From Bounded Rationality to Behavioral Economics

From Bounded Rationality to Behavioral Economics

• The motivations for the 2002 award of the Nobel Prize in Economic Sciences to Daniel Kahneman and Vernon L. Smith sheds light on a hidden on-going revolution that is challenging both the epistemological assumptions of neoclassical economics and the foundations of the decision making theory.

• At the beginning of the ceremony at the Stockholm Concert Hall, Professor Lars-Göran Nilsson delivered a Presentation Speech in which he explicitly recognized the success of the new emerging approach that challenges the traditional foundations of economic analysis:

• “Your Majesties, Your Royal Highnesses, Ladies and Gentlemen,

• Economic theory relies on the assumption that economic agents may be likened to a "homo oeconomicus".

From Bounded Rationality to Behavioral Economics

• The motivations for the 2002 award of the Nobel Prize in Economic Sciences to Daniel Kahneman and Vernon L. Smith sheds light on a hidden on-going revolution that is challenging both the epistemological assumptions of neoclassical economics and the foundations of the decision making theory.

• At the beginning of the ceremony at the Stockholm Concert Hall, Professor Lars-Göran Nilsson delivered a Presentation Speech in which he explicitly recognized the success of the new emerging approach that challenges the traditional foundations of economic analysis:

• “Your Majesties, Your Royal Highnesses, Ladies and Gentlemen,

• Economic theory relies on the assumption that economic agents may be likened to a "homo oeconomicus".

From Bounded Rationality to Behavioral Economics

• “This fictitious individual is usually governed by self-interest and makes his economic decisions by rationally evaluating the consequences of different alternatives, even in complex situations where the outcome is difficult to predict. Despite such strong assumptions, this approach has proved to be highly rewarding and has enhanced our understanding of many economic phenomena.

• “Empirical testing of postulates in economic theory confronts theoretical predictions with findings from real-world markets and economies. In general, however, since "field data" are affected by factors which scarcely allow for control and measurement, the identification of causal relationships is problematic.

From Bounded Rationality to Behavioral Economics

• Whereas economists have had to overcome such obstacles by using ingenious statistical methods, many natural scientists have been able to rely on controlled experiments to test their theories.

• These common descriptions of theoretical and empirical economic science may well have historical validity. But nowadays, they both have to be modified. With increasing confidence, researchers in psychological economics have been able to demonstrate that in some situations, individuals do not behave like "homo oeconomicus".

• “Researchers in experimental economics have developed methods for controlled laboratory experiments in economics. A number of scholars have contributed to this development, including previous Laureates: Maurice Allais and Herbert Simon thus brought psychological perspectives into decision theory, while John Nash and Reinhard Selten conducted early experimental studies. But this year's Laureates are the key figures in these two fields.

From Bounded Rationality to Behavioral Economics

• Professor Kahneman: Insights from cognitive psychology have been instrumental in establishing new theoretical and empirical results; in ongoing research, they guide thought-provoking attempts to reformulate many aspects of economic and financial theory. The new bridges across disciplines can largely be attributed to your innovative research on the boundary between economics and psychology.

• Professor Smith: Economics used to be regarded as a non-experimental science. This is no longer so; nowadays, economic experiments are routinely conducted in specialized labs all over the world. Scale or ethics may limit economic experimentation, but the methods you have developed continue to enrich our empirical toolbox. Both of you have laid the foundations for an exciting renewal of economic research. It is a great honor for me to express, on behalf of the Royal Academy of Sciences, our warmest congratulations. I now ask you to stop forward and receive your Prizes from His Majesty, the King. ”

From Bounded Rationality to Behavioral Economics

• Although in an apodictical style, Professor Lars-Göran Nilsson claims that the revision of the traditional theory of rationality has reached a critical juncture where its inspiring principles and its status as the microfoundations of economic analysis are being seriously called into question.

• He emphasizes that experimental results were already questioning the validity of the standard model of rational action in the 1950s: Allais's paradox of 1952 and the empirical study of decision processes in firms conducted by Cyert, Simon and Trow in 1956.

