• 1 Jolls, Sunstein, and Thaler (1998) is a programmatic overview. Useful surveys of BLE include Korobk (...)

1Behavioural Law and Economics (BLE), now almost twenty years old, is roughly and easily characterized as “both the development and the incorporation within law and economics of behavioural insights drawn from various fields of psychology” (Jolls, 2006, 3). BLE has been growing rapidly while its academic reception has been mixed.¹ Indeed, even the proper term for this field of research is disputed, some suggesting “behavioural analysis of law”, “legal decision theory”, or “new law and psychology” instead (Tor, 2008, 241, note 13). It seems that by now, in spite of many imprecisions, the term BLE is well-established enough so that it is generally recognised if not generally accepted. This paper discusses a few meta-theoretical questions about BLE in order to better understand both its popularity and the criticisms it has received. The aim of this paper is, however, not to provide a historical or sociological account of the inception and reception of BLE. Rather, it is to account for BLE at a meta-theoretical level, focusing on some of its philosophical and methodological assumptions.

• 2 I have discussed the issues under (1) and (2) in Cserne (2015).

2When talking about BLE, all three words are relevant. It surely matters how we understand ‘behavioural’. Yet, we can only fully appreciate the changes that this sub-discipline has generated and is likely to generate both in academia and in the real world if we look at ‘law’ and ‘economics’ as well. In particular, the degree of academic and practical openness or the strength of academic and practical resistance to BLE not only depends on (1) whether the behavioural insights, i.e. the findings of empirical research about legally relevant human judgement and decision making are framed as ingredients of an empirically informed policy science, as a branch of psychology or as cognitive neuro-science applied to legal phenomena, but also on (2) whether law is seen as an autonomous or an instrumental social practice, and (3) on how economics is seen in relation to empirical and theoretical knowledge as well as to policy. These distinctions are rarely made explicit and the implications of the various understandings are rarely discussed in the literature on BLE. This paper is an attempt to improve on this situation in some respects. In particular, given the disciplinary focus of this journal, I shall concentrate on the third issue, i.e. on how BLE connects to and provokes economics.² The main claim of this paper is that the rise of BLE provides a litmus test: it reveals methodological and substantive dividing lines in both positive and normative economics.

• 3 Thanks to Magdalena Małecka for drawing my attention to this distinction.

3To be sure, BLE is linked to economics not in one but in two steps, mediated by Behavioural Economics and Law and Economics. Given the relative disciplinary autonomy of Law and Economics with respect to economics, the methodological and substantive claims of Behavioural Economics may play out somewhat differently in Law and Economics than in other areas of economics. So when in the following I talk about economics, it is a somewhat simplified way of talking about “economics as understood within Law and Economics”.³ Strictly speaking, this component is not identical with economics tout court. What is metaphorically called a litmus test concerns the way BLE challenges the economic component of Law and Economics.

• 4 Law and Economics is a sub-discipline of legal studies as well but this is not our concern in this (...)
• 5 For a broad understanding of schools of Law and Economics see Mercuro and Medema (2006), discussing (...)

6First, BLE can be seen as a sub-discipline of Law and Economics, itself being a thematic area of economics (canonically recognised by the JEL code K).⁴ Thus, BLE can be characterised by its aim, with respect to the enterprise of Law and Economics: “Behavioural law and economics attempts to improve the predictive power of law and economics by building in more realistic accounts of actors’ behaviour” (Jolls, 2006, 3). This characterisation suggests that BLE is an improved version of Law and Economics, i.e. the term ‘behavioural’ serves as a qualifier to ‘law and economics’. This implies a certain starting point regarding the kind of economics and the approach to law which are to be “improved”, at least insofar as mainstream Law and Economics can be plausibly considered a homogeneous endeavour with a well-defined epistemic and methodological character. However, at closer look one can observe much heterogeneity within Law and Economics.⁵ Some of these dividing lines are accentuated by BLE, as I will discuss below in detail.

• 6 Parts of criminology or forensic psychology could serve as examples. Zamir (2016, 4) also lists soc (...)
• 7 Zamir (2016, 1-3) provides a succinct summary and references. On the problem of overreliance on a p (...)
• 8 See also Zamir (2016, 4): “Behavioural insights pose a major challenge to standard economic analysi (...)
• 9 The reasons for this development raise historical or sociological questions which are interesting i (...)

