to the point where a majority of courts now reject any requirement that a party take any action at all demonstrating agreement to or even awareness of terms in order to be bound by those terms.").
52. Kessler, supra note 8, at 632. By the mid-1970s, scholars and courts had come to view SFCs as a form of fraud perpetrated by sellers.
53. W. David Slawson, Mass Contracts: Lawful Fraud in California, 48 S. CAL. L. REV. 1-2 (1974) ("Sellers often mislead buyers into expecting more than the law will say that the buyers are entitled to and then, with impunity, deliver to them even less than this.");
54. Henningsen v. Bloomfield Motors, Inc., 161 A.2d 69, 87 (N.J. 1960) ("Because there is no competition among the motor vehicle manufacturers with respect to the scope of protection guaranteed to the buyer, there is no incentive on their part to stimulate good will in that field of public relations.");.
55. William C. Whitford, Law and the Consumer Transaction: A Case Study of the Automobile Warranty, 1968 WIS. L. REV. 1006+1056 Over 50 percent of the [survey] respondents thought [mistakenly that maintenance operations] had to be performed at the selling dealer or at any authorized dealer.
56. [This was likely because] dealers' verbal explanations stress the desirability of having the work performed by the dealer....").
57. Slawson, supra note 1, at 530 (arguing that SFCs are necessary to providing services at reasonable cost in mass production society).
58. Alan Schwartz & Louis L. Wilde, Imperfect Information in Markets for Contract Terms: The Examples of Warranties and Security Interests, 69 VA. L. REV. 1387+1402 (1983) (discussing influences on warranty terms).
59. George L. Priest, A Theory of the Consumer Product Warranty, 90 YALE L.J. 1297+1298-1299 (1981) (describing warranties as allocation of risk between manufacturer and consumer).
60. Robert A. Hillman & Jeffrey J. Rachlinski, Standard-Form Contracting in the Electronic Age, 77 N.Y.U. L. REV. 429+439 (2002) ("Because the best allocation of risks is not likely to vary between businesses within an industry, most businesses will offer terms similar to those offered by their competitors.");
61. Schwartz & Wilde, supra note 33, at 1390-91, 1403 (presenting model in which "the only possible equilibrium in which all the firms charge the same price occurs when that price is competitive" and suggesting that sellers are similarly responsive to consumer demands for warranty protection).
62. 499 U.S. 585, 593-94 (1991) (upholding enforceability of forum-selection clause found in SFC printed on back of consumer cruise-line ticket because "passengers who purchase tickets containing a forum clause like that at issue in this case benefit in the form of reduced fares reflecting the savings that the cruise line enjoys by limiting the flora in which it may be sued").
63. Becher & Zarsky, supra note 12, at 310-11 ("L&E scholars' Non-Intervention Approach (hereinafter NIA) is premised on competitive market forces. These forces allegedly generate a desired equilibrium at which vendors have adequate incentives to draft efficient standardized terms.... [T]he NIA relies on the existence of market pressure, generated by a crucial group of Marginal Consumers.");.
64. Katz, supra note 10, at 505 ("[T]he relevant policy question is whether state regulation of form contracts can improve on the unregulated market."). Under this view, legal intervention encroaches on individual autonomy and smacks of state paternalism, given that contracts represent private arrangements between the contracting parties.
65. Becher & Zarsky, supra note 12, at 310. This view also rests on an assessment of institutional competence: Firms as repeat players are best situated to determine the appropriate content of SFCs, whereas courts lack the expertise to analyze the specific transaction and market, and legislatures cannot readily determine whether a pro-seller clause will be offset by a pro-buyer clause.
66. Clayton P. Gillette, Pre-approved Contracts for Internet Commerce, 42 HOUS. L. REV. 975, 982 (2005) ("Confronted with a clause that in isolation appears to be one-sided and unable to decipher how that clause fits into the contract as a whole, judicial processes are likely to focus on the injuries suffered by a particular consumer and identify apparent unfairness with exploitation.");
67. Gillette, supra note 7, at 713 ("Courts may have difficulty perceiving the manner in which invalidation of a term affects other buyers, including those who would not have found the term oppressive or otherwise would have considered the term reasonable in light of the contract as a whole.");.
68. 717 ("[A]dministrative agencies... cannot readily analyze proposed risk allocations to determine whether a proseller clause is sufficiently offset by a probuyer [clause] elsewhere in the document.");
69. Robert A. Hillman, Rolling Contracts, 71 FORDHAM L. REV. 743+751 (2002) ("[B]etween business people and judges, the former should be more capable of assessing the appropriate content of form contracts because business people inevitably gain understanding and experience through their day today... activities.").