• Therefore the challenge to the theory of rational behavior initiated at the height of his success, culminated with the publication of Theory of Games and Economic Behavior by von Neumann and Morghenstern in 1944.

From Bounded Rationality to Behavioral Economics

• To understand the origins of the changes presently in progress, we must return to the 'golden age' of standard rationality theory: the 1950s. The successes achieved by linear and dynamic programming in those years seemed to justify unlimited faith in the ability of optimization models to explain all economically significant forms of behavior.

• There was a widespread conviction that it was invariably possible and justifiable to reduce macrophenomena to rational forms of behavior and to represent rational forms of behavior as problems of constrained maximization.

From Bounded Rationality to Behavioral Economics

• Increasing complexity

• Yet as models of rational decision-making became increasingly efficient, so there was a corresponding growth in the complexity of the calculation required to operate them and in the refinement of the algorithms required to resolve them. This raised the problem of whether it was legitimate to ascribe individuals with the ability to perform extremely complex decision-making processes, resolving the problems connected with them by means of highly time-consuming and refined algorithms, or whether the models of rational behavior should only be interpreted in a normative sense as techniques aiding decision-making and suitable for use by experts, not by common decision-makers.

• It was this dilemma that prompted Simon to advance his hypothesis of bounded rationality and to launch his polemic against the idea of perfect and all-encompassing rationality.

From Bounded Rationality to Behavioral Economics

• A different solution

• However, in those years a different solution of the dilemma was proposed by Milton Friedman , a solution that was very successful, providing a (fallacious) point of reference for mainstream theory.

• According to Friedman, although individuals do not possess the formal tools with which to calculate the optimum adequately, they behaved as if they do - like bicycle riders who keep themselves in dynamic equilibrium even though they are unaware of the complex equations of the dynamics of motion, or billiard players who accomplish complex trajectories with their billard balls although ignorant of the laws of rational mechanics.

From Bounded Rationality to Behavioral Economics

• Friedman, and the Chicago school of which he was a leading exponent, supplemented the 'as if' hypothesis with the further assertion that economic facts at the macro-micro level, i.e. individual preferences, were not observable, and indeed that they were irrelevant to proof of the validity of an economic theory (Friedman 1953).

• In order to explain why this position enjoyed such prolonged success, even though it was substantially misleading, we must examine a bit more carefully the terms of the problem.

From Bounded Rationality to Behavioral Economics

• Rationality as computing activity

• In its standard version, the theory of rationality rests on the following conception of human behavior:

• there exists a set of conceivable actions which every individual may undertake and which lead to certain consequences. Individuals possess a mental order of preferences concerning all the possible consequences of their actions. They evaluate these consequences, and , given the constraints, decide upon a particular action.

• They therefore make their choice coherently with their preferences and with the constraints upon them. The choice is therefore the outcome of a rational computing activity, and it matters how complex the calculations required for rationality is.

From Bounded Rationality to Behavioral Economics

• Between Normative and Positive Economics

• According to these assumptions the theory of rational choice suggests / prescribes to economic actors the best way to achieve their goals. To weaken the normative character, it is assumed (Milton Friedman) that those who failed to conform would be gradually excluded by a process of selection which permitted only ‘rational’ operators to survive. This leads to a so-called “positive” economic theory.

• In this framework it is therefore useless and uninteresting to investigate the psychological aspects involved in decision-making, because at most they could only aid explanation as to why certain individuals are unable to behave in an entirely rational manner.

• The normative approach to decision making was limpidly expounded in Lionel Robbins’ Essay on the Nature and Significance of Economics (1932), in which he defined economics as the science of choice.

From Bounded Rationality to Behavioral Economics

• In this approach, “calculation” is therefore totally independent from individual mental activities and it is even irrespective of the mental processes of single individuals. The role of economy and rationality is viewed as being strongly normative. This view has been shared by the vast majority of economists for just less than a century, as illustrated by the systematic and exhaustive work of Lionel Robbins in the well-known “Essay on the Nature and Significance of Economic Science“ (1932).

