Behavioural Economics in Unfair Contract Terms

Abstract

The domain of behavioural law and economics is winning increasing attention also in the field of consumer policy. How the insights of behavioural law and economics can be used in policy remains, to a large extent, unclear. In this article, the following question is asked: “To what extent can the insights from the behavioural literature be applied in a way to formulate concrete suggestions to policy makers?” The authors show that many of the findings of the behavioural literature are very context-specific and hence apply only with respect to particular products or services and particular consumer groups. Formulating general policy conclusions is therefore difficult. However, as far as the specific domain of standard form contracts is concerned, the authors argue that the behavioural literature has shown that the traditional remedy (mostly resulting from information economics), being to focus on information disclosure will not be able to remedy market failures resulting from failing information and the "signing-without-reading-problem." Hence, more substantive forms of intervention in standard form contracts (e.g., resulting from collective bargaining) may be indicated as a remedy.

Introduction

Traditional economic analysis of law, including economic analysis of consumer law has been largely based on the so-called rational choice theory. This theory assumes that when individuals are confronted with various choices they will choose the option that yields them the most expected welfare.¹ Within this model individuals are supposed to maximize their own preferences and there is little reason for an intervention by the legal system, unless when so-called market failures would appear. Market failures that may justify an intervention in the area of consumer policy are especially transaction costs and information deficiencies. The traditional remedies for these market failures would be to standardize information provision to consumers (e.g., by providing standard terms in contract) in order to reduce transaction costs or to improve information, e.g., via mandatory disclosure regulation and, more generally, regulations concerning the way in which information needs to be provided to consumers.

The assumptions in rational choice theory, specifically relating to Homo economicus, the rational calculator who is aware of all his preferences, knows all ins and outs of the options with which he is presented and is perfectly able to choose the option that maximizes his own welfare, have been much disputed.² Lawyers have criticized the effectiveness of the traditional remedies in consumer protection that find their basis in economic analysis.³ Individuals, and more particularly consumers, do not always act as if they maximize their own preferences. Whether the assumption that consumers are rational and perfectly able to act in their own self-interest results in verifiable predictions of consumer behaviour is debated. Even when appropriate information is provided, consumers often sign without reading (see De Geest 2002). Disclosure duties apparently do not guarantee that consumers make rational decisions of what would be in their own interests (Wilhelmsson 2006). The consumer to which the information is provided is often still not able to process the information adequately and take a meaningful decision on the basis of it (Grundmann 2002).

Cognitive psychology can be found to support this critique. Starting from the seminal work of Kahneman and Tversky, cognitive psychologists have shown that many individuals suffer from a variety of cognitive heuristics and biases as a result of which decision-making may take place in a different way than neo-classic economic models assume (see Tversky and Kahneman 1974). This literature, which is supported by powerful empirical evidence, has shown the limits of individuals to make rational choices. It has lead to the emergence of a new domain, which is now referred to as “behavioural law and economics.”⁴ Even though this behavioural approach has encountered opposition from some Chicago style law and economics scholars,⁵ others have recognized the valuable contribution of cognitive psychology for law and economics.⁶ While the behavioural approach has now entered traditional economic analysis its use is still far more debated when the behavioural literature is used at the normative level to formulate policy recommendations,⁷ even when policy makers’ interest to do so is clearly present.⁸ Many argue that the behavioural approach is still that new that more empirical research is needed for policy makers to be able to make effective use of the insights provided by behavioural literature.⁹

There is, however, probably one domain where the number of empirical studies that show evidence of heuristics and biases is impressively large and this is the area of consumer policy. That does not mean that the use of these insights at the normative level would be less debated,¹⁰ but at least the empirical evidence is that large that this merits the question if and how these insights could be translated into policy recommendations. That question is precisely the goal of our paper.

We aim at discussing the findings of behavioural law and economics as far as consumer policy is concerned¹¹ and will especially examine the potential of this research to formulate policy recommendations. The effectiveness and efficiency of traditional remedies in consumer policy that can be argued for using economic theory can (and has been) questioned. Critique, however, is often based on intuition, not on empirical or experimental studies. In this respect, we would like to point at the potential of the behavioural research, but also at its limits. Given the limited scope of our paper, we will formulate a few general observations concerning consumer policy, but focus on the example of misleading standard terms in consumer contracts. The behavioural literature is that rich that one could easily fill an entire book with this topic.¹² Given the limited space, we will necessarily have to provide a summary of the literature; the reader is referred for a more in-depth analysis and further reading to the references mentioned in footnotes.¹³

The remainder of this paper is structured as follows: First, we briefly sketch the traditional economic approach towards consumer law and standard terms in particular based on transactions costs and information economics (“Traditional Economic Approach”). Next we sketch some of the main lessons of behavioural economics for consumer protection and look at the potential policy implications of this literature (“Lessons from Behavioural Law and Economics”). We then apply these behavioural insights to the problem of standard terms, addressing how the behavioural lessons are different from the remedies proposed by information economics (“The Case of Standard Terms in Consumer Contracts”). Then we summarize EU policy with respect to standard terms in consumer contracts and formulate a behavioural critique (“European Consumer Policy”). The last section concludes (“Concluding Remarks”).

