A. The Information Problem of Mass Contracting
Let us begin with what everybody knows: Nobody reads boilerplate.Footnote 1 Doing so would be absurd. Walls of text govern many consumer transactions offline, and all interactions on the internet. The opportunity cost of engaging with this information load would be gigantic. Yet at the same time, the benefits would be very slim. The most important reason for this is that most contracts are performed without a hitch.Footnote 2 It is pointless to engage with rules governing a conflict if it is extremely unlikely that a conflict will ever occur. Beyond this, even if one were to sift through the walls of text and find some undesirable clause in the details, one could not change it. Friedrich Kessler, a German born U.S. law professor, is credited with coining the term “take-it-or-leave-it.”Footnote 3 It has been cited and examined by most of the literature on the subject.Footnote 4 It seems that with regard to contracts of adhesion the much invoked serenity prayer has been answered: People accept what they cannot change.
This rational ignorance of the recipient allows the term drafter to unilaterally impose its clauses on the other party. For how can a party object to a disadvantageous clause, if they never become aware of it? Thus, businesses are given an incentive to make the clauses as onerous as possible and use the cost savings to improve product dimensions like price that consumers actually care about,Footnote 5 or internalize the gotten gains. This power to unilaterally determine contract content to the detriment of recipients does not necessarily require market power.Footnote 6 Rather, it follows from the non-existent incentives to read contract terms. The value of most transactions in itself,Footnote 7 and the possibility that something in fact does go wrong, are so low, that any engagement with the contractual framework is inefficient.Footnote 8 For the term setter, who uses the same contract template thousands of times, this is different. One party cares deeply about cost reductions accruing thousands of times, the other party cares little.Footnote 9
Boilerplate contracts are not a new phenomenon. If a merchant likes a contract that has been drafted, why would they not copy and re-use it for all future transactions? Pre-drafted contracts are thus centuries old and have been reported even in Roman law.Footnote 10 They allow for efficiency increases in the provision of services, for drafting contracts is a costly exercise.Footnote 11 Despite this, the apotheosis of the standard form contract mirrors the standardization of services, industries and products commencing in the 19th century.Footnote 12 Standardized contracts are a part of and a pre-requisite for this standardization.Footnote 13 Think of the restrictions imposed on the size of airline baggage. Negotiating the permissible baggage dimensions individually would lead to huge transaction costs acruing to airline and consumer. It would also require that each piece of luggage is individually analyzed as to the respective agreement. With a standardized contract, the airline can simply put a box at the airport in which the flight crew fits—or does not fit—the luggage. The box as a sole arbiter of baggage compliance is not possible without standardized contracts. Standardized contracts are thus not only efficiency promoting, but have rather been a prerequisite of modern service economies for at least a hundred years. It is therefore unsurprising that the first academic examinations of unilaterally drafted and unchangeable contracts are eighty or even a hundred years old.Footnote 14 Although we lack empirical data on readership in the early 20th century, due to the congruent incentives, we can assume that these agreements were not read significantly more than their modern descendants.Footnote 15
The diagnosis of an economy being based on contracting in ignorance is therefore not a new one, yet the advent of the digital economy has put the information problem on steroids.Footnote 16 There are two reasons for this. First, the digital world allows for reproduction of information at almost zero cost, and has therefore removed the few limits that exist in the physical world, like printing costs and the physical limitations of notice boards.Footnote 17 Terms that would seem unfathomably long, and possibly nefarious, when printed are now hidden behind hyperlinks and checkboxes; lengthy notices are made to seem like just another element of an app’s user interface and experience.Footnote 18 The digital economy has also affected the information dynamics of contracting by giving rise to a plethora of new and complex business models, which require standard form contracts in order to develop. While the sale contract for an old Chevy may be scrawled on a cocktail napkin; namely because the deal is uncomplicated and the legal framework is strong and well developed, the same cannot be said for the use of an online service or social media platform, which in turn are usually governed by a barrage of contractual terms. In Europe, a recent trend in regulatory technique adds fuel to the fire. The European Union increasingly relies on the instrument of mandating businesses to regulate and disclose certain issues in their form contracts, in other words, requiring lengthy form contracts by law.Footnote 19 The simpler approach of stating statutory default or even mandatory rules, is eschewed in favor of an approach that makes businesses draft these long contract terms, and later subjecting these terms to court review. Naturally, this increases text length. Thus, the more atomized online services become, the more the online world moves towards service provision instead of pure sale, and the more regulators use the you-set-the-terms-now-we-control-them-later approach, the more pronounced the information problem will become. Although regulatory trends cannot be predicted, the inclusion of software and digital services in more and more products is to be expected.Footnote 20
Due to the technological nature of these heightened challenges, there is the possibility of a deus ex machina solution for this problem.Footnote 21 The newest candidate for a technology that could change the information dynamics of form contracting is Artificial Intelligence (A.I.).Footnote 22 Still, the hopes pinned on A.I. are the same hopes that were previously pinned on the blogosphere,Footnote 23 social media,Footnote 24 the internet of things,Footnote 25 augmented reality,Footnote 26 or machine bargaining.Footnote 27 Of course, the failure of these previous technological champions to solve the information problem gives no insight as to the effects of a watershed technology like A.I. Yet there are reasons to be doubtful as to whether A.I. will solve the information problem of mass contracting. The digital age aggravated the information problem by decreasing information reproduction cost. Artificial Intelligence has so far reduced not only the cost of information reproduction, but the costs of original information production. Artificial Intelligence is already used to reduce contract drafting costs.Footnote 28 Under this view, it seems likely that any A.I. world contains more information in general and contract terms specifically than the human intelligence world. Of course, A.I. can greatly increase consumers’ capabilities to process this information. Yet whether this will overcome the massively-stacked odds of the current information dynamics remains to be seen. It seems possible that a consumer will choose to eschew running A.I. contract-crunching-models over a set of terms, if they are presented on a take-it-or-leave-it basis anyway, leaving us back on square one with Kessler. Although nobody can predict the future, it seems that A.I. will not solve the underlying incentive issues regarding mass-contracts.
The impact of the information problem of mass contracting on the institution of contracts is profound. Freedom of contract, itself understood as the basis of market economies, imagines contracts as an expression of individual liberties, autonomy and self-determination.Footnote 29 In an ideal world, both parties, equipped with full mental capacities, meet on equal footing and thrash out a deal, which they fully and completely understand. Think of Adam Smith’s baker selling a shipment of bread to the brewerFootnote 30, or, as Kessler has put it: “[A] contract is the result of the free bargaining of parties who are brought together by the play of the market and who meet each other on a footing of social and relative economic equality.”Footnote 31 This has a political dimension, allowing individuals to create legal obligations purely based on their will, without government intrusion or judgment.