9Whether one or the other genealogy is more persuasive, there is no doubt that BLE is linked to economics. In fact, at least in the last decades, the findings of behavioural sciences have often come into the view of legal scholars and practitioners through economics, rather than directly from psychology and cognitive science. To be sure, there have been many interdisciplinary connections between psychology and law as well, often with no involvement of economics or economists. ⁶ Not all transfer routes start from judgment and decision making research (the sort of psychological research most closely associated with behavioural economists)⁷ and not all lead through economics.⁸ As Zamir (2016, 4) pointed out, “the association of behavioural studies with economic analysis of law—rather than with legal analysis in general—is largely an artefact of historical development, rather than an analytical truth.”⁹

• 10 The first step of this transfer has been discussed, analytically in the encyclopaedia entry by Angn (...)

11In sum, BLE is the result of a double interdisciplinary transfer. First, insights from psychology have been taken seriously by economists, giving rise to behavioural economics, and second, in response to both theoretical and practical impulses, Law and Economics has been transforming into BLE and/or Empirical Legal Studies.¹⁰ These transfers have brought about BLE, this somewhat strange interdisciplinary subfield, with its perceived in-between status between empirical research and policy, and hybridity of positive and normative analysis.

• 11 The idea that libertarians and dirigistes can agree on particular policies is in line with Sunstein (...)
• 12 Their most elaborate joint discussion of the underpinning conceptual and normative ideas is Sunstei (...)

13Behavioural Economics as an academic field is much broader than what policymakers or law professors have picked up from it. Many empirical researchers are cautious, and rightly so, not to draw far-reaching policy conclusions from their findings, and when they do, these depend on normative premises. Especially in the US, libertarian paternalism has often been expected and promised as a “third way”, a catholic i.e. politically all-encompassing pulling together of opposed political ideological factions —hence the seemingly oxymoronic term.¹¹ A key characteristic of libertarian paternalism is that it provides a normative benchmark or regulatory idea on how and through which means substantive goals can be pursued, namely, at least formally preserving individuals’ freedom to opt out. Reading through their various relevant publications, Sunstein and Thaler are, perhaps deliberately, never very precise about the substantive goals the government or “we” collectively should follow.¹² On some occasions, they tend to restrict their proposals to supporting individuals in reaching their goals, as the individuals themselves perceive it; on other occasions, they seem ready to accept social welfare as a collective normative goal overriding individual choice. On still other occasions, Sunstein (2013; 2014) refers to Mill’s harm principle as an imperfect way of drawing the boundaries of the domain of individual freedom of choice which can be corrected in light of our knowledge of human decision making and judgement. All in all, the normative claims associated with either BLE or libertarian paternalism do not amount to a full-fledged moral or political theory. While they contribute quantitatively new arguments to policy debates, the normative conclusions are not qualitatively new. Similar claims have been made for long, admittedly with little or at least different empirical backing than after the “behavioural turn”. Additional empirical claims about human frailties still do not commit one to a single standpoint in old oppositions, such as autonomy vs. paternalism, liberalism vs. communitarianism, or laissez-faire and regulation.

• 13 In a recent paper, Hansen (2016) discussed extensively the various definitions and characterisation (...)

15Nudge as a regulatory technique may or may not be informed by empirical research and guided by libertarian paternalism.¹³ As a conceptual matter, it is possible to have nudges which are introduced in a scientifically uninformed way and/or implemented in the service of any, perhaps sinister, goal. Also, under different normative premises behavioural economics may lead to policy implications very different from nudging. Our everyday experience shows that we routinely find ways to manipulate others (family members, colleagues, customers, etc.) in certain directions, without relying on their full cognitive and deliberative capacities. In doing this, we may be completely unreflective; or we may rely on behavioural assumptions, derived from sources as diverse as common sense, folk psychology or marketing research.

• 14 Cf. the remarks by Daniel Kahneman: “Actually, I think behavioural economics is the only hope for h (...)

16The regulatory techniques suggested by BLE are also not new in kind—we have all used menus and sticky defaults for long, marketing and business studies have analysed, and in fact advised clients on how to optimally design, choice architectures for consumers and firms. The techniques are new in that they are now suggested as policy tools for governments, thus presented in the main mode in which much of economics has been operating, at least implicitly: whispering advice in the prince’s ear (Sugden, 1986, 6-8). Policymakers provide a market for applied or potentially applicable empirical, especially experimental research that promises new results on processes of human cognition and decision-making in politically or legally relevant contexts.¹⁴

• 15 On evidence-based regulation see Cartwright—Hardie (2012), Sunstein (2011). See also Tor (2008, 291 (...)