Traditional Economic Approach

Correcting Market Failures

The starting point for any economic analysis of law is usually the well-known Coase theorem, which holds that in the absence of transaction costs an optimal allocation of resources will arise, irrespective of the contents of the legal rule (Coase 1960). This theorem certainly has its importance for consumer policy, as the starting point is that well-informed parties will in principle negotiate for an optimal quality and output of services and products. Of course, the theorem relies on the heavy assumption of zero-transaction costs; the presence of transaction costs may hence be an important reason for intervention through the legal system. Generally economic analysis holds that such intervention is necessary only to correct market failures. The most important market failures in the area of consumer policy are information asymmetries and transaction costs.¹⁴ Transaction costs relate to the costs of transaction initiation and bargaining, such as search costs, processing and storing information, conclusion of the contract, and costs related to the enforcement of the contract. A problem for consumers is that the additional costs of gathering more information to improve their decision-making can be higher than the additional benefits (Hadfield et al. 1998). Consumers may therefore suffer from what is referred to as rational apathy.¹⁵ The presence of transaction costs has been advanced as an argument in favour of government interventions, aiming at the reduction of transaction costs. An example would be to apply a fairness test to contract terms. As a result consumers would not have to worry about overly harsh terms being included in their contract, as these terms cannot be legally invoked against consumers. However, interventions based on a reduction of transaction costs are not always unproblematic. For example, an automatic renewal of contracts could be prescribed as this would decrease transaction costs for consumers that had wished to stay with the current supplier anyway. However, this measure at the same time increases the costs of switching to another supplier as well as the costs of ending the current contract (Sovern 2006, p. 1702).

Information asymmetries are also often advanced as a reason to intervene in consumer markets. Especially in the case of so-called experience goods or credence goods the quality of the product or service can only be evaluated after the purchase (experience goods) or very difficultly to not at all (credence goods) (Nelson 1970). Given the disparity between the seller and the consumer in consumer contracts, the existence of the market failure of asymmetric information is a realistic problem. If consumers cannot easily assess the quality of a product or service they will often base their purchase decision solely on price, as a result of which high quality products may be driven out of the market.¹⁶ This information asymmetry and the resulting danger of adverse selection may be reasons for government intervention, for instance aiming at the introduction of disclosure duties or mandatory quality standards (Schwartz and Wilde 1979). Economic literature has prescribed information duties as a remedy to address information asymmetry (Hadfield et al. 1998). Thus legislation could improve consumer welfare by prohibiting fraud, the distribution of false information and the imposition of duties to inform (Van den Bergh 2003).

Although economics hence recognizes the need for regulation to cure information asymmetries and reduce transaction costs, economics also recognizes that in some cases market corrections may be possible, e.g., through signalling and screening, long-term relationships and learning (Epstein 2006). Moreover, the market itself could provide for the distribution of the necessary information. Market solutions would usually be preferred over government interventions. However, in many situations these market-based solutions are unlikely to emerge (Hadfield et al. 1998, pp. 155–158). This may more particularly be the case when repeated transactions are rare and consumers are unable to learn from their mistakes. The government could be more efficient at providing information than the market, in which case the market solution is not the most efficient one. Moreover, in some cases the costs of learning may be very high in the sense that consumers could be severely hurt by bad decisions. Relying upon the market solution of consumers learning from their mistakes could then lead to high costs at the expense of consumers, which cannot be compensated through lessons learnt.

Lessons from Information Economics

Information economics teaches first of all that the value and costs of possible information asymmetry problems should be assessed. Next, the question should be posed whether a government intervention is at all justified. In this respect, information economics warns that government interventions in consumer contracts to correct market failures come at a cost. As sellers will be confronted with costs due, e.g., to mandatory information duties, these costs will be passed on to consumers where possible (Hartlief 2004, pp. 258–260). Price increases due to consumer protection may even lead to a negative redistribution since products or services would no longer be obtainable in the market for lower income groups (Van den Bergh 1997, pp. 94–95). Moreover, consumer protection may lead to consumer moral hazard, reducing consumers’ incentives to take care (Van den Bergh 1990).

The third step is that an appropriate regulatory instrument needs to be chosen to counteract the information asymmetry.¹⁷ Information economics would, in sum, advise to put information duties on the party that is the cheapest information provider.

These insights of information economics can be applied to the specific problem of standard terms in consumer contracts. According to information economics the main cause of inefficient terms follows from the signing-without-reading problem (De Geest 2002, p. 222). One remedy is to provide default rules. This reduces transaction costs since parties no longer have to draft contract terms themselves and consumer search costs can be reduced by making default terms easily available (Luth 2010, p. 141). The most important problem from the perspective of information economics, however, is the signing-without-reading problem and the resulting adverse selection. Since buyers are rationally ignorant (rational apathy),¹⁸ sellers can impose their conditions; adverse selection follows from information asymmetry (De Geest 2002). Adverse selection will lead to a situation in which consumers only (or at least mainly) observe price as a result of which the quality of terms in consumer contracts is driven down.¹⁹ Since these one-sided terms do not necessarily correspond with consumer preferences (who would, if they were better informed have preferred more favourable terms), the result is a market failure and a welfare loss (Korobkin 2003). Information asymmetry hence constitutes the most important justification for a regulatory intervention, according to information economics.

Possible Regulatory Interventions

As we already mentioned, a first type of regulatory intervention would be the provision of efficient default rules. This would then be the rule that would be preferred by the majority of contracting parties.²⁰ Given high transaction costs it is unlikely that consumers would negotiate for better terms. Hence, in practice the default rule will often be the one that applies. However, consumers and sellers who prefer other rules than the default rules should be allowed to contract contrary to the default. For that reason the default rules should be optional and not mandatory (De Geest 2002, p. 224).