• His study illustrates some key points in the economy statute and in the role of the utility theory after a decade of critical debate – often fervent – between economists of different schools. Robbins codifies a “post-Austrian” vision of the economic sciences statute in which economy and psychology must be considered fully autonomous specializations with equally independent statutes.

From Bounded Rationality to Behavioral Economics

• As reminded by Schumpeter in his “History of Economic Analysis”, the separation of economy from psychology, was ruled only after a decade or so of heated debates:

• “In principle, utility, be it total or marginal, was considered a psychic reality, a sensation that became evident from introspection, independent of any external observation […] with directly measurable proportions. I believe this was Menger and Böhm-Bawerk’s opinion” (Schumpeter, History of Economic Analysis).

• From the very beginning the resort to a psychological interpretation of “utility” was quite evident:

•  

From Bounded Rationality to Behavioral Economics

•  

• “Ferdinando Galiani (Della moneta, 1750) defined utilità as “the power of a thing to procure us felicity.” Similarly, Jeremy Bentham at first spoke of utility as “that property in any object, whereby it tends to produce benefit, advantage, pleasure, good or happiness” (An Introduction to the Principles of Morals and Legislation, 1780). But the meaning of the term has shifted continuously and even today “utility” circulates with various, albeit cognate, connotations. By referring to the principle of utility as the principle of the greatest happiness of the greatest number, Bentham himself paved the way for this terminological license. The ensuing confusion prompted W. Stanley Jevons to insist that “Utility is not an Intrinsic Quality,” but “the sum of the pleasure created and the pain prevented” .(The Theory of Political Economy, 1871).”

• (N. Georgescu Roegen, Dictionary of the history of economic ideas)

From Bounded Rationality to Behavioral Economics

• It is mostly the contrast between the French and Austrian schools that fuelled the long evolution of the notion of utility until it eventually culminates into the codified version of Robbins, Hicks and Allen.

• In few decades the conceptual links between the psychological aspects of Utility theory and the logic of decision making were broken . Pareto gave the proof that one did not need to resort to the utility function, but could resort to the simplest notion of preference: the winning analytical strategy consisted, in fact, in establishing some simple properties of preferences, completeness, transitivity, continuity and independence. These properties enabled the construction of an axiomatic model of the choice.

• Therefore a confused and heated debate on the measurability of utility dropped out rapidly , as a consequence of From the very beginning, assuming the existence of the utility function to measure the benefits stemming from a particular choice appeared unattainable.

From Bounded Rationality to Behavioral Economics

• Slowly progression led to demonstrate that it was possible to adopt a utility function that was perfectly equivalent to the axiomatic model. In the 50s there were a series of incredible disputes and much confusion with regards to this issue, as illustrated by Schumpeter in his History of Economic Analysis. (1968). But at that time the Pareto’s approach was consolidated, and most economists shared the opinion that the theory

•  

• “…. has a much better claim to being called a logic of choice than a psychology of value”( Schumpeter, History of Economic Analysis).

•  

• In the Theory of Games and Economic Behavior, published in 1944, von Neumann e Morghenstern make a further step in this direction; they provide an axiomatic approach to the theory of decision making in condition of uncertainty by formalizing the Expected Utility hypothesis, two centuries after Bernoulli original definition.

From Bounded Rationality to Behavioral Economics

• Slowly progression led to demonstrate that it was possible to adopt a utility function that could be used that was perfectly equivalent to the axiomatic model. In the 50s there were a series of incredible disputes and much confusion with regards to this issue, as illustrated by Schumpeter in his History of Economic Analysis. (Schumpeter, 1968).

• In the Theory of Games and Economic Behaviour ,published in 1944, von Neumann e Morghenstern provide an axiomatic approach to the theory of decision making and fully formalized the Expected Utility hypothesis, two centuries after Bernoulli original definition.

