Rationality, expectation, belief, preference
Rationality is defined as a property (either of a decision or of an action) in relation to a given model of rationality. Therefore, there is no absolute rationality, or itself, but only a contextualized (or relative) rationality for a given model of rationality. Models of rationality can be either explicit, often formalized (as in the model of economic rationality within neoclassical economic theory, known as homo œconomicus), or implicit (which is the case of behaviourist models or, more recently, evolutionary models, inspired by the biological paradigm). If we note with a model of rationality i, then a decision j taken into context of the given model of rationality can be of three categories (all related to ):
- rational decision – when it is validly inferred1 from, noted with ;
- irrational decision – when it is invalidly inferred from , noted with ;
- a-rational decision (or logically inferred) – when it has no relationship with type inference , noted with .
Of course, what appears as irrational or a-rational in relation to one model of rationality can be rational in relation to another model of rationality and vice versa. Thus, if we note a decision k, in relation to a model of rationality q, , we can have any of the following situation: the noted logical constant ⋈ has the meaning ‘is compatible with’ ( means ‘is incompatible with’), and the logical constant → has the meaning ‘validly implies’ (↛ means ‘does not validly imply’):
- formalization of the rational decision:
- formalization of the irrational decision:
- formalization of the a-rational decision:
Human behaviour may take place under four ‘authorities’: under the authority of rationality, under the authority of expectation (subjective), under the authority of belief, and under the authority of preference.
Of course, this typology is rather academic because, in fact, the four ‘authorities’ can be combined two by two or (which is, additionally, the most common case) co-exist in different weights, weights which, in turn, change because of the experience, as a result of environmental pressure, or as a result of both causes.
Behaviour led by rationality
It is a behaviour based exclusively on inferring the decision from the pre-accepted model of rationality (in the logically way described earlier). It is the behaviour considered in neoclassical economic theory2 (under the empire of which the efficient market hypothesis (EMH) was also constructed), which considers the human individual as having a hyper-rationality in choosing the optimal decision (i.e., the decision which extremizes an objective function under given restrictive conditions or, equivalently, minimizes the opportunity cost under those conditions). The best-known logical model of this type of behaviour is the (mathematical) homo œconomicus model. Over time, this abstract model has been relaxed in many ways (including here Simon’s concept of bounded rationality, as well as other formal relaxations), but rational behaviour remains predominant.
Behaviour led by expectation
Behaviour led by expectation is based primarily on behaviourist studies3 (behavioural economics), and the economic decision is based on a combination of rationality and emotion (fear and various aversions to risk, to loss, etc.), considering even that ‘realistic’ rationality contains emotion.4 Behaviourism accepts the important role of the individual’s past (experience) as an economic actor. In essence, an expectation is subjective (and, therefore, strongly idiosyncratic) and is manifested by the allocation of Bayesian probabilities5 events (occurring under conditions of uncertainty, which are logically equivalent to the absence of any information that would associate risks with the events concerned and which, by this allocation, would change the probabilities associated with them). The phrase ‘rational expectations’, extremely common in the literature, is quite ambiguous because being subjective, the expectation cannot be rational – because if it were rational, it would be associated with a model of rationality, so it would not be subjective (namely, free, based on free will or, at least, idiosyncratic desirability). Probably, those who use this phrase intend to say ‘rational anticipations’, which is something else entirely.6
Behaviour led by belief
Regarding behaviour led by belief, this is intended to be a more comprehensive theory related to that of behaviour led by expectations, in the sense that it is considered (in an axiomatically way) that the individual is ‘endowed’ with a (complete) system of beliefs about economic phenomenology, beliefs which underlie its decisions and actions (including, of course, religious beliefs, ethical values, etc.). This theory is closely linked to the theory of justification and, as a result, is more a post-factum theory than one which would constitute an operational predictor on the economic market (or, more specifically, on the financial market). Belief (put into relation with motivation, intention, and other psychological ‘objects’) constitutes one of the arguments of behaviourism (and evolutionism) for shaping the economic decision. In the specialty literature, the following categories of beliefs are considered:
- overconfidence in their own judgements (i.e., self-attribution – luck is attributed to their own talent; hindsight – ex post, individuals always find that they ‘predicted’ well);
- optimism, desire-oriented thinking, and greed;
- representativeness, which leads to an underestimation of the sample size (there are two failures of representativeness, especially in assigning probabilities: (1) neglect of the base rate; (2) neglecting the sample size);
- conservatism: leads to overestimation of the sample size;
- perseverance of belief: inertia in renouncing the formed belief;
- anchoring: availability errors – information that is no longer available in memory (Barberis & Thaler, 2012).
Behaviour led by preference
Behaviour led by preference is, in a way, a wrapper of all three types of behaviours described earlier, in the sense that the individual (actor or economic agent) externalizes both his model of rationality and idiosyncratic expectations and beliefs, through the preference. So, in a way, the preference is a synthesis (not aggregation, because the three bases of behaviour are, in principle, incommensurable two by two) of the whole ‘economic personality’ of the individual concerned. Although the authors did not agree on a ‘list’ of preferences (which, as mentioned earlier, are sometimes considered beliefs), in principle, there are three such preferences: (a) risk aversion; (b) aversion to loss – see Prospect Theory (Kahneman, 2012); and (c) aversion to ambiguity. In what follows, we will focus only on this type of economic behaviour.
The concept of economic preference
In economic theory, economic preferences mean anchors of behaviour of the individual (actor/economic agent) which are, in the most general sense, predictors of the behaviour concerned. There are two basic characteristics of economic preferences: (a) they are considered given, namely, they represent variables (in fact, rather, constraints) exogenous for the economic analyst;7 and (b) they are idiosyncratic, namely, specific to each specific individual (or, as the case may be, to each representative agent, if modelling based on representative agents is used), which means that they are incommensurable between any two individuals. More precisely, economic preferences act on economic behaviour analogously with the action of the principles assumed (and publicly announced) by each individual or, where appropriate, by each class of individuals considered homogeneous within that class and from the perspective of the accepted classification criterion.8 In this context, we can define economic preference, at the individual level, as the maximum probability of choice in conditions of uncertainty. The justifications for this content of economic preference are as follows:
- in connection with a choice (or equivalent, with a decision between at least two measurable alternatives9 between them), ther...