Autonomy as license to operate: Establishing the internal and external conditions of informed choice in marketing

The assumption that consumers voluntarily accept or decline marketing offerings provides the ethical justification that gives marketing as a social system its license to operate. Consumer autonomy is, therefore, the key ethical principle of marketing in capitalistic economies. However, even in domains with extensive regulatory frameworks and advanced market conditions, consumers are often ill-informed or underinformed. The resultant lack of epistemic confidence diminishes consumers’ ability to make informed choices. At the same time, consumers are by default exposed to promotional content designed to persuade them to accept marketing offerings. This threatens personal autonomy. We develop a concept of consumer autonomy which marketing regulations should protect and promote to enhance informed decision-making. We design autonomy to be robust in situations where individuals are exposed to persuasive attempts to influence them to choose a specific course of action. As such, our concept of autonomy is applicable to a range of contexts beyond marketing where it is necessary to balance external influences and individual autonomy.


Introduction
Consumers' ability to make autonomous decisions provide legitimacy to marketing as a social practice in capitalistic economies (Cluley, 2019;Villarán, 2017). Marketing is a social system based on mutual exchange (Lusch and Watts, 2018;Lüdicke, 2006) whereby providers and consumers trade goods or services for money (Anderson et al., 1999;Bagozzi, 1975;Houston and Gassenheimer, 1987). Such exchanges are valid ethically only if all parties knowingly and willingly agree to accept the exchange (Brenkert, 2008;Nixon and Gabriel, 2016). However, many consumers feel under-and ill-informed and are thereby not in a position to make informed decisions and act autonomously (EC, 2015;Eurobarometer, 2011). The issue of under-and ill-informed consumers is not limited to fairly unregulated markets: the European Union's single market provides an extensive set of marketing regulations across its 28 member states, but inadequate consumer information is nevertheless a problem across the region (EC, 2015;Eurobarometer, 2011). More than one-third of consumers across the EU single market does not feel knowledgeable (Eurobarometer, 2011) and a large proportion is unaware of key consumer rights (EC, 2015). This impedes consumers ability to exercise autonomy and make informed decisions and thereby threatens the ethical justification of marketing as a social system, where individuals' right to make informed decisions and exercise autonomy is a non-negotiable precondition.

Basic characteristics of consumer autonomy
Personal autonomy is the ability to figure out what one has good reasons to do, to align one's motivations with these reasons, and act accordingly (Buss and Westlund, 2018). Autonomy thereby comprises three different components: cognitive, volitional, and agentic. Our study is primarily concerned with autonomous decisions and therefore focuses on the cognitive and volitional aspects (the internal condition) and the market conditions that must be obtained in order to enable autonomous decision-making (the external condition). Consumer autonomy is valuable because it empowers us to accept or decline marketing offerings in accordance with and because of our personal reasons and motivations (Brenkert, 2008). This theoretical observation is manifested in European consumer law in the "Unfair Commercial Practices Directive" (EUR-LEX, 2005): A commercial practice shall be regarded as aggressive if . . . it significantly impairs or is likely to significantly impair the average consumer's freedom of choice or conduct with regard to the product and thereby causes him or is likely to cause him to take a transactional decision that he would not have taken otherwise.
Although the EU marketing directive explicitly addresses consumer autonomy (or freedom), it is nonetheless a concept which, in regulatory terms, remains fairly underdetermined. More work is needed to fully understand the importance of consumer autonomy in marketing regulation and how it can best be protected.
Consumer autonomy is associated with both internal conditions (e.g. cognitive and volitional capacities) and external conditions (e.g. epistemic market conditions such as access to information and consumer rights). The internal conditions have been subject to substantial research in analytical philosophy (Buss and Westlund, 2018), but the external conditions are largely unexplored. Reasonably, this is because the internal conditions are assumed to be stable and can be expressed in terms of exhaustive necessary and sufficient conditions that define what autonomy is in all possible scenarios. However, the external conditions are context dependent because they enable autonomous decision-making in practical contexts. For example, while the same internal conditions will define autonomy for a patient and a consumer, the external conditions will be different: in medical ethics, the predominant idea of patient autonomy is that doctors should present patients with objective information about their health condition and treatment options, but then step back and avoid making a recommendation (Kapp, 2007;Stirrat and Gill, 2005;. Autonomous choice is here assumed to require access to material information and absence of external influence. It is questionable whether absence of external influence can ever be fully obtained as any presentation of material information requires some degree of interpretation and selection and therefore can be construed as persuasion. But there is an important difference: the subjective interpretation involved in providing material information implies a degree of passive persuasion, whereas advertising usually entails fairly high levels of active persuasion. We assume that the difference in these external conditions impacts on individuals' capacity to act autonomously and we seek to develop the external conditions that are necessary for acting autonomously in the face of active persuasion. Indeed, some external influences have the capacity to undermine personal autonomy, such as harassment and coercion highlighted in the EU "Unfair Commercial Practices Directive" (EUR-LEX, 2005). This raises the critical question: what distinguishes external influences that compromise autonomy from external influences that are compatible with autonomy? This is where the discussion of autonomy becomes highly relevant to marketing ethics, as the following example demonstrates. Impulse buying is a very frequently occurring consumer behavior (Chan et al., 2017;Moser, 2018;Strack et al., 2006) that demonstrates the conflict between autonomy and marketing. It occurs when a consumer gives in to a sudden urge to buy a product in response to a marketing message, with no prior intention to make the purchase and with reduced attention to its potential adverse consequences (Piron, 1991). The phenomenon is very common indeed, with five in six Americans admitting to regular impulse buying (Kossman, 2016), quite often in response to online and offline sales promotions (Lo et al., 2016) and product reviews (Zhang et al., 2018). Impulse buying is demonstrated to be significantly correlated with consumer regret over having purchased items that, on reflection, are unwanted (Hoch and Loewenstein, 1991;Lee et al., 2015;Wood, 1998). This indicates an impairment of consumer autonomy: research into consumer psychology demonstrates that impulse buying can be understood in terms of psychological functioning and, in many cases, lack of self-regulation and self-control (Chen and Wang, 2016;Verplanken and Sato, 2011;Yi and Baumgartner, 2011). Thus, impulse buying is a prime example of a common type of consumer behavior where consumers' autonomy is significantly diminished by marketing (Baumeister, 2002).
Persuasive marketing is not always an external condition that potentially threatens autonomy: frequently it is a core part of a brand's symbolic value. Companies like Nike, Apple, and Mercedes spend vast amounts on marketing, and the symbolic values created through branding become a core part of the product (Belk, 1988). In such cases, exposure to persuasive marketing is a consumer preference rather than a threat to autonomy. Yet, this situation does not challenge the importance of protecting the consumer: autonomy does not entail shielding the consumer from exposure to persuasive marketing; rather, it is all about making sure that consumers get a fair chance to make their own decisions when exposed to persuasive marketing without feeling coerced, deceived, or misinformed.