70. Douglas G. Baird, The Boilerplate Puzzle, 104 MICH. L. REV. 933+933 (2006) ("For a higher price, I can buy a computer with a bigger screen.... So too with the attributes that are legal terms.... [For example,] I can buy a service contract that extends the warranty.");
71. Radin, supra note 8, at 1229 ("If a cell phone contains a chip that will fail in one year and also comes with boilerplate exonerating the seller of liability for consequential damages, its market value is affected in exactly the same way by these two features; for economic purposes, they are both just features of the product.");.
72. Lewis A. Kornhauser, Unconscionability in Standard Forms, 64 CAL. L. REV. 1151+1168 (1976) ("The myriad clauses in fine print represent a detailed description of the commodity to which the price... attaches. One buys not just a Chevrolet of a given model year... but also a complex of rights and liabilities, risks and insurance outlined by the purchase order form.");
73. Arthur A. Leff, Contract as Thing, 19 AM. U. L. REV. 131+144-151+155 (1970) (proposing that we "talk[] about contracts as things, as 'part of' the goods being sold"). Some courts appear to have echoed this assumption.
74. Hill v. Gateway 2000, Inc., 105 F.3d 1147, 1149 (7th Cir. 1997) ("Customers as a group are better off when vendors skip costly and ineffectual steps such as telephonic recitation, and use instead a simple approve-or-return device.");
75. Pro CD v. Zeidenberg, 86 F.3d 1447, 1453 (7th Cir. 1996) ("Terms of use are no less part of 'the product' than are the size of the database and the speed with which the software compiles listings.");
76. Digital Equip. Corp. v. Uniq Digital Techs., 73 F.3d 756, 762 (7th Cir. 1996) (""Free' tech support is not free; customers pay for it in the price of the product, and a marginal cost of zero for calling means that it will be overused, and hence can add quite a bit to the price of a computer or a software package.").
77. Radin, supra note 8, at 1231 ("The traditional picture imagines two autonomous wills coming together to express their autonomy by binding themselves reciprocally to a bargain of exchange.");
78. Randy E. Barnett, A Consent Theory of Contract, 86 COLUM. L. REV. 269+272 (1986) (referring to classical view as "will theory," under which "contractual duties are binding because they are freely assumed by those who are required to discharge them");
79. Richard Craswell, Remedies When Contracts Lack Consent: Autonomy and Institutional Competence, 33 OSGOODE HALL L.J. 209, 210-211 (1995) ("[I]f someone signs a contract but does so only because the other party holds a gun to his head, no autonomy theorist would say that the first party had consented in any meaningful way. The same would apply if the first party had been tricked into signing....").
80. This was the Supreme Court's justification for upholding a standard form contract printed on the back of a cruise ticket in Carnival Cruise Lines, 499 U.S. at 593-594.
81. Meyerson, supra note 26, at 595 ("[C]onsumers do not read form contracts, or do not understand the terms, and are thus unaware of their contents.");
82. Todd D. Rakoff, Contracts of Adhesion: An Essay in Reconstruction, 96 HARV. L. REV. 1174+1179 (1983) ("Virtually every scholar who has written about contracts of adhesion has accepted [that the adhering party is unlikely to read the standard terms before signing the document,] and the few empirical studies that have been done have agreed.").
83. Becher & Unger-Aviram, supra note 12, at 13-14 (presenting surveys that support assertion that consumers tend not to read SFCs at time of assenting to contract);
84. Robert A. Hillman, Online Consumer Standard-Form Contracting Practices: A Survey and Discussion of Legal Implications, in CONSUMER PROTECTION IN THE AGE OF THE 'INFORMATION ECONOMY' 283+283 (Jane K. Winn ed., 2006) (surveying ninety-two law students and finding that few buyers read SFCs beyond price and description of goods or services).
85. Korobkin, supra note 8, at 1205-06.
86. Todd D. Rakoff, The Law and Sociology of Boilerplate, 104 MICH. L. REV. 1235+1243 (2006).
87. Baird, supra note 38, at 936 ("The sophisticated buyer provides protection for those that are entirely ignorant.");
88. Gillette, supra note 7, at 691 (noting that under competitive conditions "the presence of reading buyers can serve as a proxy for nonreading buyers");
89. Alan Schwartz & Louis L. Wilde, Intervening in Markets on the Basis of Imperfect Information: A Legal and Economic Analysis, 127 U. PA. L. REV. 630+637-638 (1979) (suggesting that competition for informed consumers can "generate optimal prices and terms for all consumers").
90. Gillette, supra note 7, at 692 (noting that "high-end computer[] [buyers] are likely to be more dependent on their computers and more attentive to the terms that allocate risks concerning the computer's performance" than low-end buyers, such that "[s]ellers could offer different terms with computers of different levels of sophistication");.