From Bounded Rationality to Behavioral Economics

• Expected utility is nothing but a function of utility applied to uncertain events, which features a particular property that can be illustrated in the following manner: faced with two different uncertain alternatives, x and y, which can occur with p and 1-p probability, a subject will compare the utility of each of the two options by “weighing them” in relation to respective probabilities: thus the expected utility will be: U= p U(x) + (1-p) U(y) .

• Though apparently innocuous, the assumption that in conditions of uncertainty individuals decide on the basis of expected utility, contains further restrictions compared to those relating to the utility functions in deterministic conditions, as noted by Allais.

From Bounded Rationality to Behavioral Economics

• Risk Aversion, Neutrality and Proclivity

• The concept of univariate "risk aversion" intuitively implies that when facing choices with comparable returns, agents tend choose the less-risky alternative. The formal construction of this concept, based on von Neumann Morghenstern expected utility approach is largely due to Milton Friedman and Leonard J. Savage (1948).

From Bounded Rationality to Behavioral Economics

From Bounded Rationality to Behavioral Economics

From Bounded Rationality to Behavioral Economics

• Friedman and Savage’s study dates back to 1952. Here the authors try to construct an expected utility curve that represent, reasonably well, observable behaviour at the aggregate level.

• The method imposes appropriate restrictions to the expected utility function based on characteristics that relate to the behaviour of vast aggregates of individuals: for instance noticing that a large number of citizens with middle-low income risks small sums of money on the lottery and on gambling, implies there is a tendency to risk, and at the same time the fact that those same citizens insure themselves, means they are risk averse: to be compatible with these two “generic” characteristics the expected utility curve will need to have an “S” shape for the values of middle-low income.

From Bounded Rationality to Behavioral Economics

From Bounded Rationality to Behavioral Economics

• Friedman thus constructs a general shape of the curve, without testing the characteristics on a real population. In fact precise data on insurance were not taken into consideration and data relating to the income of gamblers do not even exist. With this approach any attempt to directly verify the validity of the utility hypothesis was neglected, on the contrary it was claimed to be misleading and thus was also useless to move in such direction since assumptions, at the individual level, were totally unrealistic (Friedman (1996), p. 105).

• With regards to the way in which the notion of utility was defined during that period, Friedman and Savage’s report proved to be a considerable step ahead. Yet this advancement concealed an untenable general methodological and epistemological approach that was unfortunately successful and remained for an extended period an unquestionable dogma for a vast number of economists.

From Bounded Rationality to Behavioral Economics

• Maurice Allais’s research proposed a counter trend to Friedman e Savage’s approach. He carried out experiments on individual preferences that supplied experimental proof of some systematic failures of the theory’s prediction.

• In 1952, at a symposium held in Paris, Allais presented two studies in which he criticized the descriptive and predictive power of the choice theory of the “American School” and especially Friedman’s stance, (Allais, 1953), demonstrating some experiments in which subjects underwent alternative choices in conditions of risk systematically violating the assumptions of the expected utility theory.

• Its investigation methodology overturns the prescriptions set by the Chicago school, as it is founded on observation of an individual’s behaviour, and introduces a method of experimentation with which the inherent difficulty of direct observation of individual preferences can be overcome by cross checking alternative choices.

From Bounded Rationality to Behavioral Economics

• The experiments proposed by Allais are founded on two characteristic points:

• first, the properties of the choice that characterize the expected utility function must be identified in axiomatic form; the properties are: completeness, transitivity, continuity and independence.

• Second, subjects are presented with pairs of binary choices selected in such a way that one combination of the answers involves the violation of at least one of the axioms.

From Bounded Rationality to Behavioral Economics

• Allais ‘ critique to “Neo-Bernoullian” theory of choice: an experiment refuting Savage’s principle of independence

• Do you prefer Situation A to Situation B?

• Situation A

• Certainty of receiving 100 million

• Situation B:

• a 10% chance of winning 500 million,

• an 89% chance of winning 100 million,

• A 1% chance of winning nothing

• Do you prefer Situation C to Situation D?