Saying that this ideal differs from the reality of consumer contracting would be an understatement. Not only have many consumers never concluded such a contract in their entire life.Footnote 32 But it is at odds with this concept that most legal transactions happen in complete ignorance of almost all of their contents. Even earliest commentators have noted that the unilateral imposition of a term on a take-it-or-leave-it basis is not an expression of freedom of contract,Footnote 33 and it has been universally accepted since. Thus, almost all contracts a person concludes in their life, bear little resemblance to the idea of freedom of contract. There is no doubt that almost all terms of almost all consumer transactions are imposed unilaterally. You cannot participate in modern life without assenting to a myriad of take-it-or-leave-it terms.Footnote 34 The connection between the idea of freedom of contract and actual contracts is wafer thin at best,Footnote 35 which is corrosive to the idea of freedom of contract.Footnote 36
Scholars have emphasized that this discrepancy is not actually a problem at all.Footnote 37 A person can easily agree in ignorance, because they are not promising the content of an unread document, but simply stating that they are agreeing to an—unread—promise.Footnote 38 This is undoubtedly true. To refuse this truth would mean that any unread agreement can never create a contractual obligation. Thus, to have any enforcement of any boilerplate term requires a theoretical framework as to how an obligation can be created in ignorance. The earliest German law treatise on boilerplate terms provides this framework, by outlining that assent to unread terms is akin to a declaration for a risky event, and it is within freedom of contracts to take risks.Footnote 39 Every day unread obligations are created and enforced. The issue is not whether this can be done at all, but whether systematically unread documents can create the same legal effect as understood contracts. How much blind risk is the legal order comfortable with?
It is, however, not the sanctity of the institution of contracts that keeps consumer protection advocates up at night. Rather, the information dynamics allow the term setter to skew the contractual balance of rights and obligations in their favor. Because the individual is unaware of the content of the terms, they cannot reject terms that they deem unwanted or inappropriate. As contracting in ignorance is widespread, recipients are also not protected by the power of the market. This would require sufficient readership for the recipients to form an informed minority.Footnote 40 Yet in lieu of such a minority, the unread terms become a non-salient product dimension.Footnote 41 This provides incentives to reduce quality in this dimension, while improving salient product dimensions such as price.Footnote 42 The result is a market failure in which one would expect boilerplate to hide the most egregious terms.
Although the concept of an informed minority is deeply steeped in the Law and Economics approach, the diagnosis of the information problem need not be. Many Law and Economics scholars were able to describe a failure for the market for contract terms through their respective instruments, especially when taking behavioral economics on board. Yet so was everybody else, without reference to economics. Not only has the truth that these take-it-or-leave-it agreements are not read been accepted by “[v]irtually every scholar who has written about contracts of adhesion.”Footnote 43 But also, doctrinal theories of law commonly find there to be some form of deficiency in the contracting process, which needs be compensated by law,Footnote 44 or lead to non-enforcement.Footnote 45 Although the information problem is often phrased in terms of benefits and costs of reading, it is not a problem solely posed, understood, or solved by the economic analysis of law.
Before continuing, some pruning of the terminology is required. On the one hand, in the American literature, the take-it-or-leave-it aspect of mass contracting is described under the term “contract of adhesion.”Footnote 46 The term boilerplate, on the other hand, references the phenomenon of lengthy, pre-drafted and usually contractual text.Footnote 47 Current discussions focus on the issue of online consumer contracts, specifically regarding assent to arbitration clauses.Footnote 48 The German term Allgemeine Geschäftsbedingungen refers to pre-drafted and unchanged contract terms universally imposed by one party onto another, which follows from Section 305 German Civil Code. “Standard Business Terms” is the rather clunky but official translation of this statutory term, while General Terms and Conditions would be a more succint translation for international audiences. Both German terms not only span both the paper and the online world but—rather controversially—also cover wide ranges of business-to-business contracts.Footnote 49 The concept of a contract of adhesion may be wider than that of a modern wrap contract, with German terms ranging potentially wider still. Thus, to avoid getting bogged down with the devil in the details of whether principles perceived and restated for this situation also apply to the paperFootnote 50 or the business-to-business world, this Article will focus on online contracts which are offered to a large mass of consumers. These contracts are understood to contain the purest expression of the information problem, despite this issue having broader implications. Limiting the scope of examination does not imply that the information problem is necessarily limited to the online world or is even exclusive to the business-to-consumer-context.
B. The Legal Theorists’ Response
From the perspective of legal theory, the alternative to this inconsistent approach is simple: Not considering boilerplate as contracts at all. Like most simple ideas, this one is not new. One may recall the famous passage in Hennigsen v. Bloomfield Motors, uttered nearly sixty-five years ago:
The traditional contract is the result of free bargaining of parties who are brought together by the play of the market, and who meet each other on a footing of approximate economic equality. In such a society there is no danger that freedom of contract will be a threat to the social order as a whole.
But in present-day commercial life the standardized mass contract has appeared. It is used primarily by enterprises with strong bargaining power and position […] Such standardized contracts have been described as those in which one predominant party will dictate its law to an undetermined multiple rather than to an individual. They are said to resemble a law rather than a meeting of the minds.Footnote 51
In the U.S., the classic scholars have emphasized the disconnect between mass-contracts versus freedom of contract and the similarity between form contracts and legislation. Llewellyn noted that the form contract “[…] amounts to the exercise of unofficial government of some by others, via private law.”Footnote 52 Kessler stated that “[f]reedom of contract enables enterprisers to legislate by contract”.Footnote 53 Slawson notes that they are “privately made law”.Footnote 54 Modern interlocutors have taken up this cause. Ben-Shahar has stated “[o]n a theoretical level, boilerplate is shown to be a legal phenomenon different from contract. Is it a statute? Is it property? Is it a product?”Footnote 55 Kim has stated that wrap contracts create their own laws.Footnote 56 Others have questioned whether boilerplate contracts are even contracts at all.Footnote 57
The German law equivalent to the canonical passage of Bloomfield Motors originates from a speech of Großman-Doerth in 1933.Footnote 58 Standardized contracts are described as “self-created law of the economy” (Selbstgeschaffenes Recht der Wirtschaft). Following this idea, scholars influenced by a sociologist’s perspective of law advanced a theory of norms (Normentheorie) that sought to describe boilerplate as ready-made de-facto statutes to which the recipient party subjected themselves to.Footnote 59 In this theory, boilerplate contracts are understood to be factually similar to legal norms. We find substantial overlap between the American and German perspective.
It may appear somewhat suprising then, that the German norm theory of standard form agreements was almost universally rejected in the second half of the 20th century, while in the United States, such descriptions have remained common. The explanatory memorandum on the law governing mass-contract form agreements expressly sought to establish these agreements firmly in the realm of contract, and thus rejected norm theroryFootnote 60 Throughout the German statutory provisions, pre-drafted terms are understood as contracts, albeit subject to heightened court scrutiny and special provisions on their inclusion into an agreement and application.