91. id. at 694 (noting that "[r]eading buyers may also be imperfect surrogates for nonreading buyers" because "[r]eading buyers that purchase goods in high volume, for instance, may not need the class-action device, which is displaced by arbitration, to justify seeking redress for defects in goods").
92. Korobkin, supra note 8, at 1206 (arguing that buyers "take into account only a limited number of product attributes and ignore others," such that sellers have incentive to make "non-salient" attributes favorable to themselves).
93. Oren Bar-Gill, Seduction by Plastic, 98 NW. U. L. REV. 1373+1407 (2004) (asserting that credit-card holders' tendency to undervalue penalty clauses allows issuers to set higher penalties).
94. Hillman & Rachlinski, supra note 35, at 481, 484.
95. Specifically, consumers may feel a sense of precontractual investment of time and effort in shopping that discourages them from reading contract terms that might prevent them from buying the product, or they may worry that gathering too much information will impair their decisionmaking process.
96. Id.
97. Yannis Bakos, Florencia Marotta-Wurgler & David R. Trossen, Does Anyone Read the Fine Print? Testing a Law and Economics Approach to Standard Form Contracts 3 (N.Y. Univ. Law & Econ. Research Paper Series, Paper No. 09-40, 2009), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1443256 (tracking Internet browsing behavior and finding that only one out of one thousand retail software shoppers even view license agreements for at least one second).
98. Becher & Zarsky, supra note 12, at 306.
99. Id.
100. Id. at 320.
101. Id. Specifically, online SFCs are often located on vendors' websites, as opposed to physical documents that may be easily misplaced, though Becher and Zarsky acknowledge that vendors may make it difficult to find the SFC if they keep the relevant link hidden on the website.
102. Id. at 320-21.
103. However, the fact that most software products-and certainly all of the products in my sample-present the license agreement during installation and automatically save a copy of the agreement in the installed folder diminishes the impact of this type of vendor behavior.
104. Id. at 315-16.
105. Such cognitive errors include: information overload given the complexity of SFC terms; purchasing environments characterized by noise, time constraints, and vendor manipulation; a limited ability to evaluate future contingencies; the sunk cost effect; and cognitive dissonance.
106. Id. at 312-14. 56 Id. at 321-33.
107. Id. at 333.
108. Online flora may be particularly susceptible to manipulation by firms given the ability to anonymously post reviews. However, Chrysanthos Dellarocas has noted that, "since the quality of honest content is high, firms will not be able to substantially change consumer beliefs through manipulation.".
109. Chrysanthos Dellarocas, Strategic Manipulation of Internet Opinion Forums, 52 MGMT. SCI. 1577+1590 (2006).
110. Becher & Zarsky, supra note 12, at 336.
111. Id. at 334.
112. As discussed in Part II.B, infra, Amazon allows users to indicate whether a given review is helpful or not, and Epinions allows users to rate the helpfulness of reviews.
113. Customer Reviews, Microsoft Office Professional 2007 Full Version, AMAZON, http://www.amazon.com/Microsoft-Office-Professional-2007-VERSION/product -reviews/B000HCVR30/ref=dp_top_cm_cr_acr_txt?ie=UTF8&showViewpoints= 1 (last visited Sept. 20, 2010);
114. Rating Reviews on Epinions, EPINIONS, http://www0.epinions.com/help/faq/show_~faq_rating (last visited Sept. 20, 2010) (encouraging users to rate reviews in order to "help other users determine which reviews to read and which to avoid" and providing guidelines on how to rate reviews).
115. Becher & Zarsky, supra note 12, at 336-37. As discussed in Part II.B, infra, Amazon encourages participation through follow-up emails, and Epinions encourages user participation through monetary compensation programs such as Eroyalties and Income Share.
116. FAQs: Earnings on Epinions.com, EPINIONS, http://www.epinions.com/help/faq/?show=faq_earnings (last visited Sept. 20, 2010) ("Eroyalties credits are the tool to track how much [writers] earn for writing reviews[,]" and "[t]he Income Share program rewards writers who contribute reviews [by paying them] a share of the revenue gained from providing consumers with high-quality information....").
117. Becher & Zarsky, supra note 12, at 320.
118. Existing studies indicate that online product rating websites are important but do not address the prevalence of their use by ex ante consumers.
119. Sylvain Senecal & Jacques Nantel, The Influence of Online Product Recommendations on Consumers' Online Choices, J. RETAILING 159+167 (2004) (finding through targeted behavioral studies that consumers are influenced in their online product choices by online recommendations);
120. Internet & American Life Project, Pew Research Ctr., Online Activities, Total, PEW INTERNET, http://www.pewinternet.org//static-pages/Trend-Data/Online-Activites-Tot al.aspx (last updated Dec. 4, 2009) (finding that 31% of internet users rate products online).
