COGNITIVE-BASED REGULATION: NEW CHALLENGES FOR REGULATORS?* by Fabiana Di Porto (University of Salento) and Nicoletta Rangone (Politecnico of Milan) 25 October 2013 Summary: 1. Introduction. 2. The need for a cognitive-based approach to the regulatory process. 3. Cognitive sciences in practice: enriching traditional regulatory strategies. 4. The emergence of new regulatory strategies: operational empowerment and nudging. 4.1 Operational empowerment options. 4.2 Nudge strategies. 5. Lesson drawing for regulators: how to choose among different regulatory options? 1. Introduction The economic paradigm of rational choice (if adequately informed) tends to frame reality in an interpretative scheme which does not necessarily correspond to reality itself. The cognitive sciences 1 have discredited this paradigm by stressing that people are driven by cognitive short-cuts, social norms and pressures. * This paper has been subject to blind peer-review. It will be included as a chapter in L. Arnaudo (eds), Cognitive law: A collection of Italian studies, Rome, Luiss University Press, forthcoming, 2014. The paper has been conceived and structured in common. However, paragraphs 1, 3, 4 and 4.1, were written by F. Di Porto. Paragraphs 2, 4.2 and 5 were written by N. Rangone. Comments are welcome. Suggested quotation: F. Di Porto and N. Rangone, Cognitive-based regulation: new challenges for regulators?, federalismi.it, 20/2013. 1 We should acknowledge that no generally accepted theory exists defining, classifying and establishing a clearcut distinction between cognitive sciences and behavioural studies. We use here “cognitive sciences” as a general label encompassing different disciplines which investigate human learning and decisions and their federalismi.it n. 20/2013 Indeed, the cognitive sciences have shown that the real people are all but perfectly informed and rational utility-maximisers; rather, they use mental shortcuts and heuristics, and are subject to recurring biases in decision-making 2 (such as, for instance, status quo bias, time inconsistency, loss aversion, or short term costs overvaluation against long term benefits). Such errors are due to intuitive rules which enable people to simplify choices (heuristics) used in the collection and processing of information for decision-making, such as availability, representativeness and anchoring 3. Morals too, as well as emotional contexts, can condition human decision-making 4. Dealing with these and similar cognitive errors is something regulators cannot ignore, provided that even the most sophisticatedly designed regulation may become ineffective or fail to reach its goals where all or part of the regulated group respond in an unpredicted way. In this perspective, the novelty and great contribution of these studies is an unprecedented attention to real people. By analysing real people’s decision-making mechanisms and reactions, the cognitive sciences have shown that people, firms and decision-makers, while not necessarily always being rational, are not always irrational either. As an obvious consequence, any simplification of reality by substituting one a priori for another must be avoided. In fact, even errors – due to heuristics – must be measured with reference to a given relevant market. In other words, the specific psychological reaction to a duty to do something, to an offer (of consumption), an option (of saving, investment etc.) must be analysed in the here and now. Therefore, the following may be an indication for rule-makers: for any specific relationship with the social context, such as neurosciences, social psychology and sociology. We will especially refer to contributions from both behavioural economists and neuroeconomists that apply knowledge on human behaviour and brain to economics. The main goal of this paper is to bridge such studies with a particular branch of public law, namely regulation. 2 We will use biases and cognitive errors in the conventional way, i.e. as predictable deviations from rationality (bounded rationality) only for the sake of simplicity. We are conscious that the cognitive sciences, as defined above, encompass a more multifaceted vision of human decision-making and behaviour. Here biases are not to be regarded as deviations or anomalies, as exceptions to rationality but rather they are just part of normal behaviour, as neuronally driven and conditioned by personal attitudes, culture, institutional environment, and the interactions among them. 3 This systematization is due to A. TVERSKY and D. KAHANEMAN (Judgement under Uncertainty: Heuristics and Biases, Science, Vol. 185, n. 4157, 1974, pp. 1124-1131; D. KAHANEMAN and A. TVERSKY, Prospect Theory: an Analysis of Decisions under Risk, Econometrica, Vol. 47(2), 1979, pp. 263 ss.). These studies, published in the Nineteen-Seventies, are considered the starting point of the behavioural economics which, thanks to H. SIMON’s teachings (A Behavioral Model of Rational Choice, Quarterly Journal of Economics, 1955, Vol. 69, n. 1, p. 99 ss.), uses surveys and laboratory data in order to match neoclassical economic theory with real people’s behaviour and to find reasons for eventual inconsistency. These studies have been developed by R. THALER (Mental Accounting and Consumer Choice, Marketing Science, Vol. 4:3, 1985, pp. 199-214) and then addressed to rule-makers by this Author and C.R. SUNSTAIN (Libertarian Paternalism, The American Economic Review, 2003, Vol. 93, n. 2, p. 175 ss.). 4 P. KOSLOSKI, Principles of Ethical Economy, Kluwer Academic Publishers, 2001; A.K. SEN, On Ethics and Economics, Blackwell Publishing Ltd. 1987; G. AKERLOF and R. SCHILLER, Animal Spirits. How Human Psychology Drives the Economy, and Why It Matters for Global Capitalism, Princeton University Press, Princeton-Oxford, 2009, p. 21. www.federalismi.it 2 targeted group in the relevant market, it is important to intervene on the basis of empirical evidence of end-users’ heuristics and cognitive errors. This puts a great challenge on regulators, and calls for a change in the regulatory process. Indeed, where cognitive errors have been detected, rule-makers need to decide if, compared to other strategies, intervening through regulation is the most suitable response to help individuals adopt deliberately conscious decisions. If this is the case, then regulators need to decide what type of regulation should be preferred. Our claim is that choosing among different regulatory tools depends on whether the detected cognitive errors may be overcome or not. In the first case, we argue that operational empowerment tools (aimed at educating people and increasing their cognitive abilities) might be best strategy; whereas in the second nudging (which in our narrow definition takes advantage, often in an undisclosed manner, of heuristics and errors while preserving regulatees’ choice) should be preferred. Again, there could not be a general or theoretical answer to the question of the avoidability of cognitive errors, which must be detected in the regulatory process. Establishing how the regulatory process should change in order to bring out and use evidence from cognitive sciences is the first question this paper is intended to answer (Para 2). It then looks at how well established regulatory strategies might be enriched by a cognitive-based approach (Para. 3). It further provides a closer look into two newly established regulatory strategies that put in practice and respond to the challenges of cognitive science: operational empowerment and nudge (Para. 4), showing their characteristics, weaknesses and strengths, having special regard to their attitude toward cognitive errors. It then concludes by formulating some proposals about how to choose among different regulatory options (Para. 5). 2. The need for a cognitive-based approach to the regulatory process Biases in judgment and individual perception of external stimuli 5offer crucial information to rule-makers about the reactions of end-users, and in so doing they enable the better formulation of rules and the provision of more adequate responses to the public interest they are intended to satisfy. 5 “Knowledge and personal knowledge are therefore the cornerstones of decision-making: this implies the individual ability to perceive external stimuli and to modify them through interpretation, representation and imagination. Learning can be considered as the engine that allows us to build up a representation of the situation and to carry out the steps that precede actions” (M. EGIDI and S. RIZZELLO, Cognitive Economics: Foundations and Historical Evolution, Working paper n. 4/2003, Università di Torino, p. 3). www.federalismi.it 3 Therefore, it could be useful to use this knowledge in order to draft regulatory options and to verify ex post their efficacy in the regulation life-cycle. However, this approach does not justify any automatic connection between evidence of cognitive errors and a given regulatory tool (as highlighted above). On the contrary, once the presence of cognitive errors has been established, it should be measured with reference to existing (one or several) groups of regulatees in the relevant markets. Indeed, end-users’ reactions to a regulatory strategy are strongly influenced by the existing social norms and by the (quality of) regulation; therefore – as an obvious consequence – the socio-cultural context and the characteristics of the regulatory framework become crucial elements in this analysis. Using evidence from the cognitive sciences in rule-making requires an evidence-based approach and a great change in the regulatory process. First of all, surveys and consultations should be designed to pinpoint information about cognitive errors – ideally carried out with the help of cognitive science experts. Indeed, it is important to consult in a way which could bring about potential cognitive errors. For instance, in order to evaluate the role of biases (such as inertia and procrastination) in the low switching rate in energy retail markets 6, those consulted could be consumers (and not only firms) who should be asked about their knowledge of consumption and of costs related to their consumption; if they have ever switched providers in the same or other markets; whether, if they did, they have saved money, and then checked the continuing benefits of this choice; if they are aware of alternative offers and how they heard of them 7. Moreover, consultations should be carried out carefully so as to tease out the extent to which low switching, in the example, has been influenced by other factors, such as the perception of the quality of regulation, or the perception of adequate protection of consumer interests from the behaviour of firms. Academic studies and professional documents on emotional reactions to a given issue and on social norms which might shape individual decision-making, might also provide valuable insight whose relevance might need to be verified in the area targeted by the regulatory intervention. Secondly, the regulatory impact assessment-RIA should also allow the use in the rule-making of knowledge about cognitive errors and decisional processes. Indeed, RIA (while used 6 Office of Gas and Electricity Markets-OFGEM, What can behavioural economics say about GB energy consumers?, 21 March, 2011. 