اقتصاد رفتاری دارای دو «روح» : آیا آنها هر دو ، بخشی از عقلانیت اقتصادی هستند ؟

In this paper, I argue that behavioral economics, far from being a monolithic theory, consists of two different ‘souls’ and that a fundamental asymmetry exists between them, with regard to the nature of the departures from the economics’ ‘canonical model’ they focus on. While a first class of departures deals with the major cognitive limitations and systematic biases in decision-making affecting economic behavior, a second research area investigates deviations from the classic assumption that economic agents are systematically driven by the pursuit of material self-interest. Even though on methodological grounds the two research areas share a broadly inductive approach – as they both aim at incorporating the major results obtained through several empirical methods (in particular, via experimental work) into formal economic analysis –, I claim that rather different methodological conclusions and regulatory and policy implications can be drawn, depending on the cognitive or social nature of the departures from the standard economic model under study.

As far as recent developments in economics are concerned, ‘Behavioral Economics’ (BE) represents one of the most interesting and promising research frontiers in the last decades. BE aims at establishing a new disciplinary field grounded on methodologically realistic roots, as its major goal is to increase the explanatory and predictive power of economic analysis through an increase in the degree of realism of the assumptions at the basis of economic models. As Camerer and Loewenstein (2003) point out: “behavioral economics increases the explanatory power of economics by providing it with more realistic psychological foundations. (…) We share the positivist view that the ultimate test of a theory is the accuracy of its predictions. But we also believe that, ceteris paribus, better predictions are likely to result from theories with more realistic assumptions” (p. 3). This objective has been pursued by means of a greater and greater integration between economics and psychology (see also Rabin, 2002), but also, more recently, between economics and social sciences such as, for example, cultural anthropology (see on this Henrich et al., 2002 and Henrich et al., 2004). While the core theory traditionally used in economics assumes that agents (i) are driven by well-defined, stable and self-centered preferences, (ii) use all the information available and (iii) act in order to maximize their objective functions, psychology reveals that it is often the case that such assumptions cannot be taken as an adequate and satisfactory description of human behavior. In this light, BE has been investigating both real and experimental environments where some economic agents involved make choices that appear to substantially deviate from traditional economic predictions. Hence, it deems it crucial and urgent to incorporate the major psychologically robust findings in the set of economic models’ assumptions. Such methodological stance is taken explicitly by prominent behavioral economists such as Matthew Rabin: “Ceteris paribus, the more realistic our assumptions about economic actors, the better our economics. Hence, economists should aspire to make our assumptions about humans as psychologically realistic as possible” (Rabin, 2002, p. 658). In this paper, I argue that BE, far from being a monolithic theory, (a) consists of two different ‘souls’ and that (b) a fundamental asymmetry exists between them, with regard to the nature of the departures from the economics’ ‘canonical model’ they focus on. While a first class of departures deals with the major cognitive limitations and systematic biases in decision-making affecting economic behavior (let us shortly refer to this component of BE as to the ‘cognitive component’, CC), a second research area investigates deviations from the classic assumption that economic agents are systematically driven by the pursuit of material self-interest (let us call this second component of BE the ‘social component’, SC).1 Even though on methodological grounds the two research areas within BE share a broadly inductive approach, as they both aim at incorporating the major results obtained through several empirical methods (in particular, via experimental work) into formal economic analysis, I claim that rather different methodological conclusions and regulatory and policy implications can be drawn, depending on the cognitive or social nature of the departures from the standard economic model under study. More specifically, the aforementioned ‘fundamental asymmetry’ between CC and SC within BE can be illustrated as follows. Both kinds of findings can be seen as systematic deviations from the previously received theory of rational choice. However, within the first area (CC) it is assumed, analogously to neoclassical models, that agents act according to the self-interest hypothesis, but also that they are affected by various forms of cognitive limitations and behavioral biases occurring at some stage of the decision-making process (in line with classical contributions on bounded rationality; see, in particular, Simon, 1955): several papers within CC explore non-standard phenomena such as framing effects, self-control problems, overconfidence, status quo bias and projection bias, to name only the major ones (for a comprehensive review, see Della Vigna, 2009). The second area (SC) departs from the neoclassical approach by taking a different road. Unlike traditional analyses, these studies have been focusing, so far, at both an empirical (mainly experimental) and a theoretical level, on interaction scenarios where players (i) show a concern for the welfare of others and (ii) are modeled as if they were not driven by classic selfish preferences, but by (different types of) social preferences (see e.g. Charness and Rabin, 2002 and Camerer, 2003). In particular, behavioral choices by individual agents (as well as their objective functions) can (either positively or negatively) be affected by other players’ preferences, material well-being, intentions and/or behavioral choices. In this light, the asymmetry thesis can be expressed as follows: most of the biases and heuristics on individual decision-making investigated within CC can be viewed as departures from the notion of economic rationality; by contrast, as far as SC is concerned, I argue that, in itself, the presence of either social preferences or other forms of sociality (as I make clear in Sections 2 and 3) does not make the subsequent behavior necessarily less rational.2 Despite the still relevant influence exerted by the so called canonical model of homo oeconomicus, there is today a lot of agreement in economics about the fact that humans (including economic agents) do not systematically behave in the pursuit of their own material self-interest. BE has undeniably played a key role in generating such widespread consensus on the need to depart from the so called ‘selfishness axiom’3 and consider other-regarding behaviors as part of the subject matter of economics. However, on more constructive grounds, it is still unclear what the actual driving forces behind human sociality are, with specific reference to economically relevant domains. In this regard, in the last decades several different proposals have been advanced in the economics literature in order to shed light on various social phenomena where motivational factors other than material self-interest seem to be critically at work. While, as anticipated above, BE's proposal has been to formalize sociality via different models of ‘social preferences’ (SC), other significant approaches have been advanced in the economics literature in the last decades, in order to account for various forms of sociality. In the next two sections, I maintain that in either case many socially oriented behaviors need not be viewed as departures from economic rationality. In particular, I show that modelling sociality via social preferences (SC) is compatible with the classic notion of rationality as ‘internal consistency’ (Section 2). Further, I refer to some interesting non-preference-based theories of sociality where the other-regarding behaviors under study are compatible with suitably defined notions of rationality, though not with a standard self-centered utility-maximizing framework (Section 3).