121. In regression analysis, the null hypothesis represents the default assumption that "the results observed in a study with respect to a particular variable are no different from what might have occurred by chance, independent of the effect of that variable."
122. Rubinfeld, supra note 18, at 224.
123. Software license agreements come primarily in the form of "clickwrap" agreements, which consumers sign electronically by clicking "I agree" to a standard set of terms.
124. supra note 4 (defining and distinguishing clickwrap and browsewrap agreements). Many of the software license agreements in my sample are rolling contracts-that is, clickwrap agreements that are not presented prior to purchase and instead appear upon installation. I do not distinguish in my tests between rolling contracts and regular clickwrap agreements for two reasons. First, my study focuses on the product ratings posted by buyers who have already purchased and installed the product. Accordingly, the timing of the presentation of the terms does not affect the analysis because the ex post buyer will have had the opportunity to view the terms regardless. Second, recent empirical work has shown that rolling contracts are not systematically more pro-seller than regular clickwrap agreements.
125. Florencia Marotta-Wurgler, Are 'Pay Now, Terms Later' Contracts Worse for Buyers? Evidence from Software License Agreements, 38 J. LEGAL STUD. 309+341 (2009) (finding that contracts displayed prepurchase are, on average, slightly more pro-seller than rolling contracts).
126. Software: Global Industry Guide, INFO EDGE, http://www.infoedge.com/product_type.asp?product=DO-4959 (last visited Sept. 7, 2010). Software license agreements should also be a particularly accessible example to most readers, given that anyone who owns a personal computer has entered into such an agreement.
127. Becher & Zarsky, supra note 12, at 365-66.
128. Marotta-Wurgler, supra note 64, at 317 ("[D]isagreement among scholars and consumer advocates over the enforcement of software shrink wrap and click wrap licenses have stalled recent efforts to create uniform laws to govern transactions of information goods over the Internet, such as the Uniform Computer Information Transactions Act....").
129. Altera Corp. v. Clear Logic, Inc., 424 F.3d 1079, 1081 (9th Cir. 2005) (affirming lower court interpretation of use restrictions in software license agreement);
130. Davidson & Assocs. v. Internet Gateway, 334 F. Supp. 2d 1164, 1177 (E.D. Mo. 2004) (enforcing software license agreement associated with CD-ROM game);
131. Stenzel v. Dell, Inc., 870 A.2d 133, 144-45 (Me. 2005) (enforcing arbitration clause in software license agreement);
132. 1-A Equip. Co. v. Icode, Inc., No. 0057CV467, 2003 WL 349913, at *2 (Mass. App. Ct. 2003) (enforcing forum selection clause in software license agreement).
133. Marotta-Wurgler, supra note 64, at 322. Several sets of products in my sample even offer similar functionality. For example, Symantec and McAfee offer virus scan software; Sony and Cakewalk offer music production software; H&R Block and Intuit offer tax filing software.
134. Methodologies relying upon an index are not uncommon in the empirical literature.
135. Paul Gompers, Joy Ishii & Andrew Metrick, Corporate Governance and Equity Prices, 118 Q.J. ECON. 107+114-115 (2003) (presenting index in which one point is added for every corporate governance provision restricting shareholder rights);
136. Rafael La Porta et al., Legal Determinants of External Finance, 52 J. FIN. 1131+1134 (1997) (presenting index in which one point is added for corporate anti-director rights).
137. Marotta-Wurgler, supra note 2, at 713.
138. Florencia Marotta-Wurgler, Competition and the Quality of Standard Form Contracts:The Case of Software License Agreements, 5 J. EMPIRICAL LEGAL STUD. 447+475 (2008).
139. Marotta-Wurgler, supra note 64, at 341.
140. Ronald J. Mann and Travis Seibeneicher appear to have adopted a contract bias methodology similar to that employed by Marotta-Wurgler, Focusing on terms that diverge from the UCC Article 2 default rules in ways that favor sellers.
141. Ronald J. Mann & Travis Seibeneicher, Just One Click: The Reality of Internet Retail Contracting, 108 COLUM. L. REV. 984+995 (2008). However, they focus only on nine terms rather than twenty-three, and they look exclusively at terms that favor sellers, without considering terms that favor consumers.
142. Id. at 995-96.
143. Becher, supra note 26, at 767 n.81 (describing Marotta-Wurgler's 2007 article as "an interesting attempt to examine the content of software license contracts");
144. Mann & Seinbeneicher, supra note 74, at 987 n.13 (citing Marotta-Wurgler's 2007 and 2009 articles and employing similar methodology);
145. Zev J. Eigen, The Devil in the Details: The Interrelationship Among Citizenship, Rule of Law and Form-Adhesive Contracts, 41 CONN. L. REV. 381+386 n.7 (2008) (citing Marotta-Wurgler's 2007 article);
146. Theodore Eisenberg & Geoffrey P. Miller, The Flight to New York: An Empirical Study of Choice of Law and Choice of Forum Clauses in Publicly-Held Companies' Contracts, 30 CARDOZO L. REV. 1475+1476-1477 n.3 (2009) (citing Marotta-Wurgler's 2007 and 2009 articles);
147. Robert W. Gomulkiewicz, The Federal Circuit's Licensing Law Jurisprudence: Its Nature and Influence, 84 WASH. L. REV. 199+247 n.290 (2009) (citing Marotta-Wurgler's 2009 article).