17Practical problems and/or legal phenomena confront behavioural economics with different disciplinary cultures or more broadly, different standards as to what counts as an acceptable or convincing argument. Whether behaviourally informed policy proposals are taken up by governments and their advisors depends on a number of factors. One reason for openness may be that behavioural studies are expected to provide scientific backing for common sense ideas and intuitions about how “real people” behave, as opposed to the rational self-interested agents in standard economic models. This may turn critics of orthodox economic analysis into enthusiasts of behavioural economics. The high expectations about what empirical research can provide are in part based on misperceptions of how evidence based policy is made, disregarding many distortions during the long and complicated knowledge transfer and translation process from laboratory to regulation.¹⁵

• 16 While Katz (1998, 3-5) characterises the economic approach to law, he discusses these three premise (...)

18So far, I have only demarcated BLE from libertarian paternalism and nudges. Now the central question of this paper start to come into focus: how does BLE relate to economics? Take a stylized description of standard (law and) economics as a starting point. It is not too controversial that what distinguishes the economic approach to law is that it combines (1) positive economics (a particular descriptive theory of human behaviour based on rational choice theory), (2) normative economics (“a particular set of moral and ethical principles, which it uses to evaluate existing institutions and to recommend reforms”, Katz, 1998, 3), and (3) it “follows a particular methodological program in pursuing its aims—one grounded in the separation of positive and normative issues and focused on the construction of abstract models of empirical reality” (ibid.).¹⁶ Is BLE different from standard Law and Economics in any of these dimensions?

• 17 Tor (2008, 243) defines behavioural analysis of law as “the application of empirical behavioural ev (...)

19In an (otherwise) exemplarily clear and useful paper, Avishalom Tor (2008, 241, note 13) suggests that BLE “may be viewed as a combination of the behavioural analysis of law with the normative foundations of law and economics.”¹⁷ Tor’s formulation provides a useful starting point inasmuch as it focuses our attention on how BLE relates to both the explanatory and the normative projects of Law and Economics. It suggests a radical difference in the first (positive), complete continuity in the second (normative), and a more mixed picture in the third (methodological) dimension.

• 18 On the normative character of rationality and its use as an assumption in positive theory, see Elst (...)
• 19 For critical views on this link between behavioural and standard economics, see e.g. Gigerenzer and (...)
• 20 “Long before the contemporary behavioural economics program came to prominence, the economics disci (...)

20There is indeed clear difference in the first dimension: although rational choice theory still provides a normative or a heuristic benchmark,¹⁸ and so-called biases are often defined with reference to this benchmark,¹⁹ BLE pursues an agenda of constructing a theory of human behaviour based on empirical research. In crucial respects, this theory is markedly different from homo oeconomicus or even O.W. Holmes’s bad man, even though there are overlaps in certain domains. The fact that humans behave systematically different from what simple economic models assume is a commonplace, recognised in economics as well as in law, long before the emergence of BLE.²⁰ Yet, as I shall argue in section 5, it is not clear whether (law and) economics should completely reject or overhaul its rational choice-based behavioural models. In particular, the epistemic and methodological character of these models requires close scrutiny in order to establish if it is advisable to reject them or they should be left unchanged, improved or replaced.

• 21 See Posner’s revealing remarks (1983, vi): “Ethical theories seem to invite and receive shocking ap (...)

21As for the second dimension, it is only plausible to link BLE to a particular normative agenda if Law and Economics has or implies one. Arguably there has never been even an apparent disciplinary consensus about the “normative foundations” of (law and) economics, in the sense of a philosophical grounding for the less abstract policy arguments couched in terms of efficiency, welfare or cost-benefit analysis (Kornhauser, 1980; 1998). Even though the normative relevance of individual well-being (normative individualism) has been acknowledged quite broadly, neither utilitarianism, wealth-maximisation, nor any other welfarist or non-welfarist substantive goal, nor a procedural criterion such as consent has been generally accepted as a “normative foundation”. There have been typical viewpoints yet never articulated with much clarity. The suggestions for alternative normative foundations have also remained just as controversial as libertarian paternalism, social welfare or wealth maximization.²¹ As I shall argue in section 5, it is at least questionable whether the “normative foundations” of the discipline are well-identified or even identifiable. Furthermore, and somewhat ironically, empirical research raises doubts about the theoretical and practical usefulness of the construct of “revealed preferences” that could serve as a benchmark. Like a litmus test, BLE helps to make the contended nature of normative foundations (more) visible.