• Situation C

• a 11% chance of winning 100 million,

• An 89% chance of winning nothing,

• Situation D:

• a 10% chance of winning 500 million,

• An 90% chance of winning nothing

From Bounded Rationality to Behavioral Economics

• If Savage’s postulate were justified, the preference AB should entail

• C  D

• “What one finds, however, is that the pattern for most highly prudent persons, the curvature of whose satisfaction curves is not very marked, and of who are considered generally as rational, is the pairing A  B and CD. Thus contradicts the Savage’s fifth axiom.” (Allais, The foundations of a positive theory of choice involving risk and a criticism of the postulates and axioms of the American school 1952)

• Proof:If AB, then U(100)>0.10U(500)+0.89U(100)+0.01U(0)

• Rearranging this expression gives

• 0.11 U(100)> 0.10U(500)+0.01U(0);

• And adding 0.89 U(0) to each side yelds

• 0.11 U(100)+ 0.89 U(0) > 0.10U(500)+0.90U(0)

• That means CD!

From Bounded Rationality to Behavioral Economics

• The 80s: the revision of paradigm

• The most natural route to move away from the impasse generated by the experiments of Allais was to consider the theory of expected utility too restricted and therefore to try to formulate an extended theory of expected utility.

• The new theory should be created by relaxing or slightly modifying the original axioms. Many proposals arose to review such approach, especially from the mid 70s onwards. We will cite the following: the Weighted Utility Theory (Chew and MacCrimmon) assumes a weaker form of the axiom of independence; the Regret Theory proposed by Loomes and Sugden (1982), and the Disappointment Theory, suggested by Gul (1991) are further examples along this track.

• See Hey J. D. (1991), Experiments in Economics, Oxford, Basil Blackwell

From Bounded Rationality to Behavioral Economics

• Kahnemann and Tversky’s approach was completely different and diverged in a crucial way from previous proposals as it was not limited to modifying certain axioms but rather, it restructured the problem, making reference to the mental processes involved.

• This theory also coherently fits within the analytical frame of Bounded Rationality and with the spirit of Simon, as the two authors explicitly acknowledge.

• In “The Framing of Decisions and the Psychology of Choice”, Tversky and Kahnemann analyse the process of choice in conditions of uncertainty as something that has two distinct aspects:

• the editing of events into a mental representation

• an evaluation phase.

From Bounded Rationality to Behavioral Economics

• Problem 1

• Assume to be 300 $ richer than you are today. Choose between

• A the certainty of earning 100$

• B 50% probability of winning 200$ and 50% of not winning anything

• Problem 2

• Assume you are 500 $ richer than today. Choose between

• - C A sure loss of 100$

• - D 50% chance of not losing anything and 50% chance of losing 200$

• Readers responding to the two problems will probably opt for the adverse risk option in problem 1 therefore choosing an earning that is certain (answer A). This occurred for the vast majority who participated in the experiment. Instead in problem 2 the answer is probably the one in favour of risking, thus answer B.

From Bounded Rationality to Behavioral Economics

• The majority, who picked answer A and D, violated the theory of expected utility (the independence axiom of the theory), as with Allais’ experiments.

• The two problems, in terms of expected utility are the same problem; in fact, the entity’s available wealth was considered after the choice had been made:

• Problem 1

• Case A 400 with prob=1

• Case B 300 with prob=0.5 or 500 with prob=0.5

• Problem 2

• Case C 400 with prob=1

• Case D 300 with prob=0.5 or 500 with prob=0.5

From Bounded Rationality to Behavioral Economics

• Therefore a large majority of individuals behave as risk takers when facing a problem presented in terms of loss (Problem 2) while they behave as risk averse when the same problem is presented in terms of gain (Problem 1). This behavioral inconsistency is called “framing effect”, and shows clearly that the mental representation (framing) of a problem may be crucial to elicit individual behavior.