148. Becher & Zarsky, supra note 12, at 351 (noting that these aggregates solve information overload problem faced by ex ante buyers). For an empirical model reflecting this difficulty,
149. Kushal Dave, Steve Lawrence & David M. Pennock, Mining the Peanut Gallery: Opinion Extraction and Semantic Classification of Product Reviews, 12 PROC. INT'L CONF. ON WORLD WIDE WEB 519-520 (2003), available at http://portal.acm.org/citation.cfm?id=775152.775226&type=series (follow "Full text" hyperlink).
150. Becher & Zarsky, supra note 12, at 351.
151. id. at 325 n.82, 330 (describing how Amazon allows exchange of ideas and opinions regarding products and services);
152. id. at 351 (describing Epinions as feedback system featuring aggregated numeric ratings of vendors);
153. Clayton P. Gillette, Reputation and Intermediaries in Electronic Commerce, 62 LA. L. REV. 1165+1177-1178 (2002) (noting that websites such as Epinions and eBay allow users of goods and services to post reviews of their experiences);
154. Lior J. Strahilevitz, Reputation Nation: Law in an Era of Ubiquitous Personal Information, 102 NW. U. L. REV. 1667+1706 (2008) (describing Epinions as "clearinghouse" for information about behavior of companies).
155. Chrysanthos Dellarocas, The Digitization of Word of Mouth: Promise and Challenges of Online Feedback Mechanisms, 49 MGMT. SCI. 1407+1408+1413 (2003) (citing Epinions as noteworthy example of online feedback mechanism, and describing Amazon as feedback mechanism that controls form of information by requiring rating on scale of 1-5);
156. Andrew D. Gershoff, Ashesh Mukherjee & Anirban Mukhopadhyay, Consumer Acceptance of Online Agent Advice: Extremity and Positivity Effects, 13 J. CONSUMER PSYCHOL. 161 (2003) (citing Epinions as "ever-expanding database of ratings provided by actual consumers" and analyzing issue of conflicting advice across rating systems);
157. Nolan Miller, Paul Resnick & Richard Zeckhauser, Eliciting Informative Feedback: The Peer Prediction Method, 51 MGMT. SCI. 1359+1359-1361 (2005) (describing honest reporting on reputation systems as Nash equilibrium and citing Epinions and Amazon as examples);
158. Vanitha Swaminathan, The Impact of Recommendation Agents on Consumer Evaluation and Choice: The Moderating Role of Category Risk, Product Complexity, and Consumer Knowledge, 13 J. CONSUMER PSYCHOL. 93+93-94 (2003) (investigating impact of recommendation agents under conditions of category risk and citing Epinions as example of "pure infomediary" based on customer reviews as opposed to expert opinions).
159. Rajat Bhattacharjee & Ashish Goel, Avoiding Ballot Stuffing in eBay-Like Reputation Systems, 2005 PROC. ACM SIGCOMM WORKSHOP ON ECON. PEER-TO-PEER SYS. 133, available at http://portal.acm.org/citation.cfm?id=1080192.1080203 (follow "Full text" hyperlink) (theorizing that transaction costs are necessary to prevent collusion to inflate positive or negative seller reputation on eBay, Amazon and Epinions);
160. Dave, Lawrence & Pennock, supra note 77, at 520 (using Amazon as source of reviews in content-mining experiment);
161. Juha Leino & Kari-Jouko Räihä, Case Amazon: Rating and Reviews as Part of Recommendations, 2007 PROC. ACM CONF. ON RECOMMENDER SYS. 137, available at http://portal.acm.org/citation.cfm?id=1297255 (follow "Full text" hyperlink) (focusing on effect of Amazon's algorithm-based and user-generated recommendations on consumers' ability to find items of interest);
162. Haifeng Liu et al., Predicting Trusts Among Users of Online Communities-an Epinions Case Study, 9 PROC. ACM CONF. ON ELECTRONIC COM. 310 (2008), available at http://portal.acm.org/citation.cfm?id=1386790.1386838 (follow "Full text" hyperlink) (applying taxonomy to predict Epinions users' trust of other users);
163. Paolo Massa & Paolo Avesani, Controversial Users Demand Local Trust Metrics: An Experimental Study on Epinions.com Community, 1 PROC. 20TH NAT'L CONF. ON ARTIFICIAL INTELLIGENCE 121 (2005) (on file with author) (using Epinions trust data to show that existence of controversial user demands personalized ways of predicting trustworthiness of users);
164. John O'Donovan & Barry Smyth, Trust in Recommender Systems, 10 PROC. INT'L CONF. ON INTELLIGENT USER INTERFACES 167 (2005), available at http://portal.acm.org/citation.cfm?id=1040870 (follow "Full text" hyperlink) (discussing trust models built from data provided by users on Epinions).