7 See e.g. C. WADDAMS, Consumer Choice and Rationality, Presentation at the international conference on Consumer Protection in Europe, organised by Acquirente Unico and EUI, Fiesole, 16 February 2012. www.federalismi.it 4 according to the proportionality principle) 8 is meant to help regulators to choose among regulatory options (including those drafted using information about cognitive errors) and to assess ex post their real impact. However, in order to fulfil this role, RIA should evolve both in terms of information gathering and in terms of evaluating the impact of rules. Indeed, when based on the cost-benefit analysis, the RIA tends to assume end-users as rational self-interest maximisers. Otherwise, an RIA enriched by a risk assessment might evaluate impacts of different regulatory options by taking into account the end-users’ potential biases as well as considering heuristics as a risk to be assessed in terms of outcomes and probabilities. The result of this analysis could help rule-makers to decide whether or not to deal with cognitive errors through regulation 9, and to identify a minimum threshold justifying regulation, which of course does not correspond to the simple presence of cognitive error risks, but to those risks that regulators have considered unacceptable 10 . Therefore, RIA helps rule-makers to detect if a cognitive-based regulation is really needed, bringing out factors which influence individuals’ behaviour, whether the errors are cognitive in nature or due to some other factor 11 . Moreover, in RIA the potential effectiveness of regulatory options are compared to the 8 Only major rules with a potential significant impact might be concerned: otherwise, if all the flow and stock of regulation is assessed, the analysis risks being superficial and its costs unjustified. However, the various criteria identified are far from solving the problem, being either too or to general, or requiring asking for a specific assessment in order to decide whether an impact assessment is needed. In the USA, the use of RIA at federal level is only for rules that may have “an annual effect on the economy of $100 million or more or adversely affect in a material way the economy, a sector of the economy, productivity, competition, jobs, the environment, public health or safety, or State, local, or tribal governments or communities” (Executive Order 12866, Regulatory Planning and Review, 1993). The UK suggests the use competition impact assessment for proposals which have a potential significant impact on competition (Completing Competition Assessment in Impact Assessment, 2007). At the same time, the depth of the assessment (as well as the consultation process) must be proportional to the issues at stake and the resources available. At European level, although general exceptions have not been made, the concept of a “proportionate level of analysis” for any RIA has been employed. It relates to the appropriate level of detail of analysis which is necessary for the different steps of RIA, and is connected to potential impact, political significance and the steps in the process of policy development (European Commission, Impact Assessment Guidelines, SEC(2009) 92, p. 12). 9 H. PILADES and C. R. SUNSTEIN, Reinventing the Regulatory State, in University of Chicago Law Review, Vol. 62, n. 1, 1995, p. 43. 10 Although it could prove difficult to identify the relevant probabilities and specify the contribution of a given regulation to reducing the risk of cognitive errors (C.R. SUNSTEIN, The real world of cost-benefit: thirty-six questions (and almost as many answers), 2013, forthcoming. Moreover, as scholars have underlined, the risk analysis is not a neutral technique and many steps may involve regulatory or political discretionality (R. BALDWIN, M. CAVE and M. LODGE, Understanding Regulation. Theory, Strategy, and Practice, Oxford, Oxford University Press, 2012, p. 283). 11 Indeed, the challenge is to “distinguish truly irrational behaviour from rationally made and therefore efficient mistakes” (J.D. WRIGHT and D.H. GINSBURG, Behavioral Law and Economics: its Origins, Fatal Flaws, and implications for Liberty, Northwestern University Law Review, 2012, Vol. 106, n. 3, p. 1048). For instance, information about a low switching rate in energy retail markets which characterises two countries (such as, e.g., the United Kingdom and Italy) could be a clue of a cognitive error (such as inertia and procrastination), or of a very rational behaviour if, according to a specific regulatory framework, it is more convenient (or at least less risky) to remain with the incumbent rather than switching. Therefore, a cognitive based regulation which is intended to deal with inertia and procrastination might be useful in the first example and a policy error in the second. www.federalismi.it 5 doing nothing option. Therefore, this step might be very useful in order to avoid any overestimation of social costs of cognitive errors (which, according to the critics of libertarian paternalism, could end up justifying over-regulation), as well as any under-estimation of social costs of regulations aimed at reducing such errors (a cognitive-based regulation) in terms, for instance, of individual liberty limitation 12. Such an incorrect estimation might be difficult in RIA which, by assessing the potential impacts of feasible options and comparing them with the doing nothing option, is intended to identify the option whose advantages outweigh its disadvantages (the so-called preferred regulatory option) 13 . Moreover RIA, increases end-users’ participation and rule-making transparency, being based on consultations and generally being characterized by the publication of the technical analysis performed (at least in a summary). As such, it might help in increasing the end-users’ awareness of cognitive-based mechanisms (which, being designed to encourage people, are less transparent than those which impose or forbid behaviours). Thirdly, evidence from cognitive science leads to an enrichment of the traditional regulatory options (such as those based on the command and control approach, or incentive based regulation) and to include new regulatory options (operational empowerment and nudging). 3. Cognitive sciences in practice: enriching traditional regulatory strategies The lack of knowledge about real people, their recurrent errors and biases may defeat even the most sophisticated and traditional regulatory tool, such as Command and Control (C&C), disclosure or incentive regulation. In this domain, evidence of cognitive errors may add to the understanding of one’s behaviour and strengthen the regulator’s cognitive apparatus and increase regulatory effectiveness 14 . In addition, cognitive sciences may also help re- 12 These critics of behavioural law and economics literature are mentioned by J.D. WRIGHT and D.H. GINSBURG, Behavioural Law and Economics: its Origins, Fatal Flaws, and implications for Liberty, cited above, note 11, p. 1041. 13 A comparison between a new cognitive-based intervention with what would have happened if nothing were changed might always be engaged in rule-making even when, according to the proportionality principle, RIA is not used. In both regulatory processes with and without RIA, the comparison might be more robust by using a randomly assigned control group (L. HYNES, O. SERVICE, B. GOLDACRE and D. TORGERSON, Test, Learn, Adapt: Developing Public Policies with Randomised Control Trials, Cabinet Office Behavioural Insights Team, 2012). 14 On the increasing interest toward a deeper knowledge about decision-making processes and individuals’ errors: see e.g. the Commission’s report edited by N. CHATER, R. INDERST, S. HUCK, Consumer DecisionMaking in Retail Investment Services: A Behavioural Economics Perspective, 2010 (http://ec.europa.eu/consumers/strategy/docs/final_report_en.pdf). More recently the Commission has adopted a number of initiatives to include behavioural insights into its policy-making: see e.g. the “European Consumer Agenda Boosting confidence and growth” COM(2012) 225 fin., of 22.5.2012 (ec.europa.eu/consumers/strategy/index_en.htm#agenda), and the Staff Working Document on “Knowledgeenhancing aspects of consumer empowerment, 2012 – 2014”, SWD(2012) 235 fin., of 19.7.2012 (ec.europa.eu/consumers/strategy/docs/swd_document_2012_en.pdf). In 2013 the Commission has www.federalismi.it 6 interpreting traditional paternalism, thus somehow enriching its rationale, something that many often confuse with a cognitive approach to rule-making. (a) Rethinking command and control. On the first ground, a C&C strategy with paternalistic objectives, prohibiting or imposing limits to end-users, if designed in a cognitive-based manner, can be used to avoid cognitive errors. For instance, in the case of corporate governance rules and information obligations can be imposed on listed companies to counterbalance the overconfidence of company managers. Here the rule-maker takes note of regulatees’ cognitive errors, but does not accept the risk that they occur, and adopts regulation intended to avoid or limit such errors. This approach is potentially effective and relatively inexpensive for rule-makers, as general standards are usually cheap to design and easy to implement. However, beside paternalism 15, it could lead to over-regulation anytime it bans or puts limitations onto those not affected by cognitive errors. These highly intrusive limitations would hardly be regarded as cognitive-based in the sense clarified above (para. 2): firstly, many of them were introduced several decades ago, i.e. far before the first studies on cognitive sciences and behavioural economics were known to the wider public, and eventually irrespective of the proof of any cognitive errors; secondly, their implementation was most probably not preceded by systematic gathering of data or empirical evidence of any recurrent cognitive errors; nor were they introduced following experimentation 16 . Following a cognitive-based approach could help reducing the over- regulation shortcoming of C&C: if evidence is gathered of biases affecting only a slight minority of the population in a given relevant market, then a complete ban or a set of duties and standards would be excessively restrictive of the population’s autonomy. On the second ground, if a C&C strategy is used irrespective of the proof of any cognitive errors, it can still be interpreted in a cognitive perspective. So, for instance, speed limits or the commissioned four studies (on cross-border healthcare, online gambling, hidden fees for card payments and making greener choices for online purchases) to gather behavioural evidence to be included in its future policies. Results of these studies were presented and discussed at the Conference on “Applying behavioural insights to policy-making: results, promises and limitations”, Brussels, 30.9.2013. 15 J. KLEINIG, Paternalism, Totowa (NJ), Rowman&Allanheld, 1983, esp. at pp. 18 ff.; G. DWORKIN, Paternalism, The Monist, 56, 1972, pp. 64–84 (see also Paternalims, in the Stanford Encyclopaedia of Philosophy: http://plato.stanford.edu/entries/paternalism/). According to A. OGUS, Regulation, Oxford, OUP, 1994, pp. 51-2, many paternalistic regulations could be better justified in terms of externalities. 16 Similar arguments can be made with regards to statements by Director General Testori Goggi (Behavioural insights in the Commission, appeared on European Voice, of 14 June 2012) claiming that the European Commission has been using a behavioural approach since 2008. Examples cited include the 2008 EC limitation to the use of pre-checked boxes in consumer contracts and the 2010 “user-friendly” on-screen ballot box remedy in the Microsoft I antitrust case. As remarked, this is not a truly cognitive-based approach to regulation, provided that the Commission did not engage in “serious, systematic, yet experimental, use of behavioural insights into EU policymaking”: A. Alemanno in his blog (www.albertoalemanno.eu/articles/4381082). www.federalismi.it 7 obligation to wear seat belts can be understood ex post as tools intended to counterbalance motorists’ overconfidence. Similarly, smoking bans benefit (also) those who are being coerced: indeed, they might help people who are willing to quit smoking, but are unable to do so, to take the sensible course of action 17. Ex post cognitive explanations can also be useful, for instance to judges, to justifying and interpreting a legal measure in adjudication. (b) (follows) Disclosure regulation. A second traditional regulatory strategy is disclosure18: information-related obligations and the (ex post) control of false or misleading information are copious in most legal systems 19 and are aimed at attenuating market failures, such as information asymmetries or externalities as democratic participation 21 20 , as well as fostering other social objectives, such and avoiding government failure 22. Information disclosure obligations are fundamentally neutral with regard to individual preferences and freedom of choice, as are the norms on unfair commercial practices; indeed they are choice preserving. The main limitation is that all these disclosure tools have in mind a notional, typical consumer 23; they assume that individuals may take advantage of disclosure to make rational decisions. 