A second type of regulatory intervention relates to information duties. One possibility is to introduce a so-called “duty to read,” which would hold consumers responsible for the consequences of not reading the contract. The contract terms will then, in other words, be legally binding to the consumers, even if they did not read. This imposition of a duty to read, however, is heavily criticized in the literature, as not linked to human reality (Eisenberg 1985); others consider the opportunity to read a “myth” (Ben-Shahar 2009). Another possibility is to impose requirements on business to make the contract terms timely and easily available (availability requirement) and to make the terms transparent, readable, and intelligible. This transparency would increase consumer reading, as it reduces search costs. Many, however, doubt whether these disclosure duties can really solve the problem. The costs of reading remain high as a result of which consumers may still refrain from reading (Katz 1990) and consumers may feel that obtaining the information is of little use if they believe that they have no alternative but to accept the standard terms (Howells 2005, p. 358). Moreover, disclosure duties can overburden consumers with information as a result of which consumers may rationally decide that they will not understand the information anyway (Eisenberg 1985).²¹

A third possible regulatory intervention is not only to aim at information disclosure but to regulate standard terms by “blacklisting” specific onerous terms and/or by indicating low-quality, but permissible terms, for which the seller has to provide proof that the terms are not onerous in case of consumer complaints (grey lists). The economic argument would be that if it can be shown that certain clauses would never be agreed to by consumers, these clauses should be banned (De Geest 2002, p. 225). Others, however, argue that this would amount to a restriction of contractual freedom, which will lead to welfare losses for those consumers who would read and shop for or negotiate other terms (Hatzis 2008).

The blacklisting of certain onerous terms assumes that courts strike out these terms. However, this can only lead to a slight increase in the quality of terms, but not necessarily much. A “flea market”²² may therefore still exist since the court intervention cannot be fully effective at improving the quality of terms to the level preferred by consumers: Onerous terms can be barred from contracts, but not terms that are barely permissible and still inefficient.

In sum, information economics holds that information asymmetry and the resulting adverse selection will lead to consumer contract terms of low quality, favouring the seller over the consumer. Market mechanisms (such as reputation and consumer learning) may not sufficiently correct for this problem. Hence, drafting efficient default rules could partially remedy the signing-without-reading problem and disclosure duties. A duty to read for consumers would create an incentive and opportunity for consumers to read the terms and discipline the market. Nevertheless, the literature realizes that these remedies may not always be effective and may not result more particularly in onerous terms disappearing from the market. The important issue is that information economics focuses centrally on the provision of information as the key instrument to increase social welfare, enabling consumers to discipline the market. This is based on the central assumption that if consumers would be properly informed (e.g., as a result of information disclosure) a rational decision-making process of consumers would again be possible. It is precisely that assumption which has been challenged by behavioural law and economics.

Lessons from Behavioural Law and Economics

Main Findings

Since the early writings of Kahneman and Tversky, a rich behavioural literature has emerged challenging the assumptions of neo-classical economics. Psychological experiments, both in laboratories and in real life, but also empirical tests, have increasingly challenged some of the assumptions underlying the economic analysis of law. Not only the assumptions are challenged; it is argued that the predictions about choices people make that follow from economic theory based on rational choice do not hold. Many scholars now distinguish a variety of possible cognitive problems, sometimes under different names and headings.²³

A first issue stressed in behavioural literature is that important “bounds” on human behaviour exist. This notion of “bounded rationality” refers to the fact that individuals often take shortcuts in making decisions that frequently result in choices that fail to satisfy the utility maximization prediction.²⁴ Individuals engage in “satisficing,” being that they do not choose the option that maximizes their utility, but rather the one that satisfies their aspirations (Wibisana 2008, p. 230). A related problem is that when the quantity of information provided is too large individuals have difficulties to evaluate the information accurately. This is referred to as information overload (Jacoby 1984; Schwartz 2004).

Another deviation from the standard model identified in behavioural studies relates to the fact that people pay more attention to absolute outcomes than to the probability that an adverse event may occur. In connection, people may be highly adverse against relatively small risks (e.g., of exposure to toxic chemicals) even if the probability of such hazard to occur is almost negligible. People are particularly inclined to avert from any action that has a (personal) connotation of severe suffering or death. This is sometimes referred to as the dread factor (Slovic 1987).

Another bias is that the easier an event comes to people’s mind the more likely the risk seems to occur. It is referred to as the “availability heuristic”: a mental shortcut that individuals use on the basis of which they assume that memorable events are memorable precisely because they are common. In reality these estimates can be biased and largely unrelated to the objective statistical probability of certain events occurring (Kuran and Sunstein 1999). This explains why the frequencies of highly publicized causes of death (e.g., homicides, fires, tornados, and cancer) are relatively overestimated and underpublicized causes (e.g., diabetes, stroke, asthma, and tuberculoses) are underestimated (Ogus 2000, p. 236).

A third bias, referred to as the status quo bias, describes how people have a preference for leaving things as they are. It is related to the so-called “endowment effect” as a result of which individuals place a higher monetary value on items they own than on those they do not own yet.²⁵ The results from many experiments show that, all things being equal, individuals will prefer a status quo outcome.