Consumer autonomy and marketing research
Consumer autonomy is a well-established topic in marketing theory. One stream of research focuses on the extent to which marketing methods and practices undermine, or are compatible with, autonomous consumer agency (e.g. Anker et al., 2010;Arrington, 1982;Barrett, 2000;Bishop, 2000;Crisp, 1987;Cunningham, 2003;Raley, 2006;Sneddon, 2001;Villarán, 2017). An emerging stream of work in the intersection between political economy, moral philosophy, and business studies focuses on the epistemic preconditions that need to be satisfied in a capitalistic market in order for consumers to be able to make autonomous choices (e.g. De Bruin, 2015). Finally, critical marketing scholars call into question the underpinning assumption of this article by arguing that the notion of consumer autonomy is fundamentally flawed (Arnould, 2007;Fuchs, 2001). Sociological analysis of agency points to fatal problems in the concept because consumer autonomy is in effect a social ordering mechanism which classifies the majority of the population as free agents that can exercise power through choices of consumption, while in reality consumers are subject to exploitative power relations: autonomy is a social relation and specific artifact of Western market-driven economy, which the social system of capitalism needs to reproduce and sustain its existence (Arnould, 2007;Bauman, 1988;Davis, 2016;Kozinets, 2002). We accept this critique. However, at the same time, we argue that this does not diminish the importance of improving the set of social relations and external market conditions that the concept of autonomy-with all its flaws, defects and caveats-rests upon. While consumer autonomy, as Bauman (1988) argues, is a defect form of freedom that will never be characterized by strong selfdetermination and independent formation of the individual will, it nonetheless provides a set of boundary conditions that are immensely valuable to consumers, protecting them against many forms of exploitation from powerful businesses and unfair practices such as harassment, intimidation, and misleading advertising. Against that background, it is an important task to articulate the internal and external conditions of consumer autonomy as this study seeks to do.

The internal condition of consumer autonomy
There are two competing theories of autonomy known as the coherentist and the reasonsresponsiveness accounts of autonomy (Buss and Westlund, 2018). Counter to extant theorization, we demonstrate that the theories are not mutually exclusive and we synthesize them into a coherent concept of consumer autonomy.