17 R.H. THALER and S. BENARTZI, Save More Tomorrow: Using Behavioral Economics to Increase Employee Saving, Journal of Political Economy, Vol. 112, S1, February 2004, pp. S164-S187; see also S. BENARTZI and R. LEWIN, Save More Tomorrow: Practical Behavioural Finance Solutions to Improve 401(k) Plans, New York, Penguin, 2012. 18 On disclosure regulation see: R. BALDWIN, M. CAVE, M. LODGE, Understanding regulation, 2nd ed., Oxford, OUP, 2012, espec. p. 117; A. OGUS, Regulation, cited above, note 15, pp. 121 ff.; S. BREYER, Regulation and its reform, Cambridge (MA), Harvard University Press, 1982, p. 161 ff.; F. DI PORTO, L’informazione come ‘oggetto’ e come ‘strumento’ di regolazione (il caso dei mercati energetici al dettaglio), in Riv. trim. dir. pubbl., 4/2011, pp. 975-2010. 19 See EC Commission’s Report on “Consumer Policy (July 2010 - December 2011)” accompanying the “European Consumer Agenda - Boosting confidence and growth”, SWD(2012) 132 final, of 22.5.2012, providing details on EU information regulation in various domains (food, tobacco, educational programmes, etc.), p. 10 ff. See also the Consumer Law Compendium 2007, edited for the European Commission by H. SCHULTE-NÖLKE in co-operation with C. Twigg-Flesner and M. Ebers, 2007 (www.eu-consumerlaw.org/index.html) at pp. 412-415 for implementation of EC directives encompassing some information-related obligations in the 27 member states. 20 A. SCHWARTZ and L. WILDE, Intervening in markets on the basis of imperfect information: a legal and economic analysis, 127 Univ. Penn. L. R., 1979, p. 630. 21 For consideration of public systems of information provision, see: K. YEUNG, Government by Publicity Management: Sunlight or Spin?, Public Law, 2005, pp. 360-383; ID., Regulating Government Communications, 65 Cambridge L. J, 2006, pp. 53-91; C. HOOD and H. MARGETTS, The tools of government in the digital age, London, Palgrave Macmillan, 2007. 22 E. BALLEISEN and D. MOSS, Introduction, in Balleisen and Moss (eds.) Government and Markets: Toward a New Theory of Regulation, New York: Cambridge University Press, 2009, p. 1 ff. 23 See cons. 18 of Directive 2005/29/CE. www.federalismi.it 8 However, they do not consider that information giving per se does not imply real comprehension 24 . Incalculable empirical studies exist showing that consumers do not make full use of the information provided; that information given in a dense format may discourage people from acting or may confuse them; that too much information may distract people from relevant aspects to making sensible and reflected decisions; that information overload increases transaction costs, and so on 25. For disclosure regulation to be cognitive-based scrutiny on motivational determinants of individuals’ choices and errors in the targeted group should be conducted prior to its introduction and considered in ex post assessment. Recent developments at EC level testify of an evolution of disclosure regulation to take into consideration cognitive shortcomings and build a more comprehensive strategy named empowerment, which will be analysed in more detail below (para. 4.1). (c) (follows) Legal empowerment. A third strategy is legal empowerment: legal standards are enforced not only by individuals but also by public institutions acting ex officio. Old tools (e.g. the power to challenge unfair clauses in standardised contracts 26 ) are put at disposal (also) of public bodies, transforming them into public procurators for individuals’ well being and individual rights’ enforcement. Unlike in traditional contexts where the notion of “legal empowerment” was first introduced 27, here it is to be understood mostly as “public tutoring”, where the public powers are in charge of supporting the exercise of private rights, to attain stronger compliance, no matter whether consumers are effectively in need of protection or weather they are actually affected by some cognitive error. Legal empowerment assumes consumers’ weakness, and avoids the risk of low enforcement of their rights. In that, it 24 Such an acquisition is commonly dated back to the pioneering works of H.A SIMON, A Behavioral Model of Rational Choice, 69 Q. J. Econ., 99, 1955, pp. 99‒ 100, and ID., Rational Decision Making in Business Organizations, 69 Am. Econ. Rev., 1979, pp. 493 ff. 25 Commission’s Staff Working Document on “Knowledge-enhancing aspects of consumer empowerment, 2012 – 2014”, 2012, cited above, note 14, at p. 8: “The traditional economic approach assumes that more consumer information leads to better consumer empowerment. Each new piece of information improves consumer welfare as it enables consumers to make better informed choices. But today’s society is characterised by a densemessage environment, and our cognitive capacities are limited … In practice, behavioural researchers have found that … there is a limit to how much information a consumer can or wants to absorb at the various steps of the purchasing process, and more information is not always better for consumers. There is a clear trade-off between information completeness and simplicity, and also a potential crowding-out effect”. 26 See Council Directive 93/13/EEC of 5 April 1993 on Unfair terms in consumer contracts (OJ L 95 of 21 April 1993, 29-34), as amended by Directive 2011/83/EU of the European Parliament and of the Council of 25 October 2011 on consumer rights. 27 “The purpose of legal empowerment is to enable the people - in particular, the disadvantaged - to assert their rights vis-à-vis the State. Here the State and the administration play a double role: they recognize rights and also provide the means to enforce them” (S. CASSESE, The rise of the administrative state in Europe, Rivista Trimestrale Diritto Pubblico, 2010, pp. 981-1008). www.federalismi.it 9 strengthens the deterrent effect of private enforcement. However, like C&C, it is still stuck to the notional, unrealistic concept of “average consumer”. It supports the idea that the best way to protect individuals is by more public intervention and more ex officio powers to public bodies, such as independent regulatory authorities 28 . No doubt the latter have a deeper knowledge of the commercial strategies adopted in their realms, however expertise does not necessarily cover knowledge about the real people and their cognitive errors. For instance, following a cognitive-based approach, the regulator might consider the “rational apathy” affecting consumers as the main cause for limited use of class actions 29 . In such circumstances, a possible solution might be to substitute the existing opt-in default rule (only people registered in the class will be affected by the final decision) with an “opt-out” one (the final decision or decree is binding to the all in the class unless opted out) 30, instead of (or in addition to) “public tutoring” (i.e. complementing private action with a public one or of attributing more individual rights to be enforced 31). (d) (follows) Incentive regulation. The fourth scheme to be considered is incentive regulation. Under this strategy economic rewards are associated to virtuous behaviours following an incentive scheme established in advance by the regulator. Incentive regulation is a widely known and extensively experimented tool; what interests here is that modifying the incentive structure of regulatees is considering only one motivation for action: the financial 28 In the UK the Office of Fair Trading (OFT, which will be replaced by the Competition Market Authority or CMA, starting from 2014, as established by the Enterprise and Regulatory Reform Act 2013, of 25 April 2013) can take legal action to prevent the use of unfair terms in Business-to-Consumers contracts. The aforementioned EC Consumer Law Compendium of 2007 reports that most member States do have an “extensive system of public control of contract terms” and administrative bodies almost everywhere (except for Spain or Italy) do “have standing to apply for injunctions in court” (p. 424). Enforcement of Directive 2005/29/EC “is mainly carried out by public authorities such as consumer ombudsmen (e.g. Denmark, Sweden and Finland), consumer or competition authorities (e.g. Italy, Ireland, the Netherlands, Romania and the UK) and dedicated departments of ministries (e.g. Portugal and Belgium). Other Member States run a private enforcement scheme led by competitors (e.g. Austria and Germany)”. A combination of the two is however present in most of them: see EC Commission’s “First Report on the application of Directive 2005/29/EC (‘Unfair Commercial Practices Directive’)”, COM(2013) 139 final, of 14.3.2013 (http://ec.europa.eu/justice/consumermarketing/files/ucpd_report_en.pdf) at pp. 26-27). 29 Rational apathy is understandable as the reasonable assumption that my vote will have little or no effect on the results of general elections, causing disaffection and disengagement. In the case of class action, rational apathy – together with the little economic advantage a consumer might earn from a successful class action – if combined with an opt-in scheme (like the one in place in Italy) may lead to a small number of participants, thus decreasing its critical mass and following deterrent effect. For further discussion see G. AFFERNI, Class action e danno antitrust: il caso traghetti, in Consumatori, Diritti e Mercato, 2, 2012, pp. 118-126. 30 As suggested by X. GABAIX, A. LANDIER, D. THESMAR, P. MONGIN and J. TIROLE, La protection du consommateur: rationalité limitée et régulation, Conseil d’Analyse Economique, Paris, La documentation française, 2012. 