165. supra notes 57-58 and accompanying text (describing accreditation and motivation problems).
166. Epinions is a pure product review website that exists primarily as a platform on which users can write reviews on various products, though it links to other sites that a consumer may visit to purchase products. By contrast, Amazon is primarily an online retail website that also features product ratings.
167. Infra Part II.D.2 (summarizing product review instructions on Epinions and Amazon, both of which use 5-point rating systems and provide brief descriptions for each rating value).
168. Company Information: About Epinions, EPINIONS, http://www.epinions.com/about (last visited Sept. 20, 2010).
169. Epinions had 4.5 million unique visitors during the month of December 2009. Site Profile for Epinions.com, EPINIONS, http://siteanalytics.compete.com/epinions.com (last visited Sept. 7, 2010).
170. Paul Gil, The Top 10 Most Useful Web Sites, 2010, ABOUT, http://netforbeginners. about.com/od/readerpicks/ss/useful_sites_7.htm (last visited Sept. 20, 2010).
171. About Us, CONSUMER REPORTS, http://www.consumerreports.org/cro/aboutus/index.htm (last visited Sept. 20, 2010).
172. Tara Lane, Epinions, Boing Boing Illustrates Rise in Transparent Brands, SPARXOO (Feb. 3, 2010), http://sparxoo.com/2010/02/03/epinions-boing-boing-illustrates-rise-in-t ransparentbrands.
173. Company Information: About Epinions, EPINIONS, supra note 85.
174. FAQs: Rating Reviews, EPINIONS, http://www.epinions.com/help/faq/?show=faq_rating (last visited Sept. 20, 2010).
175. FAQs: The Web of Trust, EPINIONS, http://www.epinions.com/help/faq/show_~faq_wot (last visited Sept. 20, 2010). The Web of Trust feature has been explored in depth by various economists and behavioral psychologists.
176. sources cited supra note 81.
177. Nicholas Thompson, Technology: More Companies Pay Heed to Their 'Word of Mouse' Reputation, N.Y. TIMES, June 23, 2003, at C4.
178. FAQs: Earnings on Epinions, EPINIONS, supra note 60.
179. The Top 500 List, INTERNET RETAILER, supra note 22. Compete.com indicates that Amazon had 85.9 million unique visitors in December 2009.
180. Site Profile for Amazon.com, COMPETE, http://siteanalytics.compete.com/amazon.com (last visited Sept. 7, 2010). Note that this statistic may not be indicative of the number of visitors that used the site for the purpose of reading reviews as opposed to merely purchasing products.
181. Yubo Chen & Jinhong Xie, Online Consumer Review: Word-of-Mouth as a New Element of Marketing Communication Mix, 54 MGMT. SCI. 477 (2008).
182. Amy Harmon, Amazon Glitch Unmasks War of Reviewers, N.Y. TIMES, Feb. 14, 2004, at A1.
183. Rik Myslewski, User-Generated Reviews-Blessing or Bull?, REGISTER (Jan. 27, 2009, 5:00 AM), http://www.theregister.co.uk/2009/01/27/user_generated_reviews/ (discussing admission by hardware manufacturer Belkin that employee paid rater to write positive, false reviews of Belkin products).
184. Id. (discussing Amazon's policies on reviewing consumer ratings).
185. Customer Reviews, Microsoft Office Professional 2007 Full Version, AMAZON, http://www.amazon.com/Microsoft-Office-Professional-2007-VERSION/product -reviews/B000HCVR30/ref=dp_top_cm_cr_acr_txt?ie=UTF8&showViewpoints= 1 (last visited Sept. 14, 2010);
186. Dave Gussow, Online Opinion Shapers: Everyone's A Critic, ST. PETERSBURG TIMES (St. Petersburg, Fla.), Dec. 13, 2004, at 1D (describing consumer reviewing process).
187. Becher & Zarsky, supra note 12, at 339 & n.162 (describing such encouragement as "endless[] nagging").
188. Several companies' user support representatives refused to release their license terms to me and suggested that I purchase the relevant software in order to view the terms. Surprisingly, I was almost always able to obtain these terms by calling each company's sales hotlines instead.