An interesting deviation refers to the self-serving bias, also referred to as selective optimism or overconfidence: Individuals tend to generalize information based on highly selective examples that suit them well (Ogus 2000, p. 237). Bad things are more likely than average to happen to the others (Jolls 1998, p. 1653). Many experiments have provided evidence of this selective optimism.²⁶ The overconfidence is stronger in cases when the individual has a degree of control over the event. Therefore, studies showed that 90% of car drivers thought that they drove more safely than the average driver (Ogus 2000, pp. 237–238). Unrealistic optimism leads to an underestimation of the probability that unpleasant events will occur.²⁷

We could easily continue this list with discussing yet other heuristics and biases as they have appeared from the behavioural literature.²⁸ To conclude this discussion of the behavioural literature, it is important to mention that there are not just a few studies or experiments, but that some of the described biases are confirmed by a substantial number of studies, based on experiments or empirical studies in different circumstances. Moreover, for the topic of this paper it is important to stress that the evidence also points in the direction of biased decision making by consumers. Several examples can illustrate this.²⁹ Most of the heuristics and biases point at cognitive deviations on the side of consumers and hence towards opportunities for sellers to take advantage of consumers. A well-known example provided by Bar-Gill relates to fitness club subscriptions and price plans for mobile phones: Consumers often err in the assessment of the extent to which they will make use of the respective product or service. Overconfidence (e.g., the assumption that one will visit a fitness club every week) may lead to the situation that a consumer is lured into a contract that does not maximize his utility (Bar-Gill 2007). In cases of consumer credits overoptimism may lead consumers to accept loans and repayment schedules which in reality are more difficult to comply with than anticipated (Bar-Gill 2008, p. 763). Moreover, money-back guarantees and no-risk trial periods create in the consumer a sense of ownership. These marketing strategies take advantage of the above mentioned endowment effect as a result of which the consumer will value the contract more than he would have beforehand. As a result the consumer will keep the good instead of returning it as he thought he would do (Hanson and Kysar 1999). Due to the endowment effect, consumers will tend to stay with the provider they already contracted with, even when competitors offer better options (Brennan 2005). The list could be enlarged with many other examples. The most important conclusion is that as a result of these heuristics and biases consumer decision-making might be suboptimal, leading them to enter into contracts that do not maximize their utility and hence do not maximize social welfare. This raises questions concerning the policy implications of behavioural literature.

Policy Implications

So far we made clear that much of the empirical findings of behavioural studies show that individuals (and more particularly consumers) do not always act in the way presumed by traditional economics on which much consumer policy is based. The next question is whether this should necessarily lead to consequences at the policy level. Generally, law and economics scholars are willing to accept the outcomes of behavioural studies, but for a variety of reasons³⁰ they wonder whether the finding of these human errors should necessarily lead to regulatory interventions (through aggressive regulation). Some argue that it is more fruitful to examine how individuals themselves are capable to (at least partially) remedy errors by learning in order to improve their situation.³¹ John List, for example, has shown through experiments that as participants gain experience in market transactions some of the cognitive biases disappear (List 2003, 2006). Some have, more generally, held that the so-called debiasing and rebiasing can counteract cognitive and motivational distortions (Jolls and Sunstein 2006; Tor 2008). This literature therefore does not deny the findings of behavioural economics, but is rather critical of the normative implication that regulation would be necessary to correct these human errors. For example, Ogus argues that one should be careful with paternalistic interventions, based on cognitive biases, since there may still be welfare maximization notwithstanding the biases and hence no need for regulatory intervention. Moreover, he warns that if such intervention takes place the question arises whether the benefits outweigh the costs (Ogus 2000, pp. 250–252). Korobkin and Ulen argue that in some cases more empirical research is needed for policymakers to be able to make effective use of the insights provided by behavioural literature³² and that in other cases the legal implications of a particular behavioural phenomenon may not be that clear-cut.³³ This is why some have argued in favour of so-called “soft paternalism” or “asymmetric paternalism”³⁴ to “nudge” individuals into an efficient decision-making without necessarily forcing them to do so.³⁵

This debate concerning the use of behavioural insights at the policy level, of course, also plays in the area of consumer policy. According to some, behavioural economics has already become an influential source for consumer policy (Ramsay 2007, p. 71). Behavioural insights may indeed provide arguments in favour of government intervention in consumer decision-making and could thus advance alternative reasons, other than economic arguments, for government intervention.³⁶ Behavioural insights are especially interesting concerning the use of information disclosure which is strongly advocated on the basis of information economics. Behavioural literature shows that acquiring and assessing information may be more troublesome for consumers than is implied by information economics. Especially due to information overload the quality of consumer decisions can actually decrease. The complexity of the problem at hand and possible ambiguity in the context of options may hinder the extent to which consumers are able to make decisions that correspond to their (known) preferences (Korobkin and Ulen 2000, pp. 28–32). Providing information to consumers without paying attention to the format, quantity, and effectiveness of the disclosure can be inefficient or have adverse effects (Jolls et al. 1998, pp. 1533–1535). When given highly technical or excessive information consumers may “switch off” and hence not reap the benefits of being informed. Information disclosure often does not address the form of the information, whereas behavioural research indicates that the framing of choices (via adequate presentation) influences consumer decision-making to a large extent (Jolls et al. 1998, pp. 1533–1535). In sum, in the presence of cognitive errors information disclosure may remain ineffective (Rachlinski 2003). Of course, to some extent already consumer lawyers had warned that the ideal of an informed consumer may be irrealistic. Hence, behavioural scholars are certainly not the first ones to point at biases and problems in consumer decision-making (see, e.g., Kroeber-Riel 1984, p. 662). However, behavioural economics now allows researchers to back up some of these earlier more intuitive insights with sound empirical research. Furthermore, behavioural insights can shed light on possible approaches to overcome these problems in consumer decision-making.