The fundamental asymmetry within BE can be stated as follows: while CC focuses (both theoretically and empirically) on departures from the standard economic rationality which in several cases can be characterized as ‘errors’ and make the behavior of the agents under study (at least to some extent) ‘irrational’, SC does not imply that other-regarding agents are necessarily irrational. By contrast, this research area within BE, together with the non-preference-based accounts of sociality reviewed in Section 3, calls for the integration of various forms of non-selfish rationality into the analytical framework of BE. As a consequence of this asymmetry between the two ‘souls’ of BE, I argue that also arguments in favor of regulatory and paternalistic policies ought to be – asymmetrically – mainly referred to phenomena investigated by CC, rather than SC. The point is that an increasing body of recent work has been discussing the implications for normative analysis of the findings of behavioral economics (see e.g. Kahneman et al., 1997, Camerer et al., 2003 and Thaler and Sunstein, 2003). In particular, interesting arguments have been advanced by Camerer et al. (2003) in favor of a soft version of paternalism that they call ‘asymmetric paternalism’. My point is that such arguments have to be mainly related to CC, not to SC. Camerer et al. (2003) believe that paternalism (that is “forcing, or preventing, choices for the individual's own good”) can be justified when there is “fear that even people of sound mind might not act in their best interest in certain predictable situations”15 (p. 1211). I claim that if a person is, say, inequity averse, driven by intention-based reciprocity or by other forms of (possibly non-preference-based) morality, then this has to do with how she implicitly defines her own best interests. Hence, insofar as BE's definition of ‘rationality’ does not get violated, there is no room for paternalism. By contrast, Camerer et al.’s considerations have to do with a lot of phenomena occurring within CC, as they refer to “a list of common decision-making errors that even highly competent, well-functioning people make in predictable situations” (p. 1214). In other words, I argue that the fundamental asymmetry outlined in this work entails that a crucial distinction has to be made with regard to the policy implications that behavioral findings bring about: it is mainly in the domain of CC, rather than with regard to BE as a whole, that paternalistic policies can be usefully developed. Hence, any argument in favor of soft paternalism has to be referred to one strand of literature only (though a definitely relevant one), within BE (that is CC), but not to the other one (SC).