31 In Italy, for instance, the law governing class actions has been amended in 2012 to extend the right to initiate a collective suit to individuals also (and not just consumers’ associations). This has been done based on the scarce application of the law, irrespective of any analysis of the psycho-cognitive determinants of such limited success of the tool (see new Art. 140-bis of the Consumer code). www.federalismi.it 10 one. But people may be indifferent to economic rewards and thus react in a way regulators may not predict. Also, biases in assessing the value of present money against that of future money (endowment effect) may hinder the efficacy of incentive-based regulation. Such schemes assume regulatees’ economic rationality, i.e. the very core concept cognitive sciences have been de-constructing. Certainly incentive regulation as we know it preserves autonomy and is quite easy to enforce. However, insights from cognitive errors of the regulatees might help improving the incentive scheme and increase its overall efficacy. It should be noted, though, that by combining the economic with other non-economic rationales for action (discretion), regulators may let their vision of the world prevailing on that of other stakeholders, besides the risk of capture that arises anytime regulators need establishing economic incentives that benefit some groups only. 4. The emergence of new regulatory strategies: operational empowerment and nudging In the past decades, some countries have been paying an increasing attention to the need to cope with cognitive errors, calling for the use of behavioural insights in order to draw up new regulatory strategies 32 . This move has been speeded up by the decline of neoliberalism, that until the Nineties of the past Century had favoured the homogeneity, say, convergence toward a set of well established regulatory strategies 33. The bursting onto the scene of cognitive sciences and behavioural insights has thus contributed to the emergence of new regulatory tools that are inclined to consider cognitive 32 In the United Kingdom, the Behavioural Insights Team of the Cabinet Office was set up in July 2010 and in 2011 implemented a consumer empowerment strategy called “Better choices: better deals. Consumers powering choices”, a programme that brought together the industry, the government and consumers to help the latter access their own user data so to make better deals (www.gov.uk/government/uploads/system/ uploads/attachment_data/file/31834/11-749-better-choices-better-deals-consumers-powering-growth.pdf). In 2011 the energy regulator, OFGEM, published a report on “What can behavioural economics say about GB energy consumers?” (www.ofgem.gov.uk/Markets/RetMkts/rmr/Documents1/Behavioural_Economics_ GBenergy.pdf). In the United States, the 2009 OIRA Report to Congress on Benefit and Cost of Federal Regulation mentions the need for a behaviourally informed approach. In November 2010 the European Commission published the already mentioned mighty report on “Consumer Decision-Making in Retail Investment Services: A Behavioural Economics Perspective”, showing that private investors’ cognitive errors may be lessened by the reduction and standardisation of information; in 2011, together with the Joint Research Centre (JRC) it developed the Consumer Empowerment Index (CEI) describing the perceptions and behaviours of more than 56.000 consumers of twenty-nine European countries; it then published two more documents applying behavioural insights to develop new regulatory strategies to protect consumers: one is the already mentioned Report on “Consumer Policy” SWD(2012) 132 final, of 22.5.2012, the other is the also mentioned Commission’s Staff Working Document on “Knowledge-enhancing aspects of consumer empowerment, 2012 – 2014”, SWD(2012) 235 fin., of 19.7.2012. In France, in 2009 the Prime Minister-led Centre for Strategic Analysis initiated the “Neuroscience and public policy program”, an initiative dedicated to the use of brain and behavioural research in the formation of public policy; in 2012, the Conseil d’Analyse Economique, an independent advisory body reporting to the French Prime Minister, published the above mentioned report on consumer protection. 33 J.D. WRIGHT and D.H. GINSBURG, Behavioral Law and Economics, cited above, note 11. www.federalismi.it 11 errors (they are therefore cognitive-based). This rather recent way of intervening includes means that are aimed at: (i) preventing or help individuals avoiding cognitive errors and more generally at improving their ability to take appropriate decisions; (ii) preventing or using their cognitive errors to draw up soft incentives to make them behave in a way rule-makers consider to be better for them: these techniques will be defined, respectively, as “operational empowerment” (Para. 4.1), and “nudges” (Para. 4.2). The main difference between the two, as we will make clear below, resides respectively in the very avoidance of cognitive errors and the resulting willingness of regulation to interfere in one’s set of emotional preferences. Strengths and weakness of both operational empowerment and nudge strategies are analysed in the following paragraphs in order to frame some concluding indications which might help decision makers in rule-making. 4.1 Operational empowerment options The accrued awareness that consumers should be made aware of new possibilities offered by liberalised markets pushes the regulators to experiment new strategies to empower and to make them more active. Empowerment is also used in the healthcare sector 35 34 them to help patients take self-enhancing choices and stop being passive recipients of diagnoses and prognoses 36. Operational empowerment builds on these strategies to which it adds an accrued relevance given to information about individuals’ cognitive errors 37. Operational empowerment rests on the idea that individuals depart from rational choices because they are not aware, informed, or 34 Empowerment is a multi-dimension and multi-level notion, used, with different meanings and applications in political science, psychology, sociology, management, and marketing (for further discussion see the Joint Research Center’s Report edited by M. NARDO, M. LOI, R. ROSATI and A. MANCA, The consumer empowerment index. A measure of skills, awareness and engagement of European consumers, 2011 (http://ec.europa.eu/consumers/consumer_empowerment/docs/JRC_report_consumer_empowerment_en.pdf). In 2011 the Commission has no doubts about goals and tools of empowerment. Empowerment aims at welfare maximising and enhancing competition. As per the tools, it depends on good cognitive skills, knowledge of consumer rights, information, well functioning organisations and public authorities, accessible means of redress: see Commission staff working paper, Consumer empowerment in the EU, SEC (2011) 469 fin., of 7 April 2011. 35 Psychology and healthcare studies, probably the first disciplines to employ the term, conceive empowerment as self-determination deriving from information access and knowledge to be able to make informed choices; see J. GELLER, J. BROWN, W. FISHER, A. GRUDZINSKAS and T. MANNING., A National Survey of "Consumer Empowerment" at the State Level, Psychiatric Serv., 49, 1998, pp. 498-503. 36 R.M. ANDERSON and M.M. FUNNELL, Patient empowerment: Myths and misconceptions, in 79 Pat. Educ. and Couns., 2010, pp. 277–282. 37 The Commission itself recognises the limits of not knowing consumers in their daily behaviours, away from textbook models, and acknowledges that detailed portraits of European consumers is prerequisite for “helping EU and national policy-makers design smarter regulation” (Commission staff working paper, Consumer empowerment in the EU, cited at note 34). www.federalismi.it 12 educated enough to act reasonably. Biases and cognitive errors leading to misjudgements can be overcome and regulation be justified to provide more adequate and empirically-based information and education to reach awareness and, possibly, reasonable action. The theoretical contend implied in operational empowerment is therefore still attached to the “enlightened” accounts of rational choice. Indeed, the latter does accept violations of or deviations from conventional rationality, assuming that they derive from judgement errors that decision-makers can learn to overcome and correct, by giving them the right opportunities, information and data 38. In this perspective, operational empowerment options principally make use of institutions, social norms and soft law to instruct and educate individuals to act in their best interest, while advancing some public interest 39. It gives value to self-determined virtuous behaviours and is based on the idea that cognitive errors and biases may be avoided, corrected or overcome by strengthening people’s reflexive capabilities 40. Regulatory operational empowerment uses rules to tackle cognitive errors. It does so by framing information in an over-simplified way, or by simplifying individuals’ choice itself, or by using education so to prevent errors or help individuals overcoming their biases and emotional responses. Regulatory operational empowerment tools can thus be understood as truly de-biasing techniques 41 . Indeed, they are aimed at overcoming errors while preserving autonomy and freedom of choice, and leaving individual preferences largely untouched. (a) Information simplification. The easiest and most popular way to facilitate people’s choices is through the simplification of information to be given to individuals. Empirical studies have shown how reducing and standardizing information on financial products energy bills 43 42 and have a greater effect on the choices of investors and consumers than an 38 A. TVERSKY and D. KAHNEMAN, The Framing of Decisions and the Psychology of Choice, Science, Vol. 211, n. 4481, 1981, pp. 453-458, at p. 458. 39 In this operational empowerment differs a lot from other strategies such as C&C, as in most cases it does not include hard law, neither does it necessarily have a legal basis. 40 For a recent position questioning the long-established theory of dual system (creative/reflexive) in human brain, see S.M. KOSSLYN and G.W. MILLER, Top brain, bottom brain: Surprising Insights into How You Think, New York: Simon & Schuster, 2013. 41 Recalling C. JOLLS and C.R. SUNSTEIN, De-biasing through law, Journal of Legal Studies, Vol. 35, 2006, pp. 199-241. But diverging from them, we consider de-biasing techniques only those that are really intended to overcome, not to elude, a bias (a function that is mostly typical of nudge). 42 See the European Commission’s report, Consumer Decision-Making in Retail Investment Services, cited above, note 14. 43 For details on transparency obligations faced by energy providers established at EU level see the Working Group Report on Transparency in EU Retail Energy Markets, November 2012 (ec.europa.eu/energy/gas_electricity/doc/forum_citizen_energy/2012111314_citizen_forum_meeting_working_g roup_report.pdf). www.federalismi.it 13 increased amount of information (typical of information disclosure obligations), as attention is a scarce resource 44. To overcome consumers’ inertia and attain, for instance, higher switching rates, a simplification strategy could be used, e.g. compelling utilities to inform consumers about how much they have been spending overtime (not just in the current month) or about how much they could save if some suggested behaviours (e.g. using energy saving light bulbs, installing timer for aircon, etc.) were adopted 45. A variation is the simplification of information asked of consumers (make it easy) 46 ; for instance, the switching from one supplier to another could be made simpler by using forms which have been pre-filled by the traditional supplier with non-sensitive information about the consumer (such as name and address, telephone number, e-mail, etc.). This strategy might prove effective if it results that the targeted population is affected by status quo bias. (b) Framing: making information salient. Framing information so to make it not only easy to understand but salient is also essential: empirical findings in the context of household appliances’ labelling 47 show that comparative information (like scales) is more motivating, better understood and more effective in easing choice about energy efficient products, than is information presented generically or in technical format 48 . Thus, facing regulatees affected 44 Since Herbert Simon’s work, cognitive scientists and “behaviouralists” have discredited the idea that by providing more information (i.e. by enacting more duties to disclose information) individuals can act more rationally: to the contrary, “a wealth of information creates a poverty of attention and a need to allocate that attention efficiently among the overabundance of information sources that might consume it”: H.A. SIMON, Designing Organizations for an Information-Rich World, in M. Greenberger (ed), Computers, Communication, and the Public Interest, Baltimore, The Johns Hopkins Press, 1971, pp. 38-52, at 40-41 (http://zeus.zeit.de/2007/39/simon.pdf). Indeed, information overload can lead to cognitive and decision errors, such as overconfidence. For application in the financial sector, see E. MARCHISIO, A behavioural analysis of shareholders’ meetings, forthcoming in Journal of Corporate Law Studies, 2014, (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2224589), at p. 14. 45 In the UK, since 2008, Ofgem has introduced the Energy Supply Probe, a regulation requiring energy suppliers to send consumers an annual statement setting out how much they actually spent (in GBP) and consumed (in KWh) in the last year, and how much they would spend in the forthcoming year if their usage and prices remained unchanged. 