Biased decision-making of consumers may in general be an argument in favour of paternalistically protecting consumers from their own errors. Taking into account the negative effects of information overload consumers could be made better off by even limiting choice instead of allowing consumers all possible options (hard paternalism). “Nudging” consumers to decision-making which enhances their welfare, while allowing them to choose differently, does not limit consumer choice and is therefore preferred by the proponents of soft paternalism (as referred to above). However, some will still argue that even a strategy based on soft paternalism may interfere with the free will of individuals and may hence not be welfare maximizing for all consumers.

A Few Warnings

In the previous section, we already indicated that the use of the results of cognitive psychology at the policy level does not go undisputed.³⁷ We argue that these criticisms on the behavioural approach are certainly to be taken seriously, but should not necessarily lead to a rejection of the approach. Rather, being aware of the limits of behavioural studies allows one to see the potential for using these behavioural insights at the policy level more clearly. The discussion of a few considerations may provide clarity.³⁸

A first issue often mentioned is that many of the behavioural findings are based on experimental settings. The real world environment could be systematically different than the experimental one (Posner 1998, p. 1570). This is related to the fact that the behavioural results are context specific and apply to the situation in which they were observed. Hence, behavioural insights do not allow for an overly extensive generalization of observations (Rachlinski 2000, pp. 743–744). It is important to take this context specificity into account when deriving policy recommendations from behavioural studies. The results of behavioural studies may hold only within the particular setting of that study and should be interpreted with care. Social scientists through their rigorous methodological training are aware of the risks of overgeneralizing the results of just a few experiments. An observation should result from several studies, in various settings, before policy implications can be derived. Policy makers or others who are deriving implications from these studies for policy might be less accustomed to overgeneralisation issues. Methodological training could provide a solution.

Another issue, pointing at the use of paternalistic interventions by government is that this supposes that standards set by government via regulation will necessarily be superior to the market outcome. However, even when a behavioural bias is proven to result in detrimental effects to consumer welfare and correction of this bias seems justified, public errors are as realistic a problem as private errors (Glaeser 2006, p. 134). The problem arises that policy makers, just like any other individual are prone to suffer from biased decision-making themselves. Even though this is undoubtedly a serious problem one could argue that generally policy makers may have more information and a better overview of that information than individuals. Even though policy makers may hence be confronted with biases, their decisions might be less flawed than those of consumers (Loewenstein and Haisley 2008, p. 214).

Yet another element to take into account is that government interventions that aim to correct a market failure or biased decision-making may cause several types of costs that can also be a burden to social welfare.

These and other observations may lead to cautions and warnings, especially as far as the normative use of the behavioural insights for consumer policy is concerned. However, we argue that while these limits and considerations should certainly be taken into account, they should not lead to a rejection of the behavioural approach all together. For example, the context specificity of behavioural insights implies that a behaviourally informed consumer policy should address only specific issues and refrain from providing general solutions (Luth 2010, p. 126). Also, the fact that policy makers may be biased themselves could be countered by a scientifically sound basis for policy analysis, whereby sound and preferably empirical studies are relied upon to provide a reliable basis for consumer policy. Moreover, by focusing on the limits of the behavioural approach one could get the wrong impression that only this approach has limits, whereas the neoclassical approach may have (other) limits such as the extent to which it is able to provide verifiable predictions. Finally, the fact that regulatory interventions (even in the form of so-called soft paternalism) may come at a cost should be taken into account by examining whether the marginal benefits of the regulatory intervention (in curing a market failure or remedying a cognitive bias) are higher than its marginal costs. It is, in other words, possible to take the limitations of the behavioural studies into account and still use them in a behaviourally informed consumer policy. We will now address how this could be done in the specific case of standard terms in consumer contracts.

The Case of Standard Terms in Consumer Contracts

In “Traditional Economic Approach,” we sketched some of the main lessons provided by traditional information economics as far as the market failure caused by standard terms is concerned: Because of the signing-without-reading-problem and hardly comprehensible standard contract terms, information economics suggests remedies that aim at improving the provision of information: making default rules available, but especially information duties (a duty to read and/or mandatory disclosure of information). The question we will now address is how the remedies suggested by traditional information economics³⁹ should be considered in the light of the results of the behavioural literature presented in “Lessons from Behavioural Law and Economics.”

Consumer Biases and Implications

Many of the biases we discussed in “Lessons from Behavioural Law and Economics” play a role when it comes to the ability of consumers to assess standard terms. A first issue is the already mentioned problem of information overload, connected to the propensity to read standard terms, which leads to the claim that consumers simply do not read standard terms. For a consumer to attach a fair price to terms they should have a deep understanding of the market, its cost structures and roughly know what competitors are offering, which most consumers do not possess (Ben-Shahar 2009, pp. 17–20). In order to deal with the large amount of information provided in the assessment of standard terms, consumers engage in the above mentioned process of satisficing rather than in optimal decision-making.⁴⁰ Moreover, consumers are often attracted to the product or service itself and will hence not let the standard terms impede the transaction. This relates to the so-called sunk cost effect: Consumers have already invested time and effort in choosing a product or service and hence do not wish to have wasted time and effort when at the end of the transaction process the standard terms are presented (Thaler 1980). Overoptimism and self-serving bias play a role in the sense that consumers often wrongly interpret standard terms more favourably than they really are (Solan et al. 2008). It is only after a dispute has arisen that consumers wish to know where they stand and become aware of the contents of the terms (Becher and Zarsky 2008). Moreover, other biases, like the status quo bias (which may lead consumers to stick to standard terms) and probability neglect (tendency to eliminate events with low probability as impossible), will all lead to the same result, being that consumers are even less likely to read standard terms than was already hypothesised by information economics.