• 22 This is at least my impression in comparing the ratio of empirical papers presented at annual confe (...)

24As for research methods and the relation of theoretical and empirical approaches, the practice has changed quite significantly in the last decades but there does not seem to emerge any serious argument, let alone a consensus for rejecting abstract modelling. After all, non-maximising or non-selfish behaviour is just as amenable to formal models as the interactions of homines oeconomici (Posner, 1998). Rather, what is observable is that due to improvements in data and development in methods, BLE has brought more and at its best, more sophisticated, empirical (both experimental and field) research to Law and Economics. This brings about a change of disciplinary culture within Law and Economics, with increasing prestige of empirical research and more training in empirical methods in study programs.²²

• 23 BLE normally focuses on the “identification of a departure from unbounded rationality, willpower or (...)
• 24 Jolls calls the latter approach “debiasing through law”. See also Jolls and Sunstein 2006.

27BLE is sometimes seen as continuous with standard Law and Economics. Just like standard Law and Economics has a combination of enabling and interventionist implications (to wit, depending on transaction costs, it suggests “lubricating” or “mimicking” markets; Katz, 1998, 39-42), BLE responds to biases either by “taking people’s natural tendencies as given and shaping law around them” (Jolls, 2006, 18-19)²³ or by “attempting to reduce or eliminate such human tendencies through the legal structure” (Jolls, 2006, 19, emphasis in the original).²⁴ A related narrative is that BLE provides a corrective to but is continuous with standard (law and) economics. Empirical Legal Studies indicate a quantitatively but not qualitatively new phase in the maturation of Law and Economics into what Cooter (2011) somewhat imprecisely calls „normal science”: BLE does not overhaul but supplement standard Law and Economics. Not all economists consider these developments unproblematic. The vivid language quoted below suggests that the issue of continuity is highly relevant but commentators do not always find sober rational arguments sufficient or available.

• 25 Uskali Mäki (2003) makes a similar point about the notion of scientific imperialism. Thanks to Magd (...)

32This point is worth elaborating. I shall discuss it in terms of ‘economic imperialism’. The term imperialism, as used here, is not pejorative.²⁵ It simply refers to the expansion of the domain of economics, and rational choice theory more generally, beyond markets and other substantively economic phenomena. Law is such a non-market phenomenon, so Law and Economics is clearly a result of this imperialism of economics. Yet, the methodological underpinnings of this imperialistic view of economics are far from uniform.

• 26 Versions of this non-empirical arguments for rationality assumptions have been expressed by many ec (...)

37The second non-empirical reason for economic imperialism, based on the argument about “minimal rationality” as a logical necessity or a metaphysical assumption, has been itself formulated in different versions.²⁶ This view of the rationality assumption seems to follow from Popper’s philosophy of social science. Popper’s model of the social sciences was inspired by and explicitly refers to economic theory. Situational logic and the principle of rationality are building blocks of a model for the social sciences which is distinct from and somewhat in tension with Popper’s philosophy of the natural sciences (Hands 1985). Similar arguments have been formulated by Gary Becker (1976) in building a purportedly universal model for human behaviour, based on the three assumptions of rational maximisation, stable preferences and equilibrium. Philosophically minded interpreters of Becker such as Ingo Pies (1993) suggest that his version of economic imperialism comes close to apriorism, which has been often associated with Austrian economics and Mises in particular. In this Popperian–Beckerian–Misesian view, what economists call the maximization hypothesis, and philosophers call the rationality principle is “a nonfalsifiable, yet a useful and nontautological metaphysical presupposition of the neoclassical research program” (Hands, 1985, 93) in economics. If human action is assumed or defined as rational in an aprioristic way, then at least a certain minimal version of the rationality assumption is not subject to empirical testing. “Minimal rationality” is a non-empirical assumption in explanatory accounts for purposeful human action. This is required in economic explanations for logical reasons, in order to account for human action, rather than mere behaviour.

• 27 See note 14 above.

43McCloskey has been a powerful advocate of a rhetorical or humanistic understanding of economics. Here she argues that the behavioural economics of Kahneman and Tversky is deeply flawed and should not be seen as the way forward for economics. Yet Kahneman’s remark quoted above²⁷ suggests that he is open and affirmative about behavioural economics striving for practical impact by proposing policy reforms. These represent two different versions of crossing the positive/normative divide. This already takes us easily to the next topic, the impact of BLE on normative economics.

• 28 On relevant considerations about behavioural assumptions in institutional design see Kornhauser (20 (...)