• Kahnemann and Tversky observe that

• “the path of preferences observed in the two problems are of particular interest as they violate not only the theory of expected utility, but practically all choice models based on other normative theories. It must be noted that these data are not consistent with the “Regret” model proposed by Bell (1982) and Loomes and Sugden (1982) and axiomazed by Fishburn (1982)”

From Bounded Rationality to Behavioral Economics

• They suggest that in order to understand a decision one must thoroughly analyze the cognitive processes that are at the base of the decision. It is thus necessary to understand how people represent problems, how the complex process of editing is carried out and how construction of mental models is built in order to make a particular decision.

• A suggestion that is closely related to Simon’s framework of the decisional problem is: in his view the decision is the final act of a problem-solving process. The latter constitutes the heart of a subject’s activities to face a decision, and who has to make a “rational calculation” to find out which is the best action strategy. This “calculation” though is carried out with strong restrictions due to the cognitive limits of an individual (bounded rationality); these restrictions can be generated by some systematic biases, that is, systematic deviations from what would be the results obtained by a “hyper rational” subject, an omniscient calculator of unlimited power

From Bounded Rationality to Behavioral Economics

• From computing activity to building Mental models

• Simon conducted pioneering analysis of strategic situations of this kind, in which individuals are not able to explore all the consequences of their actions and hence cannot assess these consequences with precision. The limit on calculation is in a certain sense an insuperable computational obstacle - a case in point being the game of chess, where the time required to calculate a complete strategy exceeds that of the human life-span. Even apart from such extreme cases, though, one immediately realizes that, in the majority of economic situations, decisions are taken in the presence of a time constraint. This brings with it a important implication: if economic subjects are unable to explore all the consequences of their actions, they are likewise unable to assess them; and this introduces an intrinsically uncertain element into human action.

From Bounded Rationality to Behavioral Economics

• In 1956, in a path-breaking article which constituted a first stage in construction of the microfoundations of economic behavior within organizations, Cyert, Simon and Trow carried out an empirical analysis of managerial decisions which revealed an evident "dualism" of behavior: on the one hand, a coherent choice among alternatives; on the other, a search for the knowledge required to make the choice. For several months, an observer monitored the decisions made by the executives in a particular firm and recorded a number of features entirely at odds with the prescriptions of the decision theory then current.

• The principal finding was that, when decisions are made in conditions of high uncertainty - conditions, that is, poorly structured in terms of a hierarchy and a set of rules - their outcomes are imperfectly assessed. This activates a search process intended to frame all the elements involved in the decision.

From Bounded Rationality to Behavioral Economics

• The central problem in these situations is the search of relevant knowledge. Hence, not only must subjects gather information, they must also be able to select the information that is effectively relevant to their purposes and assimilate it into the system of knowledge that they already possess. To do so, they must have a "level of competence" adequate to the situation of choice; they must, that is, implement skills of learning and problem solving which are entirely different from (and much more complex than) the "perfect" rationality presumed by traditional theory.

• The core of the decision-making process is therefore the activity of search and learning that furnishes actors with the information and knowledge they require to achieve their goals.

• Choice is only the final stage of the decision-making process and it is of secondary importance. The most important part of the process is the ability of the subjects to formulate and solve problems.

From Bounded Rationality to Behavioral Economics

• This distinction gives rise to a dichotomy between types of decision, which Cyert, Simon and Trow sharply distinguish into programmed and non-programmed:

• "Decisions in organizations vary widely with respect to the extent to which the decision-making process is programmed. At one extreme we have repetitive, well defined problems (e.g., quality control or production lot-size problems) involving tangible considerations , to which the economic models that call for finding the best among a set of pre-extablished alternatives can be applied rather literally. In contrast to these highly programmed and usually rather detailed decisions are problems of non-repetitive sort , often involving basic long-range questions about the whole strategy of the firm or some part if it, arising initially in a highly unstructured form and requiring a great deal of the kinds of search processes listed above." (1956, p.238)

From Bounded Rationality to Behavioral Economics

• Bounded rationality, unprogrammed decisions and learning are the key aspects of human behavior in organizations under ill-defined conditions.