Empirical Results

Before turning to the question whether these biases and heuristics justify interventions at the policy level we point to a number of studies that provide empirical evidence of how cognitive deviations affect consumer behaviour with respect to standard terms.⁴¹ There are quite a few studies testing the extent to which e-standard forms on internet are read. When asked whether they would read only 4% of participants reported that they would read e-standard terms (Hillman 2006). Studies that examine actual reading behaviour come up with even lower percentages; one study reports that only 0.2% of consumers accessed the standard form on the website (Bakos et al. 2009).⁴² Other studies found remarkable differences in reported reading behaviour depending upon the type of contract. Terms on downloading software are only reportedly read in full by 5% of consumers (so called “click-wrap” contracts), whereas mortgage contracts are reportedly read fully by 73% of consumers and car rental contracts also (reportedly) by 72% of consumers (Stark and Choplin 2009). This shows that whether or not contract terms are read depends strongly upon the context.

Empirical evidence also provides information on the effectiveness of disclosure duties. Availability (making the contract term accessible) does as such (not surprisingly) not guarantee that consumers read: Consumers do not read online standard forms “regardless of how accessible they are” (Bakos et al. 2009, pp. 4–5). Click-wrap contracts are available and easily accessible; that does not have a positive impact on the extent to which consumers actually inform themselves regarding the terms of the contracts they enter into. Moreover, the data suggest that terms are likely to be one-sided (favouring the seller) and generally of low quality (Marotta-Wurgler 2007). Notwithstanding disclosure duties and in contrast with the likely one-sidedness of terms, empirical evidence suggests that consumers continue (as a result of various biases, such as overoptimism) to have incorrect interpretations of the contract terms. Employees thought that they could only be fired for cause had in fact signed a contract in which they could be fired at will (Sunstein 2001) and also tenants believe that the terms of their contract are more favourable than they actually are (Müller 1970). These empirical studies hence suggest that disclosure duties are not effective in increasing consumer reading or their understanding of the conditions of the contract. These insights are surely not new for anyone working in the field of consumer policy (see, e.g., Incardona and Poncibò 2007; Richkowskiy and Döring 2008), but the empirical studies provide further support for these rather intuitive insights. As the effectiveness of information duties is undermined by consumer biases and heuristics, consumers are even less likely to be able to discipline the market than information economics have theorised. If governments wish to improve the quality of standard terms, policy suggestions that look beyond the provision of information and the reliance upon consumer vigilance should be considered (Luth 2010, pp. 204–205).

Policy Recommendations

Above we have indicated a few of the policy implications of the behavioural literature for consumer policy in general⁴³: behavioural insights predict that as a result of information overload and emotional status consumers fail to read and assess standard terms, which will lead to low quality terms; consumer vigilance will not be able to discipline the market and an intervention strategy should aim at rebiasing or debiasing consumers. These insights have been confirmed by empirical evidence with respect to the reading and understanding of standard terms in consumer contracts: the (limited) empirical evidence available suggests that consumers are unlikely to read terms and moreover, that information duties are unlikely to improve reading. Terms are likely to be one-sided (favouring the seller) and generally of low quality. As a result of these low quality terms the adverse selection problem will remain and terms will be offered that do not generally correspond to consumer preferences.

A general conclusion from this empirical evidence is, to put it bluntly, that a consumer policy that relies on consumer vigilance and information duties, simply does not work. Hence, these behavioural insights and experimental data may warrant a change in consumer protection policy as a reliance on information duties is undermined by consumer biases and heuristics.

European Consumer Policy

We will now examine to what extent the lessons from behavioural law and economics correspond to actual policy with respect to standard contract terms. Of course, it is again impossible to provide a detailed overview within the scope of this paper. We will thus show how consumer policy has, to a large extent, relied on information disclosure as the main instrument to counteract information asymmetry. This, so we argue, largely corresponds to information economics, whereas taking the behavioural insights into account may warrant a change of policy.

Procedural Rules and Substantive Tests

The main policy instrument to counteract information asymmetry in EU consumer policy is information disclosure (Howells 2005, pp. 352–353). Informed consumers are pivotal in driving efficient markets.⁴⁴ EU consumer law has to an important extent imposed several procedural rules concerning disclosure, more particularly on sellers. Contract terms need to be easily available and transparent. For example, the proposal for a directive on consumer rights⁴⁵ provides in article 31 (1) that contract terms shall be expressed in plain, intelligible language and be legible. Article 31 (2) provides that contract terms shall be made available to the consumer in a manner which gives him a real opportunity of becoming acquainted with the terms before the conclusion of the contract, with due regard to the means of communication used.⁴⁶ These procedural rules all aim to improve the situation of the consumer through the provision of information. The idea is that in this way an informed decision-making could take place (Wilhelmsson and Twigg-Flesner 2006, pp. 449–450).