Coherentist account of autonomy
According to the coherentist account, a person acts autonomously insofar as they act in accordance with and because of personal motivations that cohere with a mental state which represents their point of view (Dworkin, 1988;Ekstrom, 1993). But which mental states qualify as autonomy conducive? Key suggestions are higher order desires (Frankfurt, 1971), evaluative judgments (Watson, 1975), general intentions (Buss and Overton, 2002), or personal policies and life plans (Bratman, 1987). Accordingly, a person acts autonomously when there is psychological unity expressed as coherence between first-order and second-order motivations such that first-order motivations are being aligned with second-order motivations (Frankfurt, 1971). Understanding autonomy as absence of internal conflicts of will (i.e. misalignment between higher and lower order motivations) is remarkably relevant in applied contexts such as marketing.
Buying a product counts as an autonomous action insofar as the consumer's effective motivation coheres with their higher order motivations (Anker et al., 2010). Think of the example with the person buying a new car. The consumer had a preestablished intention to buy a new car and conducted extensive background research to find the type of car that would best fit their needs and budget. Accordingly, when they acted on their desire for a new car and actually bought one in response to a car dealer's sales promotion, it was an autonomous decision because the causal motivation was a first-order motivation that cohered with a preestablished second-order motivation. Contrariwise, think of a person who has a long-standing motivation to eat more healthily. The person has agreed with themselves that cake on special occasions such as birthdays is OK, otherwise not. Imagine the person passing by a bakery on their way back home after work and succumbing to a sudden desire for a doughnut: buying the doughnut compromises the person's autonomy because they act on a motivation that runs counter to their preestablished higher order motivation to eat healthily under normal circumstances. The coherentist account also explains why a common consumer practice such as impulse buying threatens autonomy. Impulse buying is frequently associated with post-purchase dissonance (Chan et al., 2017;Moser, 2018;Strack et al., 2006), which can be explained as misalignment between higher and lower order motivations that causes the agent to feel estranged from their actions or the reasons for doing them, because they did not really want to do them (Frankfurt, 1971).
The coherentist account is associated with a theoretical problem: the infinite regress. A conflict between motivations at any given level should be decided by reference to the next higher level of motivations. But if the conflict repeats at that level and any subsequent higher level, the result is an infinite regress which leads to paralysis of action. This is mainly a theoretical problem: in practice, consumers do not have infinitely many levels of higher order motivations and can clarify conflicts of will by reflecting on their higher order motivations at second or third level.

Reasons-responsiveness account of autonomy
On this account, a person acts autonomously to the extent that their actions are caused by firstorder motivations that cohere with second-order motivations that, in turn, are responsive to external reasons for and against acting as they do (Fischer and Ravizza, 1998;Wolf, 1990;Zimmerman, 2002). The aim of the reasons-responsiveness account is not to explore which external conditions agents should be responsive to, but to demonstrate that in order for an agent to act autonomously it is necessary to be responsive to certain external circumstances (whatever they may be). Although the account is referred to as externalist in the specialist philosophical literature, it is in our context a requirement that relates to the internal condition of consumer autonomy, because the focus is on how the agent internalizes external reasons in their personal decisionmaking.
The following example shows why reasons-responsiveness is important for consumers' ability to make informed choices. Payday loans are high-interest, short-term loans targeted at cashstrapped consumers who have an immediate need for a small amount of cash. Payday loans are often used to pay for everyday expenses such as utility bills, food, car costs, and credit card debt. In the US, the typical annual percentage rate is around 400% (CFPB, 2018), whereas the UK introduced a cap in 2014 of 0.8% per day and a total cost cap of 100% (FCA, 2018). While the daily interest rate may seem low, the APR could still be up to 1500% (MAS, 2018). Although some finance scholars recognize the benefits that payday loans may offer to customers (Servon, 2017), this form of finance is nevertheless correlated with a range of negative outcomes such as debtspirals, financial strain, and bankruptcy (Gathergood et al., 2019;Horowitz, 2017). Imagine a consumer who takes out a payday loan to replace their smartphone with a newer model. Because their existing phone is in good working order and acquiring a new one therefore is nonessential, friends have warned against taking a payday loan due to the very high interest rates. Being unable to feed one's children might justify a payday loan, they pointed out, but the perceived need for the latest phone is not a good enough reason. However, the consumer ignores these reasons and takes out a payday loan with an interest rate of 1500% APR to make the purchase. This lack of internalization of the external reasons against taking out the loan will qualify the consumer as nonautonomous exactly because of the seriously diminished reasons-responsiveness in the decision-making process. By contrast, the coherentist account would accept the decision as autonomous insofar as it coheres with the consumer's second-order motivation, regardless of receptiveness to the externally grounded reasons for and against.

Internal condition of autonomy
Acknowledging that both the coherentist and reasons-responsiveness accounts develop important aspects of consumer autonomy, we can now synthesize the definition into an overarching internal condition.
A consumer's decision to accept a marketing exchange is autonomous insofar as they have aligned their first-order motivations with their second-order motivations in response to, and coherent with, the external reasons they have for and against acting accordingly.