46 C.R. SUNSTEIN, Humanizing Cost-Benefit Analysis, Admin. L. Rev. Conference, February 17, 2010, pp. 7-8. “Lessons from behavioural (…): make it as straightforward as possible for people to pay tax or debts, for example by pre-populating a form with information already held”. “One of the central findings from behavioural science is that if you want somebody to do something you should start by making it easy for them to do it. If you want to get people to recycle their household waste, provide them with recycling bins and arrange regular collections. If you want people to pay their tax or fines on time, make it easy for them to do so” (Cabinet Office Behavioural Insights Team, Applying behavioural insights to reduce fraud, error and debt, 6 February 2012; available at https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/60539/BIT_Fraud ErrorDebt_accessible.pdf, pp. 4 and 8). 47 Report “Research on EU product labelling options and consumer understanding”, delivered by Ipsos MORI, London Economics and AEA for the European Commission, October 2012 (available at http://ec.europa.eu/energy/efficiency/studies/doc/2012-12-research-eu-product-label-options.pdf), at p. 68. 48 A similar contend is frequently made with regard to investors’ ability to process information in the financial market, where information overload can arise “especially when (information is) drafted without providing some ranking”, see E. MARCHISIO, cited above note 45, at p. 14. www.federalismi.it 14 by e.g. loss aversion, regulators might disclose themselves or ask private firms to disclose information accordingly. In California, manufacturers must place on all new automobiles window stickers that provide a numerical rating of greenhouse gas emissions on a scale of 1 to 10 (where 5 signifies that the effect on global warming is average compared to other cars). Introduced at federal level in 2013, this informational intervention is enriched by new data on fuel saving or increases in fuel costs over five years compared to the average new vehicle 49. (c) Targeted education. In the area of health care, information alone might be insufficient to change one’s habits, especially if the patient is affected e.g. by unrealistic optimism. In order to reduce unrealistic optimism affecting people when assessing adverse outcomes associated to risky activities (such as smoking), targeted education might prove particularly effective. This finding explains the shift from an “informed consent” strategy, characterised by mere information disclosure about disease and foreseen treatment 50, to the “patient empowerment” one 51. By the latter, regulators aim at giving patients the means to make self-selected changes in relation to their health, rather than simply improve their compliance to treatments 52 . More examples can be found in the financial sector, where regulators are converging toward global educational strategies to increase what is called investors’ financial “literacy” 53 . Education, especially if targeted (i.e. addressed through regulation to those that suffer from some biases, such as overconfidence), increases awareness and may be used to contrast biases. However, 49 See Environmental Protection Agency (EPA), National Highway Traffic Safety Administration (NHTSA), DOT, Revisions and Additions to Motor Vehicle Fuel Economy Label, Federal Register, Vol. 76, n. 129, 6 July 2011. 50 D. DOUMONT and I. AUJOULAT, L’empowerment et l’éducation du patient, Bruxelles, UCL, RESO, 2002, extrait available at: www.uclouvain.be/cps/ucl/doc/reso/documents/dos18.pdf. 51 I. AUJOULAT, W. d’HOORE and A. DECCACHE, Patient empowerment in theory and practice: Polysemy or cacophony?, in Patient Educ. Couns., 66, 2007, pp. 13–20. 52 I. HOLMSTROM and M. ROING, The relation between patient-centeredness and patient empowerment: A discussion on concepts, Patient Educ. Couns., 79, 2010, pp. 167–172. See also C. FESTE and R.M. ANDERSON, Empowerment: from philosophy to practice, Patient Educ. Couns., 26, 1995, pp. 139–144. 53 See OECD, The OECD/INFE High-level Principles on National Strategies for Financial Education, 2012 (http://www.oecd.org/daf/fin/financial-education/OECD_INFE_High_Level_Principles_National_Strategies_ Financial_Education_APEC.pdf). Financial literacy is defined as “a combination of financial awareness, knowledge, skills, attitude and behaviours necessary to make sound financial decisions and ultimately achieve individual financial wellbeing” A. GRIFONI and F.-A. MESSY, Current Status of National Strategies for Financial Education: A Comparative Analysis and Relevant Practices, OECD Working Papers on Finance, Insurance and Private Pensions, No. 16, OECD Publishing, 2012. In the US, following the recent crisis, a Consumer Financial Protection Bureau (CFPB) was created as a federal agency to “educate, enforce, study” (see http://www.consumerfinance.gov/the-bureau/creatingthebureau/). www.federalismi.it 15 no agreement exists as far as the efficacy of education is concerned: eventually, some contend, it could worsen some biases 54. (d) Simplifying choices. Other operational empowerment tools aim at facilitating people’s choice by making comparison among products or services easier. “Pro-choice” web applications are frequently used in the area of utilities, securities, bank and insurance services 55 ; they may be required of the private sector by public authorities, or run directly by the latter, to ensure really independent comparisons 56 . By allowing consumers to save search costs and by providing them with easy comparisons of existing commercial offers, these tools may prove very effective in overcoming inertia and status quo biases. In the UK, for instance, following the energy Retail Market Review, Ofgem implemented new regulation requiring retailers to provide consumers, via bills and annual statements, with personalised consumption information. In addition, consumers are set to be advised on the best deal available for them, based on medium usage of gas and electricity (the so called Tariff Comparison Rate (TCR). The TCR, combined with the energy reports, are meant to allow consumers to make full and accurate cross market comparisons 57. (e) Overcoming emotional responses. Another operational empowerment strategy is represented by cooling off or “timing of choice” rules. These are intended to help people make considered choices and overcome emotional responses, based on a waiting period being imposed by the regulator before a final decision (for instance, to buy) is made 58 . A cooling off rule, eventually supported by empowerment through simplification of requested 54 See e.g. L.E. WILLS, Against Consumer Financial Literacy Education, in Iowa L. Rev., 94, 2008, p. 12 (lsr.nellco.org/cgi/viewcontent.cgi?article=1213&context=upenn_wps) contending that it could worsen problems of overconfidence. 55 Reference to PQCWs (Price and Quality Comparison Websites) as tools to help people choosing and saving money can be found in the Commission’s First report on the application of Directive 2005/29/CE, cited above, note 28, at §3.4.2. On the benefits of RECAP (Record, Evaluate, and Compare Alternative Prices) regulation to help consumers comparing “complex pricing schemes that are neither transparent nor comprehensible to consumers” see R.H. THALER, C.R. SUNSTEIN, J.P. BALZ, Choice architecture, in E. Shafir (ed.) The behavioural foundations of public policy, Princeton, Princeton University Press, 2013, pp. 429-438, at 435. On “product attribute information” and “product use information” as disclosure regulatory strategies see E. BRODI, Abitudini e preferenze di consumo, MCR, 2, 2012, pp. 393-420. 56 Price comparison web applications accessible on the regulator’s portal are frequent in Italy: AEEG’s “Find your (energy) deal” (trova offerte); AGCOM’s (the telecom regulator) “Misura Internet” and IVASS’ (the insurance regulator) “Your Quotation” are all aimed at easing deals’ choice. 57 See Ofgem, The Retail Market Review– Implementation of Simpler Tariff Choices and Clearer Information, of 27 August 2013 (https://www.ofgem.gov.uk/sites/default/files/docs/decisions/the_ retail_market_review__implementation_of_simpler_tariff_choices_and_clearer_information.pdf). 58 Think of the 15-day cooling off rule for purchases made on Internet, foreseen in the EC Consumers Rights Directive 2011/83/EU. www.federalismi.it 16 information, could be a good strategy to help those that, having misjudged commercial offers, changed their provider and ended up paying more59. A cooling off rule could be inserted in contracts for consumers who have switched over from one provider to another, enabling them to return to the previous one without incurring any cost 60. (f) Strengths and weakness of operational empowerment. Operational empowerment tools analysed so far present many strengths as well as weak points. Most of them tend to preserve individual choice by easing it: indeed, they do not frame the choice environment in an aim to exploit automatic or affective responses; rather, they use framing to over-simplify the information individuals use to act, or they aim at providing tools to help individuals overcoming emotional responses. Therefore, operational empowerment is less prone to criticism of manipulation than are other regulatory tools, such as nudge (see later). With regard to targeted education, simplification and choice-easing tools, operational empowerment makes use of information-based regulation to empower individuals to take selfenhancing decisions. In that, it assumes that cognitive errors can be avoided and biases overcome, differently from nudge strategies and in line with the “consumer sovereignty” paradigm. Thirdly, easing choice and simplification can decrease the power of inertia and bring out important benefits such as an increased compliance with regulation and greater participation in public programs 61. Operational empowerment, thus, allows individuals to behave at the same time coherently with their own as well as a given public interest: while increasing market transparency and reducing information asymmetry, it also fosters competition among service providers because it makes choice easier. Furthermore, by strengthening individuals’ ability to exploit market possibilities, it serves rebalancing the demand vis à vis the supply side, a contend that is of particular value in recently liberalised markets. Despite their many advantages, operational empowerment strategies are not without limits. Tools in this box can be more expensive, for instance, than C&C strategies to design and implement: to design because they require a prior identification of groups in the relevant market that suffer from specific biases and errors; to implement because they can put an 59 C.M. WILSON and C. WADDAMS PRICE, Do consumers switch to the best supplier?, Oxf. Econ. Papers, 4, 2010, pp. 647-668. 60 Here we dissent from the Authors who classify cooling off rules among nudging tools (see e.g. R. KOROBKIN, Libertarian Welfarism, California Law Review, Vol. 97, 2009, pp. 1651-1685, at p. 1664). 61 C.R. SUNSTEIN, Empirically informed regulation, The University of Chicago Law Review, 78, 2011, pp. 1349-1429, at p. 1351. www.federalismi.it 17 excessive burden over the private sector 62. Furthermore, being operational empowerment still recent, its effects and outcomes are far from being fully assessed. Another weak point is the risk, well known in the patient empowerment debate, of aversion to be empowered. Sometimes patients do not wish to take an active role in decisions about their healthcare (e.g. patients in cancer treatment tend to be less willing to participate as their conditions worsen); and the same may happen with utility or financial services’ consumers, whose willingness to engage in costly self-education activities may be very feeble compared with the perspective gains 63. Finally. targeting too much may endanger the efficacy of operational empowerment as, for instance, consumers may decide to invest less and less in self-education, ending up being disempowered. Also, empowerment though information simplification may limit product differentiation, which is based on consumers’ accumulation of knowledge about the products, as well as technological innovation 64. In these cases, where it is proven that heuristics and errors persist despite the enactment of empowerment strategies, nudges might prove more effective. 