In addition the various legal systems have implemented a substantive test on which contract terms are evaluated, the so-called fairness test. It can inter alia be found in article 3 of Directive 93/13 on unfair terms in consumer contracts. Under this directive a term 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” (Weatherill 2005, pp. 118–120). These substantive tests (which are differently implemented in the various legal systems) allow courts to strike onerous terms from consumer contracts.⁴⁷

Moreover, EU law and more particularly the already mentioned unfair terms Directive provided so-called black and grey lists: terms on the black list should always be considered onerous, which renders them null and void; terms on the grey list can be employed, but are presumed to be onerous.

Based on Information Economics

The European unfair terms regimes seem to correspond mainly to the insights of information economics: the general view is that sellers abuse consumers and that this could be prevented by the remedies mentioned above. Nowadays, in consumer policy, adverse selection (resulting from information asymmetry) is considered to be the main rationale for intervening in consumer markets (Schillig 2008). The idea is that information duties will incentivize and enable consumers to become informed and hence relies on the idea of increasing party autonomy (Grundmann 2002, p. 280). The information paradigm in EU consumer law is therefore strongly based on rational choice theory (Micklitz 2008, p. 21). The basic idea which can be found in consumer law is that information asymmetries should be targeted by information remedies. However, information economics also recognizes that rationally apathetic consumers may in some cases not read or adequately assess standard terms. Hence, substantive tests, black and grey lists may also be justified on the basis of information economics.⁴⁸

A Behavioural Critique

The behavioural critique on the European consumer policy which is largely based on information disclosure may now be pretty obvious: it sheds doubt on the effectiveness of disclosure duties. Empirical insights do not only show that the reality of consumer decision-making does not correspond to the (economic) assumptions, but also that the resulting predictions do not hold.⁴⁹

Interestingly this behavioural critique goes hand in hand with critique that was equally formulated in legal doctrine. There it has been argued that consumers are unable to assess standard terms sufficiently (Ben-Shahar 2009, pp. 3–7) and that consumers, even after disclosure, cannot effectively make rational, well-balanced decisions of what would be in their interest (Wilhelmsson 2006, pp. 54–56). The behavioural debate and the empirical research therefore largely support the critique formulated in legal doctrine. More difficult to answer is the question what would then be the policy advice that can be formulated on the basis of behavioural literature.

Need for more Substantive Control?

From behavioural insights it can be argued that information disclosure duties have been empirically shown not to work. One thing that clearly follows from behavioural insights and the empirics is that aiming to improve consumer reading through regulatory intervention makes little sense for the simple reason that extensive informed decision-making with respect to standard terms is not likely to take place anyway (Korobkin 2003, p. 1246). The most radical proposal, being to prohibit standard terms all together (for the simple reason that consumers do not read) would for obvious reasons go too far as well. It would lead to a huge increase in transaction costs as a result of which either every simple contract would have to be negotiated in detail or the general rules of private law (e.g., contained in a civil code) would apply as default rules (De Geest 2002, p. 224). Default rules in contract law are often too general to deal with all contingencies that could arise in all business sectors. Standard terms still have an important information function that would disappear if the use of standard terms would be totally prohibited (Becher and Zarsky 2008) and of course the proposal to require parties to negotiate all terms of every contract they conclude is not be taken seriously for the simple reason that it would lead to excessive transaction costs (Korobkin 2003, p. 1245).

Another option which deserves more attention is to provide more substantive control of standard terms. Since information disclosure may not remedy the problem of adverse selection and low quality terms will remain into existence, more substantive control of terms could avoid welfare losses due to the use of one sided terms. Substantive control would focus on the content of the contract (as is not the case with information disclosure, which is a form of procedural control). To some extent, one could argue that this is already imbedded in the application of the fairness test. However, the application of the fairness test does not guarantee that only terms that are efficient would be used in the contract. Moreover, the application of the fairness test is often limited to individual cases. The effectiveness and efficiency of relying on fairness tests to enhance the quality of standard terms can therefore be debated.

First, fairness tests provide a threshold level of quality for standard terms. Which threshold level is efficient might depend on business sector specificities, such as cost structures, type of business model that is prevalent in the sector, whether the good to be acquired by the consumer is expensive, and so on. One fairness test for all consumer contracts might be set too low in one business sector, and too high in another. Regarding the level of quality that is aimed for in fairness tests, one could also have a look at black and grey lists provided in consumer protection regulation. These lists have an indicative effect regarding the assessment of the fairness of contract clauses. Terms that are barred through these black and grey lists are quite obviously onerous⁵⁰; this raises doubts with respect to the level of quality which results from applying the fairness test standard.

Second, the effectiveness of fairness tests depends upon enforcement; many difficulties regarding enforcement of standard terms regulation can be pointed out. Consumers suffer from rational apathy not only regarding the assessment of standard terms, but also with respect to challenging unfair terms in court (Trebilcock 2003, pp. 68–98). Moreover, the effectiveness of court enforcement of fairness tests is disputed (Collins 2004). Even when a court finds that a particular term is onerous, this finding only holds against that seller using that specific term. Consumers can therefore still be confronted with a similar term by other sellers, or even by that same seller. Consumers are unlikely to be aware that an onerous term is used in the contract. The “worst thing” that can happen to a seller applying an unfair term is an injunction; whether this will entice a seller to refrain from using unfair terms in consumer contracts is debatable. Behavioural insights furthermore suggest that consumers believe terms to hold up in court, even when they themselves consider them to be unfair (Eigen 2009).