The external condition of consumer autonomy Access to material information
Availability of and access to material information is such an obvious precondition of informed decision-making that its importance is easily overlooked. However, it deserves attention as it poses a central problem in marketing ethics. Wilson (1999) and Case (2012) reviewed the most influential models of information seeking and, by way of comparative analysis, it is possible to isolate one fundamental assumption: information seeking is a rational behavior initiated by a need for information. This poses a problem in our context because the fundamental purpose of marketing is to persuade consumers, and marketers therefore actively try to minimize the perceived need for material information (Case, 2012;Sher, 2011). It is thus often in the interest of the provider to not make available substantial product information. Obviously, this impedes the ability of the consumer to act as a rational agent and threatens autonomous decision-making.
The UK Advertising Standards Authority's archive of rulings provides a range of concrete examples where marketers are responsible for a substantial lack of information that leads to consumers making choices that they regret or are otherwise dissatisfied with. A recent example is Eurostar International Ltd, a train operator, where a consumer challenged whether the claim "Trains to France with Eurostar tickets from as little as £29" was misleading as only a very small number of tickets were available at the advertised price. The ASA (2018) ruled that: Given that we considered that the tickets were not available at the lead-in price in significant proportions, the lack of information on the period when that price was available and the lack of information indicating that the fares were not reasonably evenly distributed throughout the booking period, we concluded that the ad as it originally appeared was misleading.
Marketing codes of conduct therefore need to reinforce the responsibility of marketers to widen epistemic access to material information by making it easily accessible at all key consumer contact points. Extant research on consumer journeys can inform this requirement (Lemon and Verhoef, 2016). Consumer journeys are defined as the different touchpoints where the consumer interacts with the brand during the entire decision-making process across channels (e.g. online, in store, call center) and devices (e.g. mobile, tablet, laptop, TV, smartwatch). An example of a consumer journey could be someone seeking to buy a new pair of running trainers: first, they look at other consumers' reviews on a running brand website using a laptop computer; then they head into a running store to look at trainers and, while being there, they search for additional information on their smartphone, take advice from a sales rep, and finally try out a pair on the store's treadmill before deciding to make a purchase. This is a relatively simple consumer journey with a number of brand touchpoints (initial search, product selection, test, and advice) across channels (online, in store) and devices (laptop, mobile). Marketers should ensure that material information is made easily available at the average consumer journey touchpoints across the most frequently used channels and optimized for cross-device information searches.
To be operational, the requirement to provide easy access to material information at all key touchpoints in the average consumer journey needs to be qualified with a determination of what minimum level of information the consumer can reasonably expect to receive. The most substantial discussions of consumer information sufficiency are found in Holley (1998) and Ebejer and Morden (1988). Holley's analysis of information disclosure in personal selling is of particular interest: having discussed different types of disclosure that would primarily benefit the consumer (maximum disclosure) or the provider (minimum disclosure), they argue in favor of the mutual benefit rule. "The mutual benefit rule requires the salesperson to disclose enough information to allow the customer to make a reasonable judgment about whether to purchase the product" (Holley, 1998: 638). For our purposes, we qualify Holley's benefit rule by defining sufficient information as material product and service information that the average targeted consumer cannot be expected to possess. While this qualification does not specify exactly what information need to be provided to the consumer in every single case, it provides an operational guideline that can inform marketing conduct. Thus, marketers should provide the average targeted consumer with material information that they cannot reasonably be expected to possess and which allows them to reflect on the reasons for and against purchasing the product or service.

Proportionality
Sometimes consumers are not motivated to actively search for or engage with material product information (Bakos et al., 2014). The problem is sizable: across the European Union, 24% of consumers do not read terms and conditions and 36% only read them partially (Eurobarometer, 2011). A recent study found that "only one or two of every 1,000 retail software shoppers access the license agreement and that most of those who do access it read no more than a small portion" (Bakos et al., 2014: 1). To fully appreciate the scale of underinformed consumer decision-making, consider the problem that "terms and conditions" are information that sellers are required by law to provide to the customer. The problem is what could be called "data dumping" whereby businesses are overloading consumers with such an amount of data that they cannot reasonably be expected to process all the information. Autonomous decision-making is being diminished by requiring the consumer to read and process a disproportionate amount of textual information.
We propose that marketing information is proportional insofar as the average, targeted consumer is capable of processing the information within a reasonable period of time. Though open to interpretation, the condition is operational: if a brand, say, promises that you can "Create a mobile native app for your website in minutes!" (WiziApp, 2018), but requires the consumer to read pages of terms and conditions before the purchase can be made and the app be developed, then the amount of information would be disproportionate. The principle of proportionality requires businesses and marketers to adjust the amount of information they require the consumer to process before accepting the proposed exchange to the nature of the purchase. Finally, let us consider the case of voluntary ignorance where consumers consciously choose to ignore material marketing information. One may argue that consumers who actively decide not to engage with material information should carry the full responsibility if they make sub-optimal decisions. However, we suggest that consumers' unwillingness to engage with information is often a symptom of the problem of proportionality. If brands and marketers provide relevant, understandable, and proportionate material information at key brand contact points, then the consumer will have very little reason to form an active preference to disengage with marketing information.