4.2 Nudge strategies There is no clear definition of nudging, and it can include a wide variety of approaches 65 . According to Thaler and Sunstain nudging is “any aspect of the choice architecture that alters people’s behaviour in a predictable way without forbidding any options or significantly 62 Indeed, many of the (regulatory) operational empowerment tools consist of requirements that the industry is expected to implement. While the resulting costs may prove acceptable for some firms, they may be excessive for other, smaller ones. 63 E.g. despite the simplification of bills (where there is no overload of information, but information is selected and rationalised), there can still be room for biases and cognitive errors. 64 See X. GABAIX, A. LANDIER, D. THESMAR, P. MONGIN and J. TIROLE, La protection du consommateur: rationalité limitée et régulation, cited above, note 30, p. 9. 65 The precise boundaries of what should or should not account as nudging are far from set. The acknowledgment that not all strategies described in the influential book of R.H. THALER and C.R. SUNSTEIN (Nudge: Improving Decisions about Health, Wealth, and Happiness, Yale (CN), Yale University Press, 2008) are in fact nudges is not new. For instance,: in discussing the practice and ethics of choice architecture and nudging, E. SELINGER and K. WHYTE (Is There a Right Way to Nudge? The Practice and Ethics of Choice Architecture, Sociology Compass, 5/10, 2011, p. 923 ss.) clarify that many practices are mistaken or fuzzy nudges. C.R. SUNSTEIN also seems to have had “mixed feelings” about his creature: his 2011 work (Empirically Informed Regulation, The University of Chicago Law Review, Vol. 78, Issue 4, p. 1349 ss.) responds to many of the criticisms but avoids even using the term nudge (one citation in a footnote out of an 80page article). From 2013 he went back campaigning for nudging, with more sophisticated classifications and articulation (Nudges.gov: Behavioral Economics and Regulation, February 16, 2013 (forthcoming E. Zamir and D. Teichman (eds.) Oxford Handbook of Behavioral Economics and the Law). Interestingly, the latter uses the same classification and articulation that were used in the 2011 paper, but with nudging back at centre stage. We should not therefore be surprised if one strategy is categorised as nudging in one place and later no longer so. www.federalismi.it 18 changing their economic incentives” 66 . We support the idea that what should really account as nudges are interferences that actually do influence the set of preferences of regulatees acting in their emotional area (not in their economic sphere). Here by “emotional area” we mean in fact the way emotional responses of our brain (the automatic, affective thinking of the dual-process theory in psychology 67 ) to external stimuli make us behave – according to current scientific (psychological) knowledge – on average and in a specific relevant market. Below, some nudge strategies are classified in three main categories and exemplified through implementation in different fields. These nudge strategies are analysed only from a rulemaking point of view, meaning nudges introduced by rules, and not those originated by a nonlegislative approach (such as public information campaigns), an antitrust decision, or selfregulation. (a) Default rule. The default rule is based on a presumption of consent unless the interested parties exercise a positive opt-out. Considerable evidence suggests that the choice of the default rule can have a significant effect on behaviours and outcomes, provided that it is simple and essentially costless to opt-in or opt-out. This strategy “uses” inertia (i.e. that we often prefer not to change our habits unless we really have to) in order to nudge people to choose something considered to be better for them. Moreover, “the default rule might be taken to carry an implied endorsement by those who have chosen it; people may not depart from the default rule on the ground that it might have been selected because it is helpful or appropriate” 68. People are also more concerned with losses than with potential gains 69 and where the potential gains or losses of making a choice are unclear, to accept the default is often the preferred option, because it costs nothing in time and effort 70 . Therefore, default rules are designed in order to “alter outcomes” with respect to health care, consumer 66 R.H. THALER and C.R. SUNSTEIN, Nudge, cited at previous note, p. 6. Or intuition (as opposed to reasoning) in D. KAHANEMAN’s view (A perspective on judgement and choice, American Psychologist, 58, 2003, pp. 697-720). The automatic, affective system is said to require “little or no cognitive engagement, being driven by immediate feelings and triggered by our environments”: T. MARTEAU, Judging nudging: can nudging improve population health?, British Medical J., Vol. 342, 2011, p. 263-265. See also F. STRACK and R. DEUTSCH, Reflective and impulsive determinants of social behaviour, Pers. Soc. Psychol. Rev., 2004, pp. 220-247. 68 US OIRA, Disclosure and Simplification as Regulatory Tools, p. 9. See also Sec. 4. US Executive Order 13563 of January 18, 2011, Improving Regulation and Regulatory Review. 69 D. KAHNEMAN and A. TVERSKY, Prospect Theory: An Analysis of Decision under Risk, Econometrica, Vol. 47, 2, 1979, p. 263. 70 C.R. SUNSTEIN, Impersonal Default Riles vs. Active Choices vs. Personalized Default Rules: A Triptych, 5 November 2012, unpublished manuscript, p. 11 (http://nrs.harvard.edu/urn-3:HUL.InstRepos:9876090). 67 www.federalismi.it 19 protection, the availability of human organs 71 , energy use and environmental protection 72 , mortgages, savings and many other topics 73. The default rule is not new. In many regulatory areas one can find examples (take for instance the privacy regulation, where people are considered to be against the use of their personal data, while they do not an explicitly opt in). What is new is the awareness that inertia can have a great impact in determining behaviour and that this force can be used in order to make regulation more effective. This rule can be drawn in a simple way (such as in the case of automatic enrolment, for instance, in a retirement savings plan), or in a more sophisticated way (e.g., the SMarT option feature in a retirement savings plans allows an automatic annual increase in contributions of payroll savings).74 To sum up, whatever the reason, in some contexts the default rule can make regulation effective 75 and can promote automatic compliance with the law 76; though not always nor in every situation. This is to say that biases are not universal or, at least, they have to be verified in a specific relevant market and in relation to different individual preferences. On this ground, differentiated default rules (i.e. based on different abilities to opt out, eventually grouped by geographical areas or people’s past choices) might be taken take into consideration by rule-makers 77. (b) Smart information nudge. Attention is a scarce cognitive resource and there is a limit to the processing capacity of the human brain 78 . Because of such characteristics of human 71 Countries where people are required to opt -out of organ donation (such as Austria, Belgium, France, Hungary, Poland, Portugal, Sweden) report significantly higher consent than those with an opt-in policy (E.J. JOHNSON and D. GOLDSTEIN, Do Default Save Lives?, Science, Vol. 302, 21 November 2003, p. 1338). 72 D. PICHERT and K.V. KATSIKOPOULOS, Green defaults: Information presentation and pro-environmental behavior, Journal of Environmental Psychology, 28:1, 2008, pp. 63-73. 73 C.R. SUNSTEIN, Impersonal Default Riles vs. Active Choices vs. Personalized Default Rules: A Triptych, cited above, note 71. 74 By this well famous nudge, average saving rates for Save More Tomorrow (SMarT) program participants increased from 3.5% to 13.6% over the course of 40 months.(R.H. THALER and S. BENARTZI, Save More Tomorrow, cited above, note 17. See also B. ALEMANNI, Pensioni, comportamenti e regole, Analisi giuridica dell’economia, n. 1/2012, pp. 164-187). A similar default rule introduced in Italian regulation of pension top-ups (previdenza complementare) for the private sector had limited effectiveness (in 2009, two years after the entry into force of the new regulation, the workers involved made up 20% of the workforce, as opposed to 13% in 2004). 75 The percentage of citizens currently signed up as donors in those countries operating informed consent systems ranges from 4 to 28 per cent. By contrast, countries with presumed consent systems achieve donor registration rates of between 86 per cent and 100 per cent (P. JOHN, S. COTTERILL, A. MOSELEY, L. RICHARDSON, G. SMITH, G. STOKER and C. WALES, Nudge, Nudge, Think, Think, Experimenting with Ways to Change Civic Behaviour, London, Bloomsbury Academic, 2011, p. 110). 76 C. SUNSTEIN, Empirically Informed Regulation, cited above, note 62, p. 1398. 77 C.R. SUNSTEIN, Impersonal Default Riles vs. Active Choices vs. Personalized Default Rules: A Triptych, cited above, note 71. 78 “The capacity which can be allocated to various activities is limited and the limit varies with the level of arousal” (D. KAHNEMAN, Attention and Effort, Englewood Cliffs (NJ), Prentice-Hall series in experimental psychology, 1973, p. 13). However, while an important activity is under way, the attention of the subject is not necessarily increased if more crucial information is given (N.N. TALEB, The Black Swan: The Impact of the www.federalismi.it 20 perception and decision, the framing of acts, contingencies, or outcomes might change regulatees’ perception of the desirability of an option 79. Moreover, information that is salient “can have a larger impact on behaviour than information that is statistical and abstract” 80. Of course, “the practice of acting on the most readily available frame can sometimes be justified by reference to the mental effort required to explore alternative frames and avoid potential inconsistencies” 81 . However, when it is not the case, using such knowledge about framing and salience, the rule-makers can draft smart information nudges, in order to encourage end-users’ choices by providing information structured and framed to be salient so as to achieve effects on individual behaviour. This is something different from the simplification of information given to consumers as an operational empowerment strategy (supra Para. 4.1), which is limited to influencing behaviour merely by facilitating people’s choice. In the smart information nudge scheme, data are provided in a smart, say “relational” way as they include comparisons and unspoken assessments in order to orientate individual behaviours by leveraging on their emotional sphere. An example is the “tell people what others are doing” strategy: in many North American cities, energy saving has significantly increased by sending out personalized statements about energy use. Such reports rate people on their energy use compared with that of neighbours in 100 homes of similar size where the same heating fuel was used, and also compared with the 20 neighbours who were especially energy efficient 82 . Here, social influence (or the perceived behaviour of peers) is used to push householders to consume less energy, as it has been proven that perception of the norm in the pertinent community can affect human behaviours more than rules. The main benefit of this strategy is that it requires nearly no effort in terms of enforcement, while ensuring high rates of compliance 83 . However, a special attention must be granted in choosing and drafting such nudges. On one hand, their efficacy utterly depends on whether the same mental mechanism will occur Highly Improbable, 1st ed., New York, The Random House, 2007). Therefore, as H.A. SIMON had pointed out already in 1971, the information overload consumes the attention of its recipients (Designing Organizations for an Information-Rich World, cited above, note 45). 