189. The large revenue firms in my sample have a significant effect on the mean revenue (4.74 billion) considering the fact that the median revenue of 38.15 million is significantly less than the mean and the standard deviation is 16 billion.
190. The median and the mean age of the firms in my sample are both roughly twenty-one years, corresponding to a 1989 year of incorporation, but my sample includes a mix of recent startups and long-established firms.
191. W. Edward Steinmueller, The U.S. Software Industry: An Analysis and Interpretive History, in THE INTERNATIONAL COMPUTER SOFTWARE INDUSTRY 15+30-38 (David C. Mowery ed., 1996) (describing emergence of personal computer at beginning of 1980s and related software revolution).
192. Companies, HOOVER'S, http://www.hoovers.com/companies/100003475-1.html (last visited Sept. 16, 2010).
193. The mean price of the products in the sample is $224.37, with a median price of $79.99. The mean was skewed upward somewhat dramatically by a few very expensive products.
194. To get a sense of the representativeness of my sample, note that there were 994 rated software products on Epinions as of June 2010.
195. Software Product Reviews and Prices, EPINIONS, http://www.epinions.com/Software-~all/sec_~product_list/pp_~1/sort_~rati ng/sort_dir_~des#list (last visited June 15, 2010) (providing all software products sorted by rating).
196. However, this number is likely inflated because dozens of the reviewed products have been discontinued. It is difficult to determine the total number of rated software products in the relevant Amazon categories, but there are likely several thousand.
197. Marotta-Wurgler, supra note 2, at 689-702.
198. Marotta-Wurgler looked to popular texts and trade publications in developing her list of contract terms.
199. id. at 691, 694 (describing authorities).
200. Id. at 695-702 (discussing seven categories of contract terms in detail).
201. Id. at 690.
202. Id. at 691.
203. Id.
204. EPINIONS, supra note 20 (select product, then click on "Write a Review" hyperlink to reach review guidelines).
205. Create Your Own Review, AMAZON, http://www.amazon.com/review/createreview (last visited Sept. 20, 2010).
206. Multiple regression analysis is typically used to analyze data when there are several possible explanations for the relationship among a number of explanatory variables.
207. Rubinfeld, supra note 18, at 181. It is useful for determining whether a particular effect is present as well as for measuring the magnitude of the effect.
208. Id.
209. Becher and Zarsky suggest the following as factors indicating the reliability of online ratings: "the vendor's size, revenue, market share, [and] the length of time it has been operational in the online realm."
210. Becher & Zarsky, supra note 12, at 358.
211. supra note 40 and accompanying text (illustrating L&E tradeoff model forming basis for Supreme Court's decision in Carnival Cruise Lines, Inc. v. Shute, 499 U.S. 585 (1991));
212. Marotta-Wurgler, supra note 64, at 338 ("To the extent that the terms are 'priced,' we should expect that the resulting coefficient [of overall bias] on price would be positive and significant.").
213. This relationship makes sense because if ex post consumers rate a product on the basis of price, we would expect that, between two products with similar functionality, the less expensive product would be more highly rated.
214. Indeed, contrary to this assumption, Marotta-Wurgler has found that the overall effect of contract terms on price is most likely small. Marotta-Wurgler, supra note 64, at 338
215. Marotta-Wurgler, supra note 2, at 708.
216. Id. at 709.
217. Marotta-Wurgler, supra note 64, at 338-39 (describing possible differences in treatment of business-oriented and consumer-oriented products).
218. Id.
219. Fixed-effect models avoid omitted variable bias by examining the relationship between variables within groupings of observations.
220. PAUL D. ALLISON, FIXED EFFECTS REGRESSION METHODS FOR LONGITUDINAL DATA USING SAS 4 (2005) ("Fixed effects methods completely ignore the between-person variation and focus only on the withinperson variation.").
221. The fixed-effect model in this Note controls for potential differences between the consumers of different types of software by running regressions on the rating and bias observations within each product type A normal distribution is one in which 95% of the distribution lies within two standard deviations of the mean.
222. Rubin field, supra note 18, at 224.
223. The regressions are based on the following specification: Ratingi = a0 am + b + Xi + c + Zi + ui. In this equation, the dependent variable is Rating, or the product rating value assigned by ex post consumers. In addition to looking at contract bias, the independent variables are X, or the natural log (ln) of the product price; Z, or firm characteristics including the natural log of revenue and the natural log of age since incorporation; and dummy variables for Amazon software market categories.
224. The natural log is useful when a variable is highly skewed.
225. Mann & Seibeneicher, supra note 74, at 1009 (using log-transformed version of skewed variable in regression).
226. For example, a positive coefficient for contract bias would mean that more pro-buyer contracts (i.e., contracts with more positive bias measurements) are associated with higher ratings. Similarly, a negative coefficient for product price would mean that higher prices are associated with lower ratings.