46While empirical research is important for reasonable and effective regulatory design, it is just one ingredient. Empirical knowledge is not decisive on the normative question whether institutional designers or policymakers and those responsible for legal design at various levels of decision-making should base their models, recommendations and regulatory instruments on ‘true’ assumptions about human behaviour or rather on ‘false’, that is counterfactual assumptions. Institutional design also requires non-empirical criteria to determine the ‘model’ of human behaviour (a characterisation of the agents populating the institution in terms of their motives, capabilities, etc.) underlying (supporting) various policies. In this respect, the entire exercise is inevitably, though not exclusively, value-driven.²⁸

• 29 Behavioural economists, even when reflecting on difficulties in linking empirical findings and poli (...)

48The obvious first candidate for a normative benchmark would be revealed preferences.²⁹ Yet one of the general empirical findings of behavioural analysis is exactly the inadequacy of revealed preferences as the foundation of normative analysis. If preferences are, in contrast to standard assumptions of economic theory (Becker, 1976, chapter 1), empirically unstable, radically context-dependent and subject to framing effects and manipulation, how could they serve as a normative benchmark? Alternatively, should perhaps preferences not only be ‘laundered’ (Goodin, 1986, 73), but corrected by policymakers?

• 30 I used this term here, inspired by but more loosely than Rawls (1987).

51In this light, the main impact of the growth of BLE on normative economics seems to be the following. Economists increasingly come to realise that the domain of overlapping consensus³⁰ between what autonomy requires and what welfare suggests is much smaller than previously thought. Under the assumption of individual rational choice, it is likely that what individuals would freely choose and what is in their interest largely overlap. This is the basis for economists’ notorious insistence on “consumer sovereignty” (Persky, 1993). If, however, as a matter of fact, individuals cannot be generally considered the best judges of their own interest, the right course of collective action, even for a normative individualist, becomes harder to determine. From this point, theorists have moved in different directions. Hence another litmus test effect of BLE.

• 31 To be sure, historically there have been many connections between these two versions of normative e (...)
• 32 On one version of this opposition see Marciano (2007).

53The opposition can be seen in various forms: between normative views focusing on outcome and those focusing on process; between “American” and “European” style of Law and Economics (Posner 2003, chapter 7), the first being interested in micro or middle theorizing, the second in more systemic questions;³¹ or finally and perhaps most importantly, the opposition between, one the one hand, a utilitarian or Benthamite version of Law and Economics committed to or at least not in principle objecting to grand-scale social planning and engineering in a rationalist manner; and a libertarian version, more in line with Hayek, Austrian economics or contractarianism.³²

• 33 For attempts to rephrase paternalism in terms of merit goods, see Cserne (2012, 36-37).

54As a welfarist project, normative economics is not categorically opposed to collective decisions overriding individual choice if this is expected to bring about good consequences in terms of some metric of welfare. Overriding individual choice is, however, not a principle but (subject to transaction costs) a piecemeal decision. It should be noted that in some areas of normative economics such as public finance, similar ideas have been present well before the current wave of behavioural economics, e.g. under the terminology of merit goods, originally coined by Richard Musgrave.³³ For instance, in an early paper, mainly based on common sense observations about consumer behaviour, John Head (1966) argued for the public provision of certain goods in terms of behavioural patterns of consumers.

• 34 On further economic models of freedom of choice see Dowding and van Hees (2009).

55Others concluded that if preferences are incoherent or indeterminate, and welfarism based on some objective metric of well-being is paternalistic, then welfare economics should adopt new normative standards which are likely to be radically different from those familiar from a long tradition of consequentialist and welfarist reasoning in normative economics. Robert Sugden’s work is a good example of this approach. Already in the 1980s Sugden has sharply criticized the general idea of economists whispering advice in the ear of governments how to govern society. Around the same time, he has got involved in empirical research as well as theoretical modelling of bounded rationality. In the next stage and explicitly as a response to empirical findings about the instability of preferences, he has argued that unstable or incoherent preferences do not in themselves justify paternalism and suggested the so-called opportunity criterion as a normative measure for economics (Sugden 1998 and 2007).³⁴

• 35 This is, after all, what Tor (2008, 314-325) meant by “normative neutrality” and “normative relevan (...)

56Of course, welfare and autonomy do not exhaust the domain of values actually or potentially guiding normative economics. The growth of BLE may also raise awareness of this plurality of values. In principle behavioural findings can be combined with normative foundations altogether different from the currently dominant ones.³⁵