79 A. TVERSKY and, D. KAHNEMAN, The Framing of Decisions and the Psychology of Choice, cited above note 38, p. 458 (prospect theory). 80 C.R. SUNSTEIN, Empirically Informed Regulation, cited above, note 62, p. 1354; R.E. NISBETT, E. BORGIDA, R. CRANDALL and H. REED, Popular Induction: Information Is Not Necessarily Informative, in D. Kahneman, P. Slovic and A. Tversky (eds), Judgment under Uncertainty: Heuristics and Biases, Cambridge, Cambridge University Press, 1982, 112. 81 A. TVERSKY and D. KAHNEMAN, The Framing of Decisions and the Psychology of Choice, cited above note 38, p. 458. 82 L. KAUFMAN, Utilities Turn Their Customers Green, With Envy, The NY Times, January 31, 2012. 83 C.R. SUNSTEIN, Empirically Informed Regulation, cited above note 62, p. 1349. www.federalismi.it 21 everywhere in the relevant market and for all in the targeted group 84 . On the other hand, it might produce a boomerang effect, in terms of undesirable reactions of people who behave above the average (e.g. people who already save energy), which can be neutralised by adding a message indicating social approval or disapproval (e.g. through a positive or negative emoticon) 85. (c) Exploiting the emotional response. This is a nudge strategy which encourages end-users choices by exploiting their emotional responses (even by neutralizing them) in order to achieve effects on individual behaviour. The Framework Convention on Tobacco Control (FCTC) and a proposal to revise the current European directive on tobacco products 86 sets out a standardization of tobacco product packaging (the so-called plain or generic packaging) with the only remaining possibility being “to print brand and product names, the quantity of the product, health warnings and other mandatory information, such as security markings” 87. An example of a nudge to more physical activity (and so better health) could be a rule which obliges building managers to make stairs more prominent and attractive in public buildings, over the lifts, escalators or elevators (for instance by transforming them into a piano keyboard which plays when stepped on) 88. (d) Strengths and weaknesses of nudging. Nudge strategies have strong points as well as weaknesses. They are usually cheap for regulators and are innovative approaches in areas where traditional regulatory tools have not worked (meaning that they have not given a satisfactory answer to the need of the regulation on which they are based). At the same time, they risk limiting (and in the worst case, manipulating) the regulatees’ choices. A first critical point concerns the identification of the social welfare area to be maximized – a discretionary decision at the best of times –. Need the decision-makers nudge in order to 84 Behavioural Insight Team, Applying behavioural insights to reduce fraud, error and debt, 2012, p. 17. Indeed, the attempt to mobilise action against a problem may lead to drafting a message which depicts it as regrettably frequent (“Many people are doing this undesirable thing”) and in so doing they end up communicating that “Many people are doing this” (in other words, the behaviour which is to be curtailed is in fact being highlighted as that which everyone is doing (R.B. CIALDINI, Crafting Normative Messages to Protect the Environment, Current Direction in Psychological Science, 2003). 86 Possible revision of the Directive n. 2001/37/EC of the European Parliament and of the Council, public consultation document 2010. 87 E. BONADIO, Regulating Tobacco, Alcohol and Unhealthy Food and Beverages: the Impact on, and Role of, Intellectual Property, 2nd HEC Paris Workshop on Regulation: Regulating Lifestyle Risks in Europe - The Case of Alcohol, Tobacco, Unhealthy Diets & Gambling, 20 e 21 September 2012, http://appli6.hec.fr/forms/riskregulation-2012). However, quantitative data on the efficacy of this tool of nudge to less consumption does not seem to exist (some undetailed information is available in the study to support a DG SANCO Impact Assessment, Assessing the Impacts of Revising the Tobacco Products Directive, 2010). 88 One way is suggested by “the fun theory” which nudges more people to do something by making it fun. 85 www.federalismi.it 22 increase or protect the welfare of a limited group of people or that of everyone? The choice to maximize the welfare of a single target group is not without consequences: this expression of “pure” libertarian paternalism could encourage people to take optimal decisions for themselves, without considering the interests of other people 89 , for instance, by refusing to adopt an expensive system intended to limit industrial pollution (externality problem). At the same time, is the so-called libertarian welfarism intended to consider all citizens or only the weakest? And how are they to be correctly identified? Moreover, the identification of social welfare is difficult when there is a high degree of uncertainty (for instance, regarding which new energy technologies or renewable energy sources are to be incentivized), and impossible when there are a wide variety of needs to be satisfied. Another critical point is about the potential threat that cognitive based regulation might pose to liberty and individual autonomy, it being difficult to identify individuals’ true preferences 90 . These difficulties may lead the regulators towards error, and end up allowing them to impose their vision of that which is best for end-users. A third critical point concerns transparency. The nudge strategies are less identifiable than those connected to “traditional” paternalism. For this reason, these are usually strategies where end-users have not participated nor shared, unless the decision-makers take steps in this direction (for instance, explaining the expected effects of a default option during the consultation process) 91. These critical points might (partially) be settled in an open and transparent rule-making process, informed by an ex ante assessment of possible regulatory options (where needed according to the proportionality principle) 92. Other critical points relate to specific tools. On one side, the default rule is usually the most effective and least expensive (for regulators), but also the most controversial of the nudge strategies. The most widespread and well89 R. KOROBKIN, Libertarian Welfarism, cited above, note 61, p. 1665 ss. J.D. WRIGHT and D.H. GINSBURG, Behavioral Law and Economics, cite above, note 11. 91 “Perhaps there is an argument here for allowing citizens the opportunity to help design the nudges in order to increase the legitimacy of any changes to choice architecture. In the case of organ donation, a national debate or even a referendum on changing the registration system from an opt-in to an opt-out or a mandated choice system might be a useful way forward” (P. JOHN, S. COTTERILL, A. MOSELEY, L. RICHARDSON, G. SMITH, G. STOKER and C. WALES, Nudge, Nudge, Think, Think: Using Experiments to Change Civic Behavior, London, Bloomsbury Academic, 2011). 92 However, is it undoubtedly true that “administrative law should adapt the conceptual framework and the instruments by which it provides a real control of the way power is used by public authorities. (…) [because] the informality and the informational asymmetry of behaviourally-informed strategies appear to circumvent intrinsically the safeguards of the principles of legality and impartiality of public administrations” (A. ALEMANNO and A. SPINA, Nudging Legally. On the Checks and Balances of Behavioural Regulation, NYU School of Law – Jean Monnet Working Papers, Vol. 6/20013, forthcoming). 90 www.federalismi.it 23 founded criticisms concern its libertarian nature, and are basically connected to the ease of opting-out, which risks being overturned if it proves to be problematic, for any reason (for instance because it incurs costs – time or monetary – or if the opt-out possibility is not clear) 93 . On the other side, the “smart information” and “exploiting the emotional response” nudge strategies risk being insufficiently effective because the introduction of a new or different consumption model could take years in order to bear fruit. In general, it is also difficult to measure its level of efficacy due to the fact that these tools are not usually the unique regulatory strategy. For instance, even if plain packaging is adopted for cigarettes, healthrelevant information must always be available; in addition, the advertising ban and the prohibition of sales to minors must remain in force. At the same time, there are other nudges used along with the plain package (such as - in Australia - the shock images of diseases caused by smoking) which could paradoxically counterbalance the potential efficacy of the former. Indeed, the optimistic and availability bias lead us not to consider the possibility that these events could concern us. 5. Lesson drawing for regulators: how to choose among different regulatory options? The paper shows that biases in judgment offer crucial information to rule-makers on the reactions of end-users, and in so doing they enable the better formulation of rules and the provision of more adequate responses to the public interest they are intended to satisfy. In order to attain these results, the regulatory process must accordingly change. It is important to open rule-making processes through consultations organised and led in order to bring out potential cognitive errors. Moreover, when – according to the proportionality principle – RIA is used in rule-making processes, this tool should be enriched by a risk analysis which takes into account the use of heuristics as a risk to be assessed in terms of probability and effects and might help rule-makers to decide whether or not to deal with cognitive errors through regulation. Moreover, evidence from the cognitive sciences leads to an enrichment of the traditional regulatory options (such as those based on the command and control approach, and the incentive-based regulation), and has contributed to the emergence of new regulatory tools (such as operational empowerment and nudging). 