Thus, the contractual nature of boilerplate is handed down by law. A partial reason for this is that the idea of a unilateral legal norm outside of contract law was considered to be stooped in the civil law theory of national-socialist civil law reform. It is understood that the national-socialist approach to private law sought to establish the creation of other sources of law besides contract, due to its rejection of liberal individualism. Therefore, to many post-war commentators, the idea of creating law through other means than party autonomy and therefore liberal individualism, was seen as ideologically suspect to say the least. That being said, no German scholar has ever denied the highly diluted connection between mass contracts and party autonomy, but commentators ritually are obliged to locate form agreements firmly in the contractual realm. In German legal theory, although mass contracts are definitely contracts and not norms, it is clear that they are not akin to real contracts.
Given that the norm theorists viewed the law through the lens of sociology,Footnote 61 they were not academic navel gazers; rather, their approach was similar to that of the legal realists.Footnote 62 Thus the norm theorists, as do modern American theorizers, did more than fabulate on the nature of boilerplate, but rather raised specific demands on how the law should deal with standard form contracts. Like their modern American counterparts, norm theorists sought not only to describe the law, but shape it. They argued that due to their nature, form contracts should not be interpreted like regular contracts, in other words, by searching for the understanding of the parties, but rather under the principles of statutory interpretation.Footnote 63 Norm theory was also utilized to allow for increased scrutiny of these norms as to whether they created a just balance of rights and obligations between the parties. Finally, a duty to apply these norms in a non-discriminatory way was suggested.Footnote 64 Under German national law, most of these demands have been implemented through legislation in the year 1974, thus almost twenty years earlier than the EU Unfair Contract Terms Directive in 1993, which further established these in European law.Footnote 65 It could be said, that non–discriminatory application follows from general anti-discrimination law created in 2001. Despite its almost universal rejection, the demands of norm theory have largely been vindicated.
C. The Legal Realists’ Response: The Information Model and the Control Model
Thus, the discussion about the nature of boilerplate is always a discussion about the instruments to be applied to their control. A legal realist would thus not focus on the legal or even metaphysical nature of boilerplate, but rather seek to analyze what remedies the law can offer for the information problem. These can be broadly categorized into two different approaches: the information model and the control model.Footnote 66
I. The Control Model
The control model is simple: It subjects the terms agreed upon in the fine print to heightened court scrutiny. In the United States, the method for doing so is the doctrine of unconscionability.Footnote 67 Following the seminal case of Williams v. Walker Thomas Furniture Co.,Footnote 68 unconscionability requires an “absence of meaningful choice” and clauses that are “unreasonably favorable” to the other party.Footnote 69 Unconscionability therefore has a procedural and a substantive element.Footnote 70 In Europe, pre-drafted contracts are subject to much greater court control than individual agreements. This follows from the EU Directive on Unfair Contract Terms. Not only does this ban unfair contract terms, in other words, terms that “contrary to the requirement of good faith, [cause] a significant imbalance in the parties’ rights and obligations arising under the contract, to the detriment of the consumer.”Footnote 71 Additionally, the Directive provides a blacklist of terms that cannot validly be agreed in the mass-contract setting. The national laws of the Member States implement these requirements, and a grey list for which unenforceability is presumed. The threshold for court intervention is understood to be much lower than in the United States.Footnote 72
II. The Information Model
The information model seeks to give consumers the necessary tools to make an informed decision. It can thus be described as the self-help model of consumer protection. Ideologically, these approaches seek to reconcile the idea of a free-market society based on liberal-individualism with the apparent market failures it produces. But instead of throwing out the market-economy-baby with the bathwater of command-and-control regulation, it is attempted to merely inform consumers—thus maintaining the benefits of a market approach and light-touch regulation. Although contract law focuses on creating a binding legal obligation, it imposes informational requirements on businesses in exactly the same way as an information mandate. In an ideal world, meaningful assent to a contract requires knowledge of the agreement. If it is not given, there needs to be some mental gymnastics to find a meaningful understanding of the agreement. Viewing contract law in this way allows a connection to the global legal discussion on the role of information mandates in regulating business practices. This debate borders on religious conviction and spans the true disciples sitting in lawmaking commissions, reformers, who seek to save the information gospel by reinterpretation in smart disclosureFootnote 73 or other nudges,Footnote 74 to heretical sceptics.Footnote 75 Although some have sought to frame this discussion as policy-neutral, it should be apparent that the information model has always contained a clear view on how economies should operate.Footnote 76 That being said, sometimes the information model is advocated not out of ideological but practical reasons, in other words, that it is the second best, but politically achievable form of control.
The main expression of the information model in mass-contracting is the issue of term adoption or “manifestation of assent”, as referred to in U.S. contract law. As stated in the Restatement of Contracts: “The formation of a contract requires a bargain in which there is a manifestation of mutual assent to the exchange and consideration.”Footnote 77 Term adoption sets informational requirements for a clause to be considered as agreed upon between the parties, in other words, part of the contractual agreement. In the United States, these issues have garnered growing attention under the many forms of wrap contracts.Footnote 78 Shrinkwrap poses the question whether there is sufficient information about the content of an agreement if the agreement is packaged inside the shrink-wrapped box of a product. Using this image, clickwrap poses the question of sufficient notice in the online marketplace if contract terms are accepted by clicking a checkbox.Footnote 79 Browsewrap poses the same question, yet simply for the issue of whether scrolling through a wall of text is sufficient notice.Footnote 80 Despite their conceptual origins as expansions of shrinkwrapping contracts, online wrap contracts do not actually wrap anything at all. Rather, the judicial and academic debates on term adoption are mainly concerned with the design of information provision on the internet. The information model can accurately be described as a checkbox model or a user interface model.
Recently, in U.S. discourse, the requirements for term adoption have been restated in § 2 of the Restatement of the Law of Consumer Contract.Footnote 81 § 2(a) of the Restatement of the Law of Consumer Contracts requires notice, opportunity to review, and assent.Footnote 82 § 2(b) of the Restatement of the Law of Consumer Contracts loosens the grip of these requirements.Footnote 83 This section allows for the adoption of terms available for review only after the contract is concluded, if there was reasonable notice to the existence of the standard contract term, a reasonable opportunity for review is given at a later time and the consumer may terminate the transaction without unreasonable negative effects.