• This new vision leads March and Simon to completely redefine the description and analysis of “planning”: planning is no longer a static and mechanic activity based on rational decisions immersed in a world of complete information. Planning is now based on “organizational learning”. Search therefore becomes a key activity in organization, as well as being a resource that can be differently improved within different organization, giving rise to differentiation in organizational performances. Adaptation is the crucial element that may generate differentiation and sub-optimalities.

From Bounded Rationality to Behavioral Economics

• Note that an important part of Simon views on planning seems a re-elaboration of Hayek’s ideas on knowledge and “Sensory order”, applied to the working of organizations. Learning and problem solving are considered the core features of organizations, through search processes. The idea that learning consists in changes in the classifications of the relevant phenomena has been confirmed and further developed by recent experimental results which have shown that the creation of mental categories to represent a problem is the basic driver for the generation of a division of problems and therefore the source of the correlated biases in decision-making (Egidi, 2002).

• Hence, a relevant conceptual improvement is that not only do organizations learn, but they make errors during this process, and – as March’s behavioral description shows – since adaptation may easily lead to sub-optimal organizational configurations, errors may be systematic and stable in the long run.

From Bounded Rationality to Behavioral Economics

• Routines and the evolutionary approach: the origins

• "New business procedures would then be analogous to new mutations in nature. Of a number of procedures, none of which can be shown either at the time or subsequently to be truly rational, some may supplant others because they do in fact lead to better results. Thus while they may have originated by accident, it would not be by accident that they are still used. For this reason, if an economist finds a procedure widely established in fact, he ought to regard it with more respect than he would be inclined to give in the light of his own analytic method."

• (Roy F. Harrod, 1939, Oxford EP)

From Bounded Rationality to Behavioral Economics

• Routines and the evolutionary approach: a formal representation

• Within the theory of computation tradition considers "routines" to be synonymous with "programs" , i.e. list of instruction in a (artificial) language. Turing wrote his celebrated article on "Computable Numbers", proposing his computing machine, in the same year - 1936 - in which Alonzo Church was completing his work on Lambda Calculus, and it soon became clear that the two systems were equivalent.

• Therefore a consolidated tradition exists which considers routines to be computable programs, which can be represented by a set of condition-action rules.

From Bounded Rationality to Behavioral Economics

• Routines and the evolutionary approach: a formal representation

• 1 A set of states or configurations of the problem X=x1, x2,….. xn, and two sub-sets P X and G X . P and G are respectively the set of starting states and the set of goals. PG=

• 2 A set of actions: M=m1, m2, ....mn

• 3 A set of rules defined by a transition array A(i,j).

• A(i,j)=mw means that by applying to the state xi the move mw you get the new state xj . xj is called successor of xi ,which in turn is called the predecessor of xj .

• A Routine is therefore a program, i.e. set of states/actions allowing to achieve the goal.

From Bounded Rationality to Behavioral Economics

• The NK model has the following main features: the evolution of an organism, or in general of a complex biological system, is guided by its ‘fitness’, that is, by its reproductive success in the environment; the characteristics that determine the fitness of an organism can be represented in a discrete space because they are a list of ‘traits’ that can assume different values.

• In Kauffman’s originary approach ‘traits’ can be proteins or genes, each of which can assume different “configurations” or “values” (alleles). An organism is characterized by N traits, each of which assumes a given value. A mutation is nothing else than a change in the value of a trait (allele), and therefore to explore the effect of single mutations on the organism fitness we have to change the values of the traits, one at the time. A crucial property of the traits is 'epistasis': when a mutation is introduced, it normally happens that the effect on the organism’ fitness depends on the values of other traits.

From Bounded Rationality to Behavioral Economics

From Bounded Rationality to Behavioral Economics

• “The assumption that each gene contributes to overall fitness independently of all other genes is clearly an idealization. In a genetic system with N genes, the fitness contribution of one or another allele of one gene may often depend upon the alleles of some of the remaining N-1 genes. Such dependencies are called epistatic interactions “ (Kauffman, 1989, p. 539).

• Kauffman shows that as the epistatic interaction grows, the number of local optima increase, and an organism, which is affected by mutations, once it has reached a local optimum, may remain trapped.