93 The feasibility of the opt-out is the centrally important issue of this tool, and if the ease to opt out is not guaranteed, this approach “sows the seeds of an illiberal system of control” (R. BALDWIN, The New Scholarship: Celebrating the ‘I’ in Ideas, LSE Law, Society and Economy Working Papers, 5/2012, p. 12; see also R. BALDWIN, M. CAVE and M. LODGE, Understanding Regulation, cited above, note 10, p. 123 ss.). www.federalismi.it 24 However, the very presence of cognitive errors and biases, even if recurrent in given relevant markets, does not (and we think that should not) per se justify a regulatory intervention (free markets often providing significant protection against such errors) 94 . His rationale must, indeed, be found in a general interest which is unattainable because of the cognitive error and biases detected in a given relevant market. Nor should it justify the use of a single regulatory tool or the preference toward one tool among the many in a toolkit. Our contention is that many if not most regulatory tools could benefit from adopting a cognitive-based approach to the rule-making process. Nudging and operational empowerment are “sons” of the cognitive momentum, they bear a “cognitive DNA”. Other regulatory strategies, being less recent, have not necessarily been confronted by cognitive errors. Indeed, we contend that applying a cognitive-based approach to these might help reducing some of their weak points. An exercise can thus be made to reread them through the lenses of the cognitive sciences, i.e. in relation to the cognitive errors each regulatory strategy aims to avoid, modify or is affected by. For instance, a prohibition to negotiate (C&C) could be preferred to a duty to disclose information about conflicts of interest affecting a financial adviser, anytime it is demonstrated that (among other things) the targeted population in the relevant market is affected by overconfidence, while lacking any financial education. Table 1 at the end of these notes summarises the characteristics, weaknesses and strengths of the regulatory strategies analysed so far, having exclusive regard to their attitude toward cognitive errors. While confirming that whatever selection of a regulatory strategy must be made on a case-by-case basis, having regard to the specific relevant market and the relevant goals (public interests) to be pursued, the main findings of the paper are now summarised so as to provide some indications that might prove useful for regulators. Sometimes a unique regulatory strategy may have limited efficacy and a combination of different strategies might be suggested to regulators. This is the case of the default rules, which tend to be coupled with framing of information regarding opt-out (an example of operational empowerment) 95. Similarly, nudging which tends to “exploit or neutralise” emotional responses is often matched with other – traditional – regulations (such as, in the case of tobacco, the provision of health relevant information and the prohibition of sales to minors). 94 For instance, operational empowerment tools are often introduced and implemented by private firms, without any regulatory constraint. 95 Sometimes default rules can be alternatives to mandates or bans (C&C) or operate to complement other regulatory requirements (C.R. SUNSTEIN, Empirically informed regulation, cited above, note 62, pp. 1398-99). www.federalismi.it 25 Sometimes tools belonging to different strategies may be used incrementally to gradually raise the effectiveness of regulatory intervention. Moreover, “smart information nudges” could increase the efficacy of both traditional information disclosure regulation and “information simplification” operational empowerment. So for instance, traditional information disclosure tends to increase the amount of data provided, causing an adverse reaction of real people to complex information, who, consequently, will go in the opposite direction of the desired behaviour (for instance, people will hold onto risky investments, and they will not save energy). If it is proven that too much information has effects equal to not being informed (e.g. about financial risks and potential savings), then some operational empowerment tools, such as framing or targeted education, could be justified to strengthen the efficacy of traditional regulation based on information. Also the “simplification of choice” tool could be more effective (e.g. in urging provider switching) if coupled with “targeted information”. However, if a feedback of low switching rates still results, regulators could consider overcoming inertia by using “smart or relational information nudges”. In other words, they could exploit one social norm (e.g. emulation within a group) and emphasise positive models in the community (e.g. those who switched are better off) so to create new virtuous habits. Of course, the efficacy of such nudges utterly depends on whether the same mental mechanism will occur everywhere in the relevant market and for everyone in the targeted group 96. In other cases, one (cognitive based) strategy might be preferred and considered an exclusive regulatory answer, where is proven that the empowerment might decrease the effectiveness of nudging 97 . Moreover, where cognitive error is evitable, we would suggest to prefer operational empowerment regulation, which is more transparent than nudging and preserves regulatees’ autonomy. This is especially true in the case of default rules, the only nudge strategy that is really effective. Indeed, even if opting out is low-cost and easy to exercise, it might end up imposing a vision of end users’ welfare that belongs to the regulator. In such 96 Even the Cabinet Office Behavioural Insight Team stresses what seems clear: this type of nudge could be effective if it strengthens a pre-existing social norm, and that “it will not be appropriate to highlight a descriptive social norm, in particular where large numbers are engaged in non-compliant or problematic behaviour” (Applying behavioural insights to reduce fraud, error and debt, cited above, note 47, p. 17). 97 However, a mix of regulatory strategies is not always the best solution. An experiment concerning organ donor registration suggests that “nudging by itself is more effective than nudging and thinking in encouraging organ donor registration. Tailored information booklets, based around behavioural economics principles, created increases in organ donor registrations and were more effective than combined information and discussion. It seems that the added effect of discussion generates some uncertainty. (…) Since the discussions themselves allowed participants to explore the complexity of the topic and reflect in more detail, this enabled them to see different sides of the issue” (P. JOHN, S. COTTERILL, A. MOSELEY, L. RICHARDSON, G. SMITH, G. STOKER and C. WALES, Nudge, Nudge, Think, Think, cited at note 91). www.federalismi.it 26 circumstances, an empowerment strategy could prove less intrusive and more autonomypreserving, provided that its effectiveness in overcoming or avoiding cognitive errors is proven. It might indeed happen that an empowerment tool increases, instead to decreasing, the incidence of a given bias (such as overconfidence). To sum up, when educational strategies and information (operational empowerment) do not allow the avoidance of cognitive errors (meaning that cognitive errors are unavoidable), regulators might use them in order to build soft incentives (nudges) so as to attain results. However, this may not be the best option when the protected values are particularly sensitive (e.g. health, safety, etc.) and need a strong answer 98 : in such circumstances, more effective regulations, such as those based on command and control, might be preferable. A final note regards the possible consequences on rule drafting and implementation of adopting a cognitive based approach. Addressing real people, gathering information about individuals’ errors and reactions, consulting so as to let group preferences emerge, targeting regulation, i.e. the “core” of the new approach, entails some degree of differentiation when drafting rules and applying them 99. In the realm of regulation differentiation, indeed, this has often been associated with greater compliance. It follows that differentiating rules on the basis of recurrent errors and biases within a relevant group would be not so different, with the positive outcome of adding weight to regulation and increase the probability of greater efficacy in spreading virtuous behaviours. Furthermore, if following an accurate analysis of the targeted groups (differentiated by age, income, gender, geographic location, etc.) it results that an educational strategy was not effective in overcoming or avoiding cognitive errors, then nudge strategies such as “smart information nudge” could be justified. We should caution, however, that using nudging as a last resort strategy may be a source of serious consequences: if not effective, nudge strategies risk leaving the public interest they are intended to satisfy without adequate protection. 98 For instance, R.H. THALER and C.R. SUNSTEIN claim that “we are all aware that for environmental problems, gentle nudges may appear ridiculously inadequate – a bit like an effort to capture a lion with a mousetrap. (…) Even libertarians tend to agree that when externalities are presents, markets alone do not achieve the best outcomes. (…) People who celebrate freedom of choice are well aware that when “transaction costs” (technical terms for the costs for entering into voluntary agreements [between polluters and damaged people in order to get the first to clean up]) are high, there may be no way to avoid some kind of government action” (Nudge, cited above at note 66, pp. 184-185). 99 For instance, “personalized default rules might be demographic; (…) alternatively, personalized default rules could be very narrowly targeted. If enough information is available about someone’s past choice or personal situation, we could design, for that person, a default rule with respect to health insurance, privacy, rental card agreement, computer setting, and everything, else” (C.R. SUNSTEIN, The Storrs Lecture: Behavioral Economics and Paternalism, Yale Law Journal, 2012, forthcoming, p. 35; see also Empirically informed regulation, cited at note 62, pp. 1399). www.federalismi.it 27 Therefore, the answer is always, both for operational empowerment and nudging, the endusers’ participation and transparency of the rule-making process. www.federalismi.it 28 Table 1 – Regulatory strategies in relation to cognitive errors STRATEGY REGULATORY TOOL CHARACTERS PROS IN RELATION TO COGNITIVE CONS ERRORS Command & Control Bans, duties, standards (e.g. seat belts) Avoids risk of cognitive errors; General rules that apply to everyone (disrespectful of cognitive errors) Low cost design (except for standards) and implementation; Can help reducing creative compliance Paternalism; Risk of cognitive errors by the rule-maker; Over-regulation and excessive limitations on those not affected by the cognitive error Disclosure regulation Ex ante information disclosure duties; informed consent; Ex post (control of) prohibitions of false information; misleading adversiting, un fair commercial practices Neutral with regard to individual preferences and cognitive errors Preserve autonomy Does not ensure effective comprehension; Can esacerbate information asimmetry (information overload) Legal Empowerment Ex officio powers by public bodies to co-enforce individual rights; Class actions, collective reddress; Simplification of litigation (Alternative Dispute Resolution) Avoids risk of cognitive errors; Assumes weakness of regulatees (disregarding cognitive errors) Strengthens deterrent effect of private enforcement; Does not limit private autonomy Heavily paternalistic (complements individual autonomy with public intervention); Lacks delegation; Stuck to notional (average) consumer Incentive Regulation Economic incentives; differentiated tax regimes; subsidies (e.g. benefits for energy savers; higher taxes for polluters) Assumes rationality of regulatees; General rules that apply to everyone (disregarding cognitive errors) Preserves autonomy; Easy to enforce Operational Empowerment Simplification of information and smart disclosure; Framing; Simplification of choice tools (e.g. price comparison apps); Targeted education Overcoming emotional responses (e.g. cooling-off rules) Aimed to avoid or overcoming cognitive errors (truly de-biasing techniques); Emphasis on: education, information simplification, overcoming emotional responses Preserves autonomy; Reduces information asymmetry; Strengthens demand vis à vis the supply; Promotes competition Can increase compliance with the law and participation in public programs Can be costly to define an incentive scheme; Allows regulators to let their vision prevail; Lets economic interests prevail on other motivations (prone to capture by private interests) Costly to design; Efficacy in overcoming biases not assessed yet; Aversion to be empowered Nudging Default rules; Smart information nudge; Exploiting emotional responses Is not aimed to avoid cognitive errors; Exploits bias Libertarian paternalism Low cost design; Potentially successful in changing behaviours where other tools fail 29 Risk of manipulation; Lack of transparency; Low coercion but still limits autonomy Efficacy not assessed yet