In the European Union, term adoption is an issue of national law, as the European Directive on Unfair Contract Terms in Consumer Contracts provides few specifics regarding the informational design of mass-contracts. The requirements of § 2(a) of the Restatement of Consumer Contract Law will be familiar to German lawyers. In Germany, boilerplate term inclusion is governed by the BGB, which similarly requires a notice, an opportunity to review the terms, and assent to these terms.Footnote 84 Although much older, German law therefore mirrors this part of the Restatement, yet does not provide the additional leeway of § 2(b) of the Restatement of the Law of Consumer Contracts. Beyond term adoption, an additional information requirement arising from European law, is that the terms must be intelligible and contain no surprising clauses.Footnote 85
III. Information and Control in the U.S. and Europe
Given the universal nature of the information problem, it is not surprising that the legal frameworks in the U.S. and Europe are somewhat similar. Yet, there are two important differences. The first is a difference in emphasis. Although it is difficult to measure, there seems to be a much stronger focus of U.S. jurisprudence on issues of assent. Sibony has articulated the uncontroversial intuition that “US law seems to trust markets more and EU law seems to trust courts more.”Footnote 86 It would follow, that the Europeans would focus more on the control model, Americans more on the information model. This shows in the fact that U.S. jurisprudence provides an unerring mass of high-profile cases which deal with term adoption on the basis of the wrap-principles. Compared to this wealth of jurisprudence, Europeans are paupers. Contracts rarely fail the inclusion test, and it is treated as a formality. This is not to say, that the European approach has not affection for checkbox design. Yet in the context of form contracts, the requirements are seen as a hoop to jump through—rather than as safeguards of contractual equity.
This difference is also evident in the academic sphere. In the U.S., issues of term adoption are controversial. Recently, the shrinkwrap line of jurisprudence has come under fire, for undermining the requirement of assent provided outlined in Section 2-207 (2) UCC.Footnote 87 The Restatement of the Law of Consumer Contracts has also come under substantial criticism.Footnote 88 Much of which is centered around the concept of assent in Section 2 of the Restatement. In Germany there has not been significant activity to this effect. Section 305 BGB, which stipulates the requirements of term adoption in standard form contracts and thus would govern the equivalent wrap scenarios, has been the subject to almost no academic interest.Footnote 89 The provision of assent is not understood to be of any importance and assent is routinely construed to have been given tacitly if there was sufficient notice. Curiously, the requirements of notice, opportunity to review and—implied—assent have come under criticism as subverting contract law in the U.S.Footnote 90 In the German context these very requirements were viewed as establishing mass contracts as contracts.Footnote 91
Another major difference is the existence of a separate legal framework for issues of data collection in Europe, specifically the GDPR. In the U.S. the issue of consent to data collection and usage, in other words the assent to privacy notices and policies is routinely framed as contractual.Footnote 92 In Europe, it is viewed as consent to an infringement of a right to personally, akin to consent to medical treatment. Again, legal theory argues whether these consent declarations are of contractual nature or not.Footnote 93 But what matters is that they are subject to an entirely different and fully harmonized framework, in other words the EU GDPR. The boundaries between GDPR and contract law are often murky.Footnote 94 Thus, although the European Court of Justice loves to rule on the design of checkboxes, and app design, it usually does so outside of contract law.Footnote 95
IV. The Interplay between Information and Control
Before evaluating the pros and cons of these models, it is important to understand their interplay. Both the information model and the control model respond to a limitation of freedom of contract. Mass-contracts do not live up to the lofty ideal of well-informed contractual self-determination. The information model thus seeks to create this ideal, or at least a mirage of it. The control model by contrast uses the failure of freedom of contract to legitimize court control. Because the ideals of freedom of contract are not met, the courts are granted far greater leeway in content control.
A closer look at the dynamics shows that the control model and the information model are mutually exclusive. If the information model succeeds, it delegitimizes the control model. If a person were well-informed enough for their declaration to be viewed as a pure—or pure-ish—expression of freedom of contract, then a court deeming the reached agreement to be unfair or unconscionable would be nothing more than pure paternalism—or the less common gender neutral phrase parentalism.Footnote 96 Although such a “welfarist”Footnote 97 approach is not outside the academic or political spectrum, it requires different justification than a simple failure of freedom of contract. This connection seems intuitive to the American lawyer, because procedural unconscionability effects material unconscionability. For the European approach, this is less obvious. This is due to the fact that the procedural elements, in other words the procedural breakdown of contracting, are viewed as a separate requirement to term control itself. In German law, one must first determine that: (a) The term in question is actually an adopted standard business term before (b) subjecting it to heightened scrutiny. Thus, procedural elements of term adoption just open the gates to the kingdom of control, but do not seem to govern that kingdom at all. Despite this, in the European context it is equally universally acknowledged that term control follows from a failure of freedom of contract. If this failure were remedied, it would logically lack a basis. Put simply, contract terms are controlled, because they are not read. If they are read, then why control them?Footnote 98
The reverse relationship is less obvious. The efficacy of term control takes away the necessity of the information model. If terms are controlled because there is a failure of freedom of contract, then why make efforts to safeguard freedom of contract at all. If clauses are controlled, why must there be opportunity to read and notice? Put simply again, contract terms are controlled because they are not read. Why should we try to design them in a way, which allows them to be read? Therefore, we can establish that the more control exists as to the content of an agreement, the less control needs to exist with regard to information.
This has repercussions for the ideological underpinnings of contract. As mentioned above, the information problem in mass contracts can be described through the instrument of economic analysis, but is also understood by doctrinal approaches. Once we get to curing the disease, the approaches taken may matter. On the one hand, for the law and economics approach, the exclusivity of the information model and the control model is especially stark. Unless one argues with behavioral biases or some other rationality defect, term control cannot be rationalized economically, if the information model works.Footnote 99 On the other hand, an information model that does not work cannot promote efficiency. Thus, the law and economics approach has a hard time rationalizing that the law implements both models. This is due to the fact that law and economics is built on a consequentialist and therefore outcome-oriented approach. Conversely, approaches which view contract as a procedural tool for the expression of freedom may have less issue with rationalizing that the requirements for self-determination need to be given, even if they are not used. That being said, this point is still a somewhat pyrrhic victory for skeptics of law and economics. Term control is rationalized throughout doctrinal approaches with a breakdown of the process of bargaining. Theorizing that this potential bargaining is important in light of a universally accepted breakdown of bargaining seems strange. It calls those approaches into question.
V. Mass Contracts without the Information Model
We have seen that the information and the control model seek to combat the same problem, but are mutually exclusive on a theoretical level. Where the information model fights fire with fire, the control model fights fire with a ban on fire. Which model works better against the information problem is an empirical question. How informed are consumers about contract terms? The evidence suggests low readership rates, even if they are not zero.Footnote 100 That being said, there is a broad consensus that the information model is ineffective. We have also seen that the information model is tied up with the idea of contractual self-determination. Yet at the same time it is common in the literature to view standard form contracts as something less than contracts, more like laws, making the link between contractual self-determination and mass contracting seem less essential. Thus, the rational response would seem to be to abandon the information model altogether. We will explore how such a system could work and what advantages it could bring.