Epistemic integrity
Access to material and proportional information is necessary, but not sufficient: sometimes information is true but misleading because it is intended to exploit weaknesses in consumers' decision-making processes and thereby undermine autonomy (Hastak and Mazis, 2011). A clear example is misleading use of health, nutrition, and process claims in health food marketing. Many consumers are known to conflate nutrition and health claims: a nutrition claim states that a product contains a certain ingredient, whereas a health claim adds that the ingredient has a positive impact on health. Health and nutrition claims are heavily regulated in developed markets and the requirements for the use of health claims are higher and more restrictive. However, as many consumers interpret nutrition claims as de facto health claims, marketers can intentionally use true and substantiated nutrition claims to influence consumers to believe that the product is healthy, although a health claim would not be substantiated (Anker et al., 2011). Likewise, consumers tend to interpret process claims such as "organic" as health claims, although unhealthy products (such as chocolate bars and fizzy drinks) can be organic because the claim is about how the product has been produced and not about what nutrients it contains and whether these are good for your health (Anker et al., 2011). We therefore theorize that marketing communication has to respect the epistemic integrity of the consumer by not intentionally exploiting known vulnerabilities in their decision-making process through true but manipulative information.

Epistemic relevance: Context specific information
Obviously, marketers need to provide material information that is relevant to the intended use of the product. Relevance can be determined via evidence obtained through controlled tests carried out in an environment representing the intended context of consumption (Anker, 2016). For example, a sleeping bag marketed to serious mountaineers must be able to withstand freezing conditions: if the test team tries it out on a mountain expedition in sub-zero conditions and it is still comfortably warm, then a justified marketing claim can be made as to the product's ability to be used for mountaineering in sub-zero conditions. However, establishing epistemic relevance often proves much more complex, a complexity that is rooted in what we will call the "context of consumption discrepancy." This discrepancy occurs when the official context of consumption differs substantially from the average context of consumption. A telling example of context of consumption discrepancy is the marketing of sports drinks. These drinks often promise performance-enhancing benefits when used during physical exercise, and the product claims are often scientifically substantiated: laboratory tests have proved sports drinks to enhance performance (Cohen, 2012;Heneghan et al., 2012). Yet, the problem is that this only holds true under very specific circumstances, that is, when used during high-intensity workouts prolonged for more than 90 minutes (Cohen, 2012;Heneghan et al., 2012). For most nonprofessional athletes, sports drinks have no performance-enhancing properties: it is just flavored water with sugar. This would not have been a problem had the products been targeted at top athletes only, but sports drinks are often mass marketed to recreational athletes for whom there are no performance-enhancing benefits. Thus, the context of consumption discrepancy means that-although narrowly true in the idealized context of consumption-the marketing claims are unsubstantiated and thereby irrelevant and misleading in the average context of consumption. Marketers thus have to ensure that all factual claims are substantiated against the average context of consumption.
Marketing regulations also need to take into account potential contexts of consumption. If products or services have potentially serious adverse consequences when used in unintended ways, then marketing regulations need to reflect this. Accordingly, if using a product in a certain way or context may cause harm to the user or other consumers, then marketers have to clearly warn against using the product in that way or in such contexts. One example is pharmaceuticals: painkillers (over-the-counter and prescription) can easily be misused and, in worst-case scenarios, unintended consumption may be lethal. It is necessary for marketers to clearly inform about any such significant adverse effects, however statistically unlikely they are.
To sum up, a piece of marketing information is epistemically relevant insofar as it meets the following three conditions: (i) clear communication of the key product and service properties in the intended context of consumption; (ii) in cases where the likely or foreseeable context of consumption differs from the context of consumption conveyed via marketing communications, then substantiation of product benefits relative to the average context of consumption is necessary; (iii) clear warnings are communicated in terms of potential contexts of consumption where product use poses nontrivial risks to the user, other users, animals, and the environment.