From Bounded Rationality to Behavioral Economics

• Call K the average number of genes that contribute to the fitness variation of the organism when a mutation occurs, i.e. the average number of epistatic interactions. K may vary from K=0 (total independence) to K=N-1 (total interdependence). In the first case, (K=0) the effect of a mutation on the fitness depends solely upon the single gene that is affected by the mutation; therefore by comparing the different effects of different mutations on the same gene, we can discover the allele that produces the higher increase on the fitness. If we sequentially do the same comparison on all the genes, we can discover for every gene the alleles that make the best contribution to the fitness; therefore we can increase the fitness of the organism until his maximum value, acting on every gene independently.

• This means that an organism with zero epistatic interactions may achieve an optimal configuration in response to a sequence of random mutations.

From Bounded Rationality to Behavioral Economics

From Bounded Rationality to Behavioral Economics

• The evolutionary justification of Rationality

• Experimental economics have demonstrated that human decisions display systematic deviations from fully rational ones, and that in many cases ‘errors’ persist even when the rational solution has been explicitly presented to the subjects.

• This happens both in individual and team decision-making. In the case of teams and organizations, it may happen that systematically erroneous decisions are made by organizations, and that they remain trapped in sub-optimal routinized strategies, which are not changed even when they are highly sub-optimal. Levinthal and March (1993) single out a number of ‘traps’ into which an organization may fall during the process of organizational learning;

• in the field of evolutionary approach to organizations, Nelson and Winter emphasize that market mechanisms may not be able to select the best organizational structures and, again, inefficient firms may survive in the long run. Kauffmann’ NK model gives a vivid demonstration of these processes.

From Bounded Rationality to Behavioral Economics

• Violations of rationality are recognized as systematic, not random variations from the predictions of the theory of expected utility. A number of experiments (Camerer) show in fact that very few subjects, once the biases connected to their choices are shown to them, adjust their behavior.

From Bounded Rationality to Behavioral Economics

• From rationality to intelligence

• When the limits to the human capacity for mental calculation were experimentally demonstrated, it became clear to Simon that this ability was an aspect - an important but not unique one - of the mind’s more general capacity to manipulate symbols and to create mental models of reality.

• The awareness that human decision-making can be understood only if mental activities are viewed as symbolic manipulation urged the author on the need to build up a bridge to psychology. Simon's research shifted to a different version of the problem, subjecting the various mental abilities essential to explain human actions - memorisation, categorisation, judgement, problem solving, induction - to increasingly intense experimental scrutiny.

From Bounded Rationality to Behavioral Economics

• Both bounded rationality and behavioral economics approach holds that the crucial aspect of the decision-making process is the ability to construct new representations of problems.

• The evolution of the analytical tools and the experimental outcomes is shifting the focus from the coherence/incoherence of choices to the representation and editing of problems

• How mental models used by individuals and institutions to frame problems are constructed is becaming a crucial issue to be addressed by the decision theory, an issue that will yield a better understanding of human innovative activities within institutions.

From Bounded Rationality to Behavioral Economics

• With the weakening of the Friedmann approach, the cognitive characteristics of the choice processes are attaining a central position again.

• To follow such a course implies to inextricably connect psychology and economics: a demanding and relevant task, which raises questions on one of the fundamental statutes of the neoclassical construction of economic theory, Lionel Robbins’ idea that economy is a normative science, whose tenet is the logics of the means and ends.

From Bounded Rationality to Behavioral Economics

• Kauffman NK model suggests that biases and anomalies may be explained as sub-optimal peaks in a rugged landscape. Anomalies are driven by the process of categorization which induces individuals to adopt solutions which are locally stable, albeit imperfect. This formalization may aid understanding of why mental models and “ideologies” persist over long periods of time with remarkable stability when they have been largely falsified, and of what processes induce individuals to discard previous theoretical approaches. Re-definition of the categories constituting the building blocks of a solution requires a complex process may be similar to what Popper calls “critical thinking”.

• Rationality therefore emerges essentially as the capacity to get rid of our prejudices.