1. Shifting the Needle of Self-Determination
Neither German law nor United States law fully commits to either the information or the control model, but rather uses a combination of both information mandates and court control.Footnote 101 Instead of taking a strict “either information or control” approach, it is more apt to view information and control to be complementary, in other words, each being able to compensate the shortcomings of the other. Although this of course expresses the same thing but inside-out, it allows us to view contractual self-determination and non-consent-based statutory obligations as two extremes on a sliding scale. Unless we are at the extreme ends, we are always looking at a blend of contractual-self-determination and non-consent-based obligations. German judgements influenced by norm theory argued that a ready-made statutory order could not create obligations for another person ipso iure, but rather that the recipient had to make a declaration of submission, which was a declaration of will.Footnote 102 Thus, finding boilerplate to be akin to norms rather than contracts is not an extreme departure from the status quo, but a slight shift of the needle on the sliding scale. In a recent publication, Kim has proposed a taxonomy of adhesive agreements,Footnote 103 filling out the sliding scale. If the only tool available is a hammer, every problem must be considered a nail. If the only way for parties to create legal obligations is through contract, every obligation requires a contract. Given that there is universal consensus on the realities of online mass contracts, the question remains where they fit on the spectrum between individual and collective rule making. When we seek to abandon the legitimacy created through the information model, we shift the needle towards the more collective spectrum.
2. An Issue of Constitutional Rights?
That does not mean to say that the needle can be placed at will, rather the use of the information model is dependent on preferences, politics and ideology. In some areas of life and law, the information model is a quasi-ethical necessity. There are certain decisions that are so bound up with personhood and respect for basic rights, that there is very little choice but to have them be made by the individual, even if this individual might not reach, or could ever reach, the standard of rationality and informedness that such a decision should reasonably require, they must make the decision anyway as best they can. A textbook example of such decisions would be consent given to medical treatment, or a determination about post-mortem organ donation. It is assumed that given the necessary information, a person can make the correct determination about their own body. Even if there were evidence to the contrary, namely that people neither read nor understand even the simplest surgery disclosures, there seems little alternative but to using the framework of self-determination to justify putting scalpel to skin. Yet again, these scenarios show the inverse relationship between information and content-control. If a person has informedly objected to post-mortem organ donation, it would be very difficult to control whether this decision was fair, just, or conscionable. It would be unclear against what baseline the decision could even be measured. This shows that informed consent is a normative necessity in many contexts. However, it is argued here, that mass contracts, specifically dispute resolution clauses, are not one of them.Footnote 104
The argument could therefore be made that if human rights are involved, the information model always becomes a normative necessity. Such a strong inference is unfounded, as many contract law disputes can be framed as an issue of human rights. This includes the issue of consumer arbitration.Footnote 105 German law can give some, possibly cautionary, insight on the role the control model can play in human rights issues, as it has great experience with phrasing civil law disputes as clashes of human rights. Germany follows the Kelsenian System of having a separate constitutional court which may overturn decisions by all other courts if, and only if, they violate constitutional rights. Thus, many higher level court decisions are written with a view to constitutional meaning human rights review. A prime example of this is the 2021 decision by the Federal Court of Justice regarding the removal of a user post on Facebook.Footnote 106 In evaluating whether the removal was justified due to a violation of Facebooks anti-hate-speech policy, the court had to decide if and how large social media platforms may regulate speech through user policies, which are considered standard form contracts. It is clear that this is a matter with significant effect on the exercise of free speech, a fundamental constitutional right, which the court expanded on at length. It applied the control model, with a view to constitutional rights. Yet despite the huge impact of the decision on how an algorithmic-democratic society functions, the information model demanded that the court had to lead Facebook through the hoops of term adoption and analyze the design of the checkbox, that the hate-speech spouting user had clicked when creating his account. After all, they could not be bound by speech regulating policies, without sufficient assent. In a case like this, focusing on the form of contracting, rather than the content, verges on the bizarre. The constitutional weight of the case bulldozes through any impact of whether the policies in question were read, pre-drafted, clicked eccetra. What if the user had not actually clicked the checkbox, or the browsewrap was defective in some technical aspect? Would this make us more or less comfortable of with allowing hate speech on Facebook? While in some contexts, human rights issues may require the use of an information model, in other contexts the use of the information model borders on the absurd.
D. World without Boilerplate
Against this backdrop, in a world without the information model, we assume that a business may unilaterally set terms in an online mass contract setting. The requirements for this are that there is some basic form of contractual agreement, thus providing an anchor for the creation of legal obligations. This core bargain is what the Restatement of Consumer Contract has called “the dickered terms” in reference to Llewelyn.Footnote 107 Yet if an online contract is concluded, abandoning the information model assumes that this contract is governed by “invisible”Footnote 108 terms. While Rakoff argued that the invisible terms should not be valid without affirmative justification,Footnote 109 and recently Hoffman has advocated a battle against the empire of forms by enforcing only contracts subject to market discipline,Footnote 110 my goal here is the opposite; in other words, to make these invisible terms as invisible as possible. The invisible terms govern the transaction without notice or assent, as a pure expression of unilateral rulemaking. They are invisible, yet adhesive to the dickered terms.
If the contract terms do not need to be clicked, checked, or browse-wrapped, how can interested parties learn the scope of their rights and obligations? And more importantly, beyond the interest of contract aficionados, how can a court determine what was actually agreed upon? The answer for the courts is the same as under current law: Interpretation of party conduct and the burden of proof. If a party seeks to rely on a standardized wording, it must prove that it was its common and indiscriminate practice to use the term sheet in question for contracts of this type. As not to deprive consumers of information, this contract must be easily found on the website of the business. This obligation would go beyond the current practice of having a section for “terms and conditions.” Rather the different versions of contracts would need to be accessible allowing consumers to find “their contract.” However, the obligation would exist in the abstract, not in the context of each and every transaction. For large companies with a plethora of fast-changing services, this would create an online archive similar to the archives commonly found for user manuals. This obligation will actually make it easier for consumers to become aware of their rights. Think of this: If you want to re-program your robot vacuum, is it quicker to look for the manual in a dusty box in the attic, or to simply find the PDF document through a search engine? The same is true for contracts.
It shall not be overlooked that there may be a generational element at play here. Note however that the dusty box in the attic is somewhat metaphorical, as in the online space the comparison would be finding terms on a website or finding them in some documentation received in an email in the browsewrap process. In this context, finding the terms through a search engine is clearly easier, indifferent of age.
I. Advantages of a World without Boilerplate
1. Breaking the Information Bottleneck
The information model creates an information bottleneck at the time of contracting. Yet for any information mandate to be effective, or daresay smart, it needs to provide the information at the most relevant and opportune point in time. The information model cannot accommodate this, because it ipso iure needs to inform the recipient before contract formation. It is obvious that this is not the best point in time to inform consumers about their rights and obligations. At this point in time, it is very unlikely that any of the provisions will ever become relevant. Barnhizer has correctly pointed out that, “the time of contracting is not the best moment for promoting the ability of adherents to affect the terms of their transactions or to protect themselves against predatory or pathological contract terms.”Footnote 111 This is different if a conflict has already materialized.