Therefore there seems to be a need for more substantive control regarding standard terms, going beyond the fairness test. The idea of substantive control implies that a body would (ex ante or ex post) test the quality of the standard terms, applying criteria of efficiency and examining whether the terms lead to a maximization of aggregate social welfare. Such a substantive control could take various forms.⁵¹ One possibility would be an ex ante or ex post administrative control of consumer standard terms (Gillette 2005, pp. 975–1013). Another option would be to have a negotiation between all parties involved (consumer organizations and organizations representing business) who would, supervised by an independent authority (perhaps like an ombudsman) jointly negotiate model standard terms (Collins 2004, pp. 798–802). Of both models examples exist: Israel has experience with pre-approving standard terms (on a voluntary basis) and in the Netherlands draft model forms of standard terms are negotiated within the framework of the social economic council where businesses as well as consumers are represented.

Naturally, although both models may have advantages, they have problems as well, as a result of which they should not necessarily be presented as the miracle solution. One question which of course arises is whether the model standard form (either negotiated or via administrative control) will necessarily be leading to efficient standard terms. One could at least presume that since in this case more parties are represented that may act upon more information the likelihood of efficient terms is larger than when terms are one-sidedly imposed by business. Another problem is that the use of model terms may have anti-competitive effects since firms would no longer compete on the basis of the quality of standard terms. But then again one can wonder to what extent differences in standard terms today have any competitive impact on consumers. Empirical evidence shows that this is not at all the case.⁵²

In addition to other dangers (like weak consumer representation and regulatory capture), the danger of a model contract form is its generality: it may be attractive for the average consumer or the majority of consumers, but not necessarily for consumers that may wish for onerous terms to be included in their contracts, as long as they are compensated for that through a lower price. That paternalistic effect of model terms will be more burdensome in cases where model terms would be mandatory instead of merely optional. In fact, when negotiated terms are presented to both firms and consumers as the default option, the policy intervention can be characterized as a nudge following the insights of libertarian paternalism. The default is changed to a setting that is considered more efficient from a social welfare perspective, but actors are allowed to opt out. A mandatory prescription of negotiated terms would correspond to a more rigid form a paternalism, which encroaches on the freedom of contract of both sellers and consumers.

In sum, we do not claim that a policy strategy aiming at more substantive control of standard terms is a miracle solution. But at least it seems, on the basis of behavioural literature and empirical evidence more promising than to continue a policy of information disclosure of which it is by now known that this is not effective in improving social welfare. Of course, further studies and especially empirical evidence may be needed from legal systems where experience with those model terms exists to be able to fully assess their effectiveness.

Concluding Remarks

Micklitz already pointed out that the information paradigm in European consumer policy corresponds to rational decision-making, which does not concur in all circumstances to insights from behavioural economics (Micklitz 2008, p. 21). The goal of our paper was to put forward our view towards using behavioural economics at the normative level: take the lessons from behavioural economics seriously and more particularly address its normative consequences for consumer policy. Using behavioural insights for policy should in our view be done quite cautiously for the simple reason that although there now is an impressive body of behavioural literature, the discipline is still relatively young and much more research may be needed to come to firm conclusions.

The existing insights from behavioural studies show that these can have an important added value in addition to the traditional economic paradigm of merely intervening in consumer markets to cure the market failure caused by information asymmetry. That policy, which is strongly based on information disclosure and consumer vigilance, so the evidence shows, may not be effective in really leading consumers towards a more informed decision-making, especially as far standard terms are concerned. Of course this shows an inherent weakness of the behavioural approach which is, given the fact that a general theory is lacking, today still better at listing anomalies without providing an alternative integrated theory.⁵³

In the case of standard terms in consumer contracts, adding behavioural economics to information economics may lead to a call for more substantive control of standard terms. We do realize that this recommendation is not unproblematic and will lead to an often heard criticism on the behavioural approach, being that it often leads to paternalistic calls for regulation based on cognitive biases (Ogus 2000, pp. 250–252). However, the alternative for paternalism is relying on autonomous consumer choice which may not lead to wealth maximization either in the case where this choice is biased. Of course a paternalistic intervention may come at high costs especially for the vigilant consumers which do not have a preference for the additional protection. Moreover, the behavioural studies with respect to standard terms⁵⁴ pointed to the context specificity of the results: on average consumers claimed, for example, not to read contract terms, but when it concerned standard terms for a mortgage or the lease of a car a majority of consumers claimed that standard terms were read.⁵⁵ Context specificity does not permit to deduce generalisable results from the experiments. Whether, for example, a substantive intervention via model terms may be desirable may thus well depend upon the specific market and the existence of biases in that particular context.

Using behavioural insights, it can be argued that consumers are unlikely to be able to discipline the market in providing efficient standard terms, even when they are aided in this endeavour by information remedies. Biases and heuristics impair consumer decision making and consumer vigilance. Neither competition nor litigation will induce sellers to draft efficient terms in consumer contracts. In policy, the dependence on consumer vigilance and information duties to stimulate this vigilance is likely to be insufficient. Hence, behavioural insights argue for a more limited dependence upon information disclosure and consumer vigilance in standard terms in consumer policy, which is a clear contrast with several policy initiatives. While we caution against overgeneralizing results from behavioural studies, in our view behavioural insights can and should be used to draft effective and efficient behaviourally informed consumer policy.