Consumer understanding
For information to be useful, it must be easy to understand. This obvious observation raises the complex question of when marketers can be said to have provided understandable information. We propose a solution that provides a corrective to existing regulation in the EU.
In an early article that has informed contemporary debate of deception in advertising, Carson et al. (1985) discuss the proposition that material marketing communications should be understandable to all consumers in the relevant market. However, as they rightly point out this is a far too strong requirement: some consumers may be ignorant and unable to understand perfectly appropriate product information and marketers are not to be blamed for that (Carson et al., 1985). Imagine, for example, a situation where a consumer who is living in a rural location buys a digital audio broadcasting (DAB) radio, but cannot use it at home as planned because they can only receive analogue radio. The consumer did not understand the implication of the product information that the radio was unable to receive analogue signals, as they were unaware of the difference between DAB and analogue radio. In this situation, the consumer may reasonably be said to not have been keeping up to date with relevant developments in the radio market over the last 20 years. The consumer has not been sufficiently attentive and observant and, as a consequence, is now underinformed and therefore unable to make an autonomous decision. But this is not due to any fault of the marketing of the product. (We will return to the issue of consumers' epistemic responsibility.) The notion of the average consumer is frequently used as a benchmark against which consumer understanding should be defined (Incardona and Poncibò, 2007). While this notion is central to regulatory frameworks that enable consumer autonomy (e.g. EUR-LEX, 2005), it is not operational without qualification. The problem being the average consumer may potentially be uninformed and generally inattentive to important information about the contemporary consumer market place. If the average consumer is underinformed, then they cannot meaningfully be used as the epistemic benchmark. This is not just a theoretical possibility: in 2011, over one-third of all citizens in the EU did not feel knowledgeable as consumers, with some countries reporting more than half of their citizens feeling underinformed (e.g. Bulgaria 72%, Portugal 62%, and Greece 58%; Eurobarometer, 2011). Thus, the numerical notion of average consumer cannot be used as a benchmark for relevant consumer understanding unless it is augmented with a qualitative corrective. The EU has developed a regulatory framework that does just that. The Unfair Commercial Practices Directive (EUR-LEX, 2005) provides consumer protection from unfair business-toconsumer practices, notably protecting the consumer against deceptive and misleading marketing in order to ensure autonomous decision-making. The Directive, which is currently in force, defines the notion of the average consumer as follows: . . . this Directive takes as benchmark the average consumer, who is reasonably well-informed and reasonably observant and circumspect, taking into account social, cultural and linguistic factors, as interpreted by the Court of Justice . . . The average consumer test is not a statistical test. National courts and authorities will have to exercise their own faculty of judgement . . . to determine the typical reaction of the average consumer in a given case. (EUR-LEX, 2005: §18) However, assuming that the average consumer is reasonably well informed is often flawed: as we have just seen the majority of consumers in some EU countries within the jurisdiction of the Directive do not feel knowledgeable as consumers (Eurobarometer, 2011). Furthermore, being reasonably well informed in a general sense is not necessarily sufficient to ensure consumer autonomy. First, some product categories require more information than others to enable autonomous consumer decisions. For example, buying a bag of potatoes is likely to require less information and cognitive processing by the consumer than, say, purchasing a summer holiday for all the family in Indonesia. The latter may require substantial information about price comparisons on flights and hotels, local prices on food and transportation, visa restrictions and assessment of information from the Foreign Office (which in the UK at least advises against all but essential travel to some areas), and so on. The cognitive state of being well informed is relative to the context of consumption and, as such, different levels of consumer understanding is required for different types of products and services in order for the consumer to be able to make an autonomous decision.
Our proposition is to change the focus from the average consumer to the targeted consumer. By focusing on the target group as the benchmark, material information needs to be tailored and made understandable to the actual end user instead of the normative ideal of the average consumer defined as a reasonably well-informed individual. However, target groups may vary dramatically in size and also comprise of rather heterogeneous groups of consumers with few commonalities apart from one key segmentation characteristic. Thus, the notion of the target consumer is still too broad. To see why this is the case, let us use the previous example with a brand that produces outdoor clothing and equipment: part of the product range may be sleeping bags that are targeted at all consumers in a given market that are interested in wild camping. The target group comprises young people (e.g. scouts), parents to young families, and retired people, who all share an interest in exploring nature. This preference is the key consumer characteristic and would be defined as a psychographic segmentation variable (Wedel and Kamakura, 2000). Apart from sharing this one segmentation variable, the target group is very heterogeneous (i.e. demographic segmentation characteristics differ across the target group). The implication is that even in a specific target group the cognitive capacity to process and understand material information may be significantly variable. On the unqualified assumption that marketing information should be understandable to the target group, we would have to assume that all targeted consumers should be able to understand all product information to the same degree; otherwise, the brand would have failed its responsibility to inform the consumer. This is a very strong requirement, given the heterogeneous nature of manyif not most-target groups. We therefore introduce the following principle: the notion of the average consumer is relative to the target group such that material information should be understandable to the average consumer in the target group.

External condition of autonomy
We can now define the external condition as follows: A decision to accept a proposed marketing exchange is autonomous to the extent that: (i) the consumer has had easy access-at key touchpoints and across channels and devices-to information that is relevant, proportionate, sufficient and understandable to the average, targeted consumer; and (ii) the decision is formed in response to the consumer's critical reflection on the information.