The information model seeks to facilitate an informed decision on contracting, rather than mandating intelligible disclosure on the rights and obligations of parties at the moment those rights and obligations become relevant. It is argued, however, that the second goal might actually be achievable through information mandates. If and when a dispute has arisen, businesses can and should be hit with the full arsenal of smart disclosure. Focusing on this later point in time will be infinitely more fruitful than providing information when most people are simply not interested. Of course, this ex post contract information does not prevent harmful terms. If a person is told after the fact that they have sold their soul to the devil, it will be of little solace that the notice is easy to understand. Therefore, it has to be re-emphasized that abandoning the information model is suggested in conjunction with strong term control.
Beyond the bad timing of the information, the design, language, and content of contracts are equally unfit for the purpose of informing consumers about their rights. Ex post contractum information grants more license in these aspects. It does not need to communicate all contract provisions or structure them into legalese contract clauses. Rather, certain scenarios can be determined, and the relevant rights —whether they are arising from the agreement itself, precedent, or statute— can be explained. By contrast, writing a contract that explains the most common scenarios bit by bit would create unmanageable contract length. Ex post contractum this is not an issue. The user can select the scenario relevant for them, and ignore the rest. Note also that the contractual information on liability is often incomplete, as a contract need only contain the information that deviates from the legal framework. You cannot understand what a waiver of liability means if you don’t know the rules for liability in the first place.Footnote 112 Finally, the contractual text might not actually represent the real deal at all, but rather a paper deal which is routinely changed in ex post bargaining.Footnote 113 This can be disclosed ex post contract.
Abandoning the information model seeks to fundamentally change the relationship between the contractual document and information mandates. Dropping the information model means dropping contractual documents as sources of consumer information. Think of the contract as a source code of a program, the resulting information as the program’s interface. The consumer does not need to understand the code, if the program is easy to use. Dropping the information model would allow for such approaches. Contracts themselves could remain legalese, as long as their corresponding information materials are not. Under European Law, Article 5 Unfair Contract Terms Directive requires terms to be transparent—in other words, that consumers are able to foresee the economic consequences of a given term.Footnote 114 Transparency is supposed to allow consumers to understand their contractual rights, which requires transparency to exist before contract formation and during performance.Footnote 115 Beyond the information model, it would be sufficient if the contractual information presented at the point of an issue is understandable to the consumer. If the explanations provided to consumers are transparent, the contract itself need not be.
2. Courts Should Focus on Material, not Formal Decisions
Reducing the importance of the information model focuses judicial decision making on the material rather than the formal aspects of the case. If courts have fewer procedural tools, for example, questions of term adoption, at hand, they will need to rely on material control more. Take the issue of consumer arbitration. A real benefit for consumers comes to exist, if the law focuses on how and to what extent businesses may refer their customers to arbitration, and which procedural safeguards need to exist for consumer arbitration being a de facto default in mass contracts. This is done by taking attention away from the examination of whether the business in question has jumped through the correct hoops of contract inclusion.
The interplay between formal and material methods of term control can be seen in the jurisprudence on GDPR. As shown above, in German contract law, term adoption is not a major issue, but emphasis is given to term control. The opposite is true for consent under GDPR. The reason for this is that content control of privacy policies and consent declarations is not well developed and commonly misunderstood under GDPR.Footnote 116 The GDPR provisions do not mention term control at all. The Unfair Contract Terms Directive with its term control is only referenced in Recital 42, not in the actual statue. The recitals are merely guidelines to inform teleological application of European Law.Footnote 117 Courts are thus left with a vague feeling that term control is possible under the Unfair Contract Terms Directive, but its application and legal source are unclear. Rather than entering these murky waters, they focus on the clear formal hoops for consent from the GDPR.
The ECJ’s focus on formal aspects can be seen in the decision regarding Meta’s data collection practices.Footnote 118 The Court had to determine whether Facebook’s much-criticized tracking of users over the entire internet through tracking cookies and other instruments, and Meta’s practice of combining Facebook user data with accounts of other services like Instagram, and WhatsApp was compliant with GDPR.Footnote 119 The privacy policies, on which GDPR consent is based, are of course read no more than form contracts. Therefore, term control under the Unfair Contract Terms Directive would have allowed the Court to focus on whether practices of tracking in exchange for a free service are, all in all, a fair deal.
This did not happen. The Court ruled that the consent provided in the privacy policies could only be valid if Facebook offered an alternative service not based on data collection—in other words, for a fee.Footnote 120 In lieu of such an alternative, consent was not valid. The court is thus again preoccupied with maintaining the requirements of a meaningful decision, rather than reflecting the breakdown of the information model in this context. Recall that under Article 3 EU Unfair Contract Terms Directive a clause is “unfair if, contrary to the requirement of good faith, it causes a significant imbalance in the parties’ rights and obligations arising under the contract, to the detriment of the consumer.”Footnote 121 Focusing on the choice a consumer has in regard to clauses that they most likely did not read does not inform this standard. Therefore, the core issue is ignored, while details of choice architecture and user interface design are emphasized. As to the fairness of the core bargain struck in any use of online social media—personal data for social media services—we are still none the wiser.
3. Court Mandated User Interface Design is Bad Design
The design of digital user interfaces should be left to designers, not the courts. Designing digital user interfaces and experiences with a view to legal requirements of contract law, makes these interfaces and experiences significantly worse. If a camel is a horse designed by a committee, it should not be surprising that app UI designed by contract law comes with a few extra bumps. European readers will be familiar with the now ubiquitous cookie banners, while the contractification of the online world means that a large number of web services now require some form of sign in. Without evaluating the efficacy of cookie banners to improve privacy, they cannot be said to have improved the user experience of the internet. Undoubtedly, these banners have some effect on the expressed preferences for data collection, otherwise companies would not employ methods to affect consumer choice.Footnote 122 Yet they also annoy people by interrupting natural browsing.Footnote 123 Additional checkboxes make sign-ups more cumbersome and browsewraps take unnecessary time to scroll through. Individually, these detriments are miniscule, but one must be wary of a death by a thousand cuts.
4. Legal Certainty and Flexibility
Leaving the information model behind creates legal certainty or more precisely, it reduces the cost to attain it.Footnote 124 The information model puts organizational and financial costs on businesses, which need to attain legal sign off for every form of term inclusion. Think of all the unnecessary boilerplate printouts, signs, and checkboxes which populate our lives. This is even more pronounced, if the law of term inclusion is under constant flux as it is in the United States, which can be especially poignant if one focuses more on information than control. Under U.S. jurisprudence, the issues of term inclusion require significant thought. Of course, sophisticated players like Uber, Meta or Samsung are able to navigate these obstacles. Yet this sophistication incurs unnecessary costs.
Beyond this, the information bottleneck created by the fiction of consent is a barrier for new and innovative methods of contracting, all of which must go through the process of courts establishing the requirements for the new kind of wrap. The information model is especially restrictive in contract formation that is not text based. Consider the Amazon Dash Button, a now discontinued little button which consumers could put in their cupboard next to the laundry detergent, which would order additional detergent and other everyday consumables on Amazon at the push of the button. It seems clear that consumers–having purchased and installed this button–would understand its workings and the following contractual obligations. Yet the Amazon Dash Button was ruled by a German court to violate information mandates on consumer contracting.Footnote 125 This shows that the information model might curtail innovation for non-text-based contracting, which seems especially detrimental to voice assistants and other forms of Artificial Intelligence.