227. Note, however, that the absence of a statistically significant relationship does not guarantee that there is no relationship; it may result from insufficient data. Rubinfeld, supra note 18, at 185. Thus, these results do not prove the null hypothesis. They merely suggest that we must fail to reject the null.
228. While the magnitude of this effect may seem rather small, it is not negligible given that the star rating values range only from 1 to 5.
229. For Amazon, model 2 shows a coefficient of bias of -0.133 when controlling for price. In model 3, the coefficient is -0.134 when controlling for price and firm revenue. With controls for price, firm revenue and firm age in model 4, the coefficient is -0.129.
230. A very low R2 value is one indication that the multiple regression model leaves an unacceptably large portion of the change in the dependent variable unexplained.
231. Rubinfeld, supra note 18, at 188 n.24.
232. Id. ("[A] low R2 does not necessarily imply a poor model (and vice versa)."). Generally, R2 and adjusted R2 (which adjusts for the number of explanatory terms in a model) are of secondary importance unless the regression equation is being used to predict future outcomes, which is not the purpose of this.
233. Id. at 216 ("Typically, an R2 is low in cross-section studies in which differences in individual behavior are explained. It is likely that these individual differences are caused by many factors that cannot be measured.").
234. supra Part II.D.1 (describing calculation of bias index).
235. Korobkin, supra note 8, at 1206 (describing as "salient" terms that buyers care about and as "non-salient" those that buyers do not).
236. Id. (discussing sellers' incentives with respect to salient and non-salient attributes).
237. supra note 55 and accompanying text.
238. The seven sub-indices include: acceptance of license, scope of license, transfer of license, warranties, limitation of liability, maintenance and support, and conflict resolution.
239. supra note 110 and accompanying text.
240. 499 U.S. 585, 593-94 (1991);
241. supra notes 37-40 and accompanying text (explaining L&E approach to SFCs accepted in Carnival Cruise Lines).
242. supra notes 51-61 and accompanying text (discussing Becher and Zarsky's chokepoints).
243. In my sample, the number of reviews per product on Epinions ranges from 1 to 28, with a mean of 3.63 and a median of 1, whereas the number of reviews per product on Amazon ranges from 1 to 258, with a mean of 34.82 and a median of 12.
244. Becher & Zarsky, supra note 12, at 326 ("A possible impediment to the success of these data flows is that ... they rely on the ex ante consumers' knowledge of these existing hubs of information.").
245. Becher and Zarsky argue, however, that search engine results will present ex ante consumers with flora criticizing sellers and products.
246. Id.
247. supra notes 51-61 and accompanying text.
248. Becher & Zarsky, supra note 12, at 351.
249. Id. at 352.
250. For a previous study summarizing the difficulties with such a data mining effort Dave, Lawrence & Pennock, supra note 77.
251. This chokepoint is important because, as Korobkin has noted, "[e]fficiency requires not only that buyers be aware of the content of form contracts, but also that they fully incorporate that information into their purchase decisions."
252. Korobkin, supra note 8, at 1217-18.
253. For a recent study using clickstream data to refute the informed minority theory Bakos, Marotta-Wurgler & Trossen, supra note 50, which finds that only 0.02% of retail software consumers choose to access SFCs and those that do spend too little time to have carefully read them.
254. This focus is driven by the fact that the fundamental question of L&E scholarship is whether "existing institutional arrangements" may exert market pressure on sellers. Rakoff, supra note 44, at 1243 (emphasis added).
255. Hillman & Rachlinski, supra note 35, at 488-92 (asserting that courts can apply blanket assent where "e-consumers have some opportunity to read the standard terms before deciding whether to enter into the contract"). Hillman has since suggested that, akin to cigarette warning labels, mandatory website disclosure of standard terms may backfire in the short-term by "insulat[ing] businesses from claims of procedural unconscionability," but that disclosure may also have salutary long-term effects in raising public awareness.
256. Robert A. Hillman, Online Boilerplate: Would Mandatory Website Disclosure of E-Standard Terms Backfire?, 104 MICH. L. REV. 837+840+856 (2006).
257. FAQs on Privacy Practices, Seals & Programs, TRUSTE, http://www.truste.com/about_TRUSTe/faqs.html (click on "Consumers FAQs" hyperlink) (last visited June 16, 2010) ("TRUSTe [provides] privacy seals to Web sites that give [consumers] proper notice of its [sic] privacy practices... [O]ur privacy standards meet state and federal laws and exceed industry norms.").
258. supra notes 51-61 and accompanying text (laying out Becher and Zarsky's theory).
259. Becher & Zarsky, supra note 12, at 351 (discussing aggregated reviews as method of obtaining "concise information that encapsulates the entire [standard-form contract]").