Consumer epistemic responsibility
So far, the onus has been on marketers to provide consumers with easy access to relevant and understandable information. However, the consumer also has a significant epistemic responsibility to be receptive to and engage with the information in the right way. The concept of consumer responsibility is underresearched, as Wells et al. (2011) observe, but there is growing interest in the topic.  have made a significant contribution to a first-generation conceptualization of consumer responsibility and there are numerous topical articles in the area, in particular focusing on consumer responsibility in sustainable and ethical consumption (e.g. Gallego and Lenzen, 2005;Lenzen et al., 2007;Rodrigues and Domingos, 2008). Recently, Caruana and Chatzidakis (2014) introduced the notion of consumer social responsibility, emphasizing the active responsibility of consumers in activating corporate social responsibility.
We introduce an epistemic notion of consumer responsibility and argue that consumers-in order to be capable of making informed choices-have to demonstrate a willingness to engage with brands over time to stay sufficiently knowledgeable about the relevant context of consumption. It is useful to distinguish between pre-and postpurchase consumer responsibility as this study is concerned with prepurchase conditions only, focusing on regulations of marketing communications that aim at persuading consumers to make a purchase. We are concerned with two prepurchase functions: first, being receptive to relevant information; second, using information to make decisions (reflexive agency).
It is easy to establish why receptiveness to marketing information is an important consumer responsibility. Imagine that a brand has done an impeccable job of offering their target consumers high-quality, easily available, and understandable information at key brand touchpoints, but some consumers actively ignore the information when deciding to purchase the product. The absence of relevant information in the decision-making process impacts negatively on consumer autonomy. However, it is the consumer and not the marketer that is at fault. Consumers thereby have an obligation not to ignore material product information that is made available as part of the marketing of the product and to engage proactively with relevant information.
While receptiveness to and understanding of information are cognitive dimensions, the notion of reflexive agency draws on both the cognitive and volitional components of the internal condition of autonomy as it requires the consumer to be reasons-responsive and align higher and lower order motivations. To make an autonomous decision, consumers have not just to be adequately informed, they also need to use this information to guide their decision-making. It is necessary to critically reflect on the implications for their own circumstances of accepting a marketing offering and then align their motivations such that they make decisions that truly reflect their own reasons and motivations. This is the case when consumers are effectively motivated by first-order motivations that are coherent with their corresponding second-order motivations. To see how this works in practice, think back to the example with the consumer that wanted to take out a payday loan with a very high APR of 1500%. Further imagine that the consumer has a second-order motivation not to take out the loan, but nevertheless-after having reflected on the product information and listened to the precautionary advice from their friends-decides to go ahead with the loan. They are thus well informed of the reasons for and indeed against taking out the loan and also higher order motivation not to proceed with the loan, and yet they decide to ignore those reasons and act against their own higher order motivation. By accepting the loan, the lack of reflexive agency violates the cognitive and volitional requirements of the internal condition of autonomy because the consumer is not being reasons-responsive and acts on first-order motivations that are misaligned with their second-order motivations.

Nudging autonomy
While the consumer is responsible for exercising reflexive agency to act autonomously, we argue that it is possible for marketers to stimulate this activity by nudging consumers. Nudging is a wellestablished approach in behavioral sciences whereby individuals are influenced to make decisions that lead to certain types of desirable behaviors by making small-but significant-changes in the environment in which the individual is immersed (Thaler and Sunstein, 2008). Nudge scholars often distinguish between two types of information processing: Automatic System (or System 1) processing is fast, unconscious, and often relying on heuristics to arrive at conclusions with minimum amount of cognitive processing; Reflective System (or System 2) processing, by contrast, is slow, conscious, reflective, and cognitively demanding (Abdukadirov, 2016;Saghai, 2013;Thaler and Sunstein, 2008). Nudges typically use carefully designed Automatic System heuristics to activate a certain desired behavior (Abdukadirov, 2016;Saghai, 2013). One of the most frequently used examples-first featured in Thaler and Sunstein's seminal book Nudge (2008)-is from the Schipol Airport in Amsterdam where images of black houseflies were etched onto men's urinals to encourage users to aim for the fly and thereby reduce spillage. The intervention reportedly reduced spillage by 80% by harnessing Automatic System processing that influenced gents to concentrate while urinating by aiming at the fly.
Social proofing is one method that can be used effectively to nudge consumers into reflexive action by exposing them to information about how others behave (Aronson et al., 2014). For example, providing hotel guests with normative information such as "The majority of guests reuse their towels. You can save the environment by doing the same" is demonstrated to positively influence guests to reuse their towels (Goldstein et al., 2008). We argue that marketers can use social proofing to influence consumers to actively reflect on the reasons for and against making a purchase or accepting a deal. For example, marketing information about payday loans could be augmented with social proofing messages such as "The majority of consumers who have been taking professional advice from a bank advisor have chosen not to take out payday loans" or "The majority of consumers have trouble repaying payday loans. Don't do it if you're in doubt!" Building on insights from dialogical marketing communication practices (Ballantyne, 2004;Ylimäki and Vesalainen, 2015), future research can develop different types of nudges specifically designed to activate consumers' critical engagement with material marketing information and thereby enhance consumer autonomy in the marketplace.
However, some might argue that the idea of trying to nudge someone into being more autonomous is an oxymoron because it is impossible to manipulate others to enhance their autonomy. It is a contradiction. We accept that our proposal is influencing consumers without their knowledge and consent, but we argue that this is permissible because it does not limit the total set of choices available to the consumer or influence them to favoring any one option over another. To be compatible with autonomy, nudging should be limited to activating reflexive agency without prescribing any direction of choice or decision. In fact, our version of nudging is more libertarian than the one originally introduced by Thaler and Sunstein (2008). They advocate "libertarianpaternalism" which uses nudging to influence people to make choices that are in their own best interest. We take a less moralizing stance by arguing that consumers should not be nudged into acting in their own best interest, we should simply use nudging to optimize the conditions for exercising consumer autonomy.