5. The Frame of Contracting is Harmful
Abandoning the information model sets the correct mental frame for mass contracts, in other words, that they are unilaterally imposed private rulesets; something less than contract. By making the terms invisible, or visible only to interested parties, it becomes clear that these are unilaterally imposed, and lack meaningful assent. This framing is not only closer to the truth, but is likely to influence how consumers react to the imposed clauses.
Under current law, form contracts look just the same as regular contracts, especially to lay people. Although browsing through or clicking them is a hollow exercise, the form of contracting is maintained. This framing as a contract carries real weight. Contracts maintain the “image of a ‘binding commitment‘ that symbolizes efficiency, effectiveness, free choice and legal protection for both sides.”Footnote 126 Yet as we have seen above, this could not be further from reality. Assent to form contracts is “fiction piled upon fiction”.Footnote 127 Wrap contracts and their text-wall cousins thus put the veneer of non-form-adhesive contracts onto something that is de-facto unilateral private law making,Footnote 128 and, in doing so, legitimize potentially nefarious, business practices.Footnote 129
This has a psychological effect on consumers. Research has found a disconnect between the moral weight of contracts as a serious obligation, and the everyday submission to unread terms.Footnote 130 Although people understand it is unrealistic to read terms, they cannot shake the framing of contract when attributing responsibility for transaction harm.Footnote 131 They do not make the leap that a contract is not that serious, if there is no real opportunity to read it. It is suggested that this has to do with a correlation of the morality of promise keeping and the morality of contract.Footnote 132 Promise keeping is a moral value that is important to people,Footnote 133 thus breaking a promise is immoral.Footnote 134 The hollow assent, even if it could not realistically have been given due to reading constraints, thus affects how the resulting obligations are understood. To put it succinctly, “assent seems, for the average consumer, to cleanse the transaction – to press the reset button, morally as well as legally.”Footnote 135 The proposed reason for this is the desire to feel in control of a situation that is not within the persons control.Footnote 136 More recently, it has been proposed that bloodless assent to unilaterally imposed terms serves the psychological function, of making peace with the status quo.Footnote 137 The result is contract double think, where we are forced on a daily basis to promise things which are entirely unknown to us, and then pretend that these are real promises only to not feel at the mercy of corporations. It becomes understandable, that boilerplate and its empty consent contribute to the degradation of the legal system.Footnote 138
The information model is central to perpetuating these fictions, as it safeguards the, albeit theoretical, possibility of meaningful term understanding and thus meaningful assent. In doing so, it shores up the cultural norms surrounding contractual exchange onto something that is fundamentally not a meaningful contractual agreement.Footnote 139 Therefore, the form of mass agreements as unread but assented-to contracts matters because it phrases the blanket assent to unread terms in the realm of a promise. A study by Wilkinson-Ryan compared consumer reception of an onerous and unread clause when presented in a contractual setting or framed as an online policy.Footnote 140 It found that subjects “believed policies embedded in a contract were more likely to be legally enforceable, judged those policies as more fair, and imagined that they would be less likely to challenge those policies in court.”Footnote 141 Similarly, it is argued that the form matters because of consumers’ commonsense intuitions about the law, in other words, that contracts will be enforced.Footnote 142 Even compromised consent therefore psychologically shifts blame to the consumer.Footnote 143 Against this background, the assumption that unread disclosures are at best harmless has been challenged.Footnote 144
Formalities may create the feeling of obligation, which firms may then leverage to minimize consumer conflict.Footnote 145 Increased participation in the contract formation stage has been argued to result in greater contract compliance.Footnote 146 However, abandoning the information model seeks to achieve the opposite. Instead of adapting the reality of consumer choice to the law, the law should adapt to consumer choice. Given that there is no meaningful assent, the formal language of assent should not be utilized. Because factually hidden and unknown terms govern the mass-contract, they should also appear as factually and hidden terms.
If we abandon the information model, mass consumer contracts begin to look a lot more like the unilateral company policies that they are. This change should make consumers more critical of their content, and more aware that these terms were in fact not born of their consent, but imposed on them. It is argued here that the resulting non-compliance is a good thing. Although it is commonly understood that contract compliance promotes market efficiency,Footnote 147 and thus promoting non-compliance seems questionable, there are two reasons to seek non-compliance. The first is that ex post control is reliant on non-compliance.Footnote 148 Compliance with a contract term much reduces the likelihood of a legal dispute ensuing, and thus acts to curtail ex post control. Therefore, non-compliance is a necessity for the control model, and increasing it is beneficial to the reduction of unfair terms. That being said, this connection is highly dependent on the framework in place for the representation and aggregation of consumer interests—in other words, class actions, private enforcement, public action, eccetra. However, it can be hypothesized that there is actually too much compliance with boilerplate terms, as the mere existence of form contracts deters people from taking action. Steering against this by abandoning the information model is a step in the right direction. Beyond this, an efficiency argument can be made. Given the realities of form contracting, there are reasons to believe that boilerplate terms are inefficient terms.Footnote 149 In effect, this is an application of Akerlof’s “Lemons” framework.Footnote 150 The compliance with inefficient terms is however not an increase to market efficiency.
6. Approaches to Regulating Consumer Arbitration
These remarks have been of general nature; however, they bear special significance to consumer arbitration. The reason is that in the U.S., consumer or employment arbitration clauses have been the bone of contention of most disputes involving online consumer contracts,Footnote 151 and thus have also given rise to new proposals on how to deal with them.
The arguments on consumer arbitration, whether it creates shadow law unknowable to the public,Footnote 152 or even alternative legal universes,Footnote 153 or rather is a fully legitimate tool that is less expensive and time-consuming for firms than going to court,Footnote 154 shall not be explored here. As this Article is on the methods of contracting, it will solely assert that concerns about consumer arbitration are best-availed through term control, not information mandates.