Autonomy and misleading marketing
We have developed the concept of autonomy to be robust in contexts where external factors exert influence on the individual to take a certain course of action. We will now briefly test our concept of consumer autonomy against Sher's (2011) framework for assessing misleading marketing to check if our external conditions safeguard against undue influence on the consumer's ability to make informed decisions. Sher distinguishes between two forms of misleading marketing: deception and manipulation. Deception is an intentional attempt to deceive consumers by making false claims, omitting material information, or misrepresenting facts. Manipulation occurs when marketers exploit vulnerabilities in consumers' decision-making processes such as excessively long or deliberate obfuscation of terms and conditions, targeting vulnerable consumers, or camouflaging promotional content to make it difficult for consumers to recognize adverts (e.g. native advertising, paid for product endorsements in consumer-created YouTube videos, advertorials that are indistinguishable from editorial content). Our concept of autonomy protects the consumer against both forms of misleading marketing. On the one hand, the requirement for access to material information protects against false claims and factual misrepresentations and thereby safeguards against deception. On the other, the requirement for proportionality protects against data overload, whereas epistemic integrity protects against obfuscation and hidden marketing strategies, thereby safeguarding against manipulation.

Brand as information proxy
Perhaps consumer autonomy is not that important after all. One could argue that information asymmetries, data overload, and increasingly technical information about consumer products mean that consumers cannot realistically process the information necessary for making rational informed choices and acting autonomously. A better solution would therefore be to accept that autonomy is unrealistic and therefore focus on brands as shorthand devices in information processing. It is a commonplace assumption in brand management that well-trusted brand names work as quality proxies such that consumers automatically transfer the trust in a brand to new products by the same brand (De Chernatony, 2006;Keller, 2008). This reduces the need to read terms and conditions and product information as the brand is trusted by default. For a company like Toyota, the consumer uses the parent brand as a shortcut for reliability. This works because the consumer has recourse. If the car is not reliable, Toyota's brand is damaged, which they want to avoid, because it is beneficial to continue having satisfied consumers. There is no need for Toyota to make available technical details on reliability under various driving conditions relative to specific driving styles, and so on, because the consumer simply takes the brand as a proxy for reliability. 1 Tempting as it is, this solution is also high risk. Research has demonstrated that brands consciously use the proxy function to exploit vulnerabilities in consumers' decision-making processes. For example, established brands of a high quality (e.g. healthy breakfast cereals) are extended with new product categories of a lower quality (e.g. unhealthy breakfast bars), but the consumer automatically transfers the quality perception of the original brand product to the new category extension exactly because the brand functions as a quality proxy (Anker et al., 2011). To avoid misleading marketing and protect the consumer, it is paramount that we find ways of facilitating and protecting classic consumer autonomy and informed decision-making by operationalizing the internal and external conditions of autonomy.

Meta-critique
We will conclude the article by addressing one important line of criticism. Some will argue that developing a concept of autonomy that is designed to ensure consumer autonomy is effectively promoting a neoliberal view of the sovereign consumer. Such a critique could build on Baudrillard's (1998) explorations of the consumer society and the argument that consumers in capitalistic societies live in a marketing-constructed world of signs and symbols (simulacra) that is designed to reproduce itself by constantly generating new desires for products and services. Consumers are a necessary constituent in capitalism and as such the idea of consumer autonomy is a misnomer: the truth is that consumers are being deceived to think that they are autonomous and marketing regulations is a means to legitimize the false belief in consumer sovereignty. While this seems like a harsh critique, we can easily reach the same conclusion from different premises. A key insight from Luhmann's (1995) theory of social systems is that any given social system will always aim to reproduce itself, which it does through a system of communications that constantly reaffirm its own purpose and legitimacy. Marketing is necessary to sustain the capitalistic system, and consumer autonomy is one way of locking people into the ecosystem because they erroneously get a sense of freedom and control. From a Foucauldian (1972,2002) perspective, any type of societal organization-market-or state-based-is a system of power. Marketing is one of the means by which one type of social organization-capitalism-exercises power. The idea of consumer autonomy is neither true nor false, but a social function in a specific system which is central for its appropriate functioning and classifying consumers as autonomous makes them no less pawn in a game; rather, it confirms their status as such.
We accept the critique as relevant: our concept of autonomy can certainly be used to legitimize a neoliberal capitalistic market economy. However, we do not find this to undermine the importance of our study. We would like to meet the critique by clarifying our underpinning assumptions and the nature and scope of our concept of autonomy. First, we assume that there is such a thing as individual autonomy and that genuine autonomy is valuable per se: autonomy seems to be something profoundly important to humans, regardless of which political system they are subject to (Dworkin, 1988). Second, we also assume that any type of social system will exercise some influence on the individual to behave and think in certain ways. Our concept of autonomy is applicable to social and political systems where personal autonomy is valued, whether as a normative ideal, a human right, or an operational function to reproduce a certain social order or system. We value autonomy for different reasons, but it seems that we almost always have a reason to protect it. Our contributions to a concept of autonomy that is robust in the face of external persuasion is of critical importance, because we more often than not find ourselves in contexts where we are under influence to act in a certain way, but want to retain our integrity to make up our own mind and decide for ourselves. Consumers or not . . . .

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.