Naturally, U.S. consumer advocates have attempted to curb consumer arbitration through legislation, with the most recent attempt being the proposed Forced Arbitration Injustice Repeal (FAIR) Act.Footnote 155 Recently, a proposal of a group of almost 170 law professors, signed an open letter urging the U.S. Consumer Financial Protection Bureau to require meaningful consent to consumer arbitration clauses for financial products and services.Footnote 156 Meaningful consent is envisioned to be given when it is declared after the dispute arises.Footnote 157 Under German law, a similar provision exists, albeit only for consumer forum selection clauses.Footnote 158
This approach is rooted both in the control model and the information model. In some way, it is informational, as it seeks to break the information bottleneck. At the time ex post contractum the incentives to engage with contract terms are much higher. The information model is amended to apply to a later point in time. That being said, requiring an ex post contract decision is nothing short of a ban on pre-dispute clauses. The German Code of Civil Procedure bans pre-dispute forum selection clauses. The control model is in full effect here.Footnote 159
For consumer arbitration clauses, German law has provided a reduced version of the information model, namely manufacturing specific assent. The terminology is based on Llewelyn who noted that boilerplate is covered only by blanket assent, not specific assent.Footnote 160 Specific assent advocates agree that we can never get people to care about everything that is in boilerplate enough to understand it, but propose that individual terms can be highlighted in such a way that consumers will understand them.Footnote 161 Thus, blanket assent is turned into specific assent and the information problem disappears. Even if specific assent does not make consumers read or consciously note the terms, specific assent seeks to provide a transactional hurdle that signals the burdensome nature of the clause in question.Footnote 162 It is this very system that German law uses to regulate consumer arbitration clauses. The German Code of Civil Procedure requires that a party signs a consumer arbitration clause in a document separate from the main contractual agreement.Footnote 163 This separate document may not contain any other provisions than the arbitration clause, which makes it factually impossible to include such a clause in a standard form contract.Footnote 164 This is seen as the reason why consumer arbitration is not common in Germany.Footnote 165 German law, quite curiously, despite being deeply steeped in the control of terms through blacklists, whitelists and general clauses, has opted for two different approaches to dispute resolution and forum selection clauses. Section 1031 German Code of Civil Procedure is, at least to my knowledge, the only provision on specific assent in German contract law. It provides a “proven and high” standard of consumer protection.Footnote 166 The information and assent requirements are so high, the information model so effective, that they have the same effect as simply putting the term on a blacklist. Taking specific assent this seriously amounts to an effective ban under the guise of the information model.
Of course, requiring the signature of a separate document is more extreme than the forms of specific assent advocated by Kim or Ayres & Schwartz.Footnote 167 Undoubtedly, the methods suggested would leave room for more actual agreements than the German law on consumer arbitration clauses. Yet these attempts show that modifications to the information model always require a delicate balance. The Goldilocks principle applies: Make the requirements too harsh, and the effect is the same as a ban. Make them too soft, and there is no effect at all. Finding just the right balance, where there is some information effect, yet the obstacles imposed are not so onerous as to prevent all transactions is a very difficult enterprise.
Even if specific assent is set “just right,” some general criticisms persist. As Barnhizer has pointed out, this approaches “focuses on a symptom of the problem–lack of assent to the terms of a wrap contract–and consequent attempts to cure that symptom.”Footnote 168 If consumers do not understand arbitration clauses,Footnote 169 neither specific assent nor ex post assent saves the information model. Although it can be argued that ignoring boilerplate is fully rational, this does not make the person who does so a perfectly rational homo oeconomicus. Beyond this, the common criticism of specific assent is that no decision is made in isolation.Footnote 170 If every contract term, or further every element of every transaction, is raised to specific assent, then our lives would be filled with only one activity: assenting to things. There is an inherent limit to how many terms can be made “salient” to the consumer.Footnote 171 Clearly, specific assent has to be limited to the most pressing issues.
Specific assent promotes information and in doing so influences control. Put simply, specific assent undermines ex post control of terms.Footnote 172 It is theorized that courts will be more likely to enforce terms if they have been agreed upon through specific assent.Footnote 173 Making terms more prominent comes with the implication that consumers know, or should know, of their existence.Footnote 174 Psychologically, there is research indicating that people feel more bound to obligations if they are connected to formalities.Footnote 175 Of course, this is not a bad thing, if specific assent works, in other words, if consumers understand their obligations and factor them into their choice. Whether this is the case is an empirical question, with the aforementioned research giving reasons to be doubtful. Assent increases the pressure on consumer understanding and puts a greater emphasis on the expressions of consumer choice. This might be harmful, because specific assent can be obtrusive and is subject to manipulation. Think again of the most widespread use of specific assent, the cookie banner. Even these obtrusive assent prompts are significantly influenced by interface design, i.e. dark patterns.Footnote 176
Dipping our toes into the waters of dispute resolution clauses has shown that some expressions of the information model, in other words, ex post assent or specific assent, may be effective. They do, however, come with drawbacks where consumer understanding is limited. They may compound if specific assent is used throughout the legal system. As expressions of the information model, they seek to maintain more of freedom of contract versus content control, or in economic terms, allow for efficient transactions below the threshold of court control. It is doubtful, that such a fine-tuned approach, which gets the informational load “just right” is necessary for the issue of consumer dispute resolution clauses.
E. Conclusion
Psychologically, it might be more comfortable to live in an ideal world in which we can decide every one of our legal obligations with meaningful assent, as this would put us in full control. However, control is an illusion, at least with regard to consumer contract terms. Legal scholars agree that we live in a cold, harsh world, which requires the assent to unread agreements in order to function. This Article suggests that we should face this reality, rather than covering it up with the forms of contracting. The main reason for this, is that the form of contracts legitimizes these practices, and in the long run may de-legitimize actual contracts.
Much of what makes contract is wrapped up in ideas safeguarded–in theory but not in practice–by the information model. A cynic could therefore rephrase the title of this Article as A Farewell to Contracts in Mass Contracts. This would be short-sighted, as this Article attempts to protect the core of contracting; the bargain, the dickered terms, while stripping the veneer of contract from unread walls of text. This is an attempt to provide a boon to contract as a social institution, to combat its further erosion, rather than an attack on it.
Maybe there is another ideal world. A world in which puny and irrelevant technicalities, cumbersome contracts, and annoying banners– which almost never pertain to anyone– do not pop up in our faces. A world in which we can continue without engaging with the nitty-gritty of wrap contracts just to order some take-out. In which these rules do not have the teeth to bite us. Meaning that we are free of our labors of deciding whether we prefer consumer arbitration for a social media app we delete minutes later, and joined back to safely ignoring them, all watched over by term control of loving grace. Maybe this ideal world is also unattainable, but steps towards it seem possible.
Acknowledgement
The author thanks Gralf-Peter Calliess for the invitation to the ICtDRA Conference on Informed Consent to Dispute Resolution Agreements, Nicholas Mouttotos for organizing the draft submissions, Daniel Barnhizer, Stephen Ware, Gerhard Wagner, Sebatian Eller and Gesine Müller for helpful comments. Additional thanks to Oren Bar-Gill and Omri Ben-Shahar for discussions that shaped more basic variants of these ideas many years ago.
Funding Statement
No specific funding has been associated with this Article.
Competing Interests
The author declares none.
Author Biographical Information
Peter McColgan, Post-Doctoral Research Fellow at the Chair for Private Law, Commercial Law, and Law and Economics of Prof. Dr. Gerhard Wagner, LL.M. (University of Chicago), Humboldt-Universität zu Berlin; part of this Article is based on the research published in German in Peter McColgan, Abschied Vom Informationsmodell Im Recht Allgemeiner Geschäftsbedingungen (2020).
Open access
