Economia Politica. Rivista di teoria e analisi Economia Politica. Rivista di teoria e analisi
Non-Technical Summary
Homo economicus, the central figure of economic theory, is described as a completely rational egoist who maximises a utility function defined by stable preferences on all possible objects of choice. Moreover, his/her preferences are exogenously given and are not influenced by economic behaviour.
In recent years, as a consequence of the difficulties that the theory finds in the explanation of certain important economic and social phenomena, economists have questioned whether this subject is really useful for the objectives of their studies. Therefore, it is helpful to ask if homo economicus, after being sovereign for many years, is in danger of being deposed. If this is the case who is going to take his place? In order to answer this question we need to examine the elements that characterise the subject we wish to represent. From the considerable empirical evidence available it is possible to find aspects that contradict those taken for granted by the economic ideal-type. In fact, there emerges an individual whose preferences seem to be influenced by the socio-economic context in which he/she is involved and whose actions are triggered by motivations that are more complicated than those dictated by mere self-interest, such as altruism, desire for justice etc.. In addition he/she is limited in his/her choices by imperfect cognitive capacities. As a consequence any representation of this subject should be able to consider three principal aspects: the endogeneity of preferences, egoistic and non-egoistic motivation and limited rationality.
This paper is concerned with the endogenous change of preferences, even if one of its objectives is to argue that the separate analysis of each of these aspects is not useful for a global understanding.
The passage from the hypothesis of exogenous preferences to that of endogenous preferences can be defined as capable of taking into consideration the effect produced by certain economic and social variables on individual preferences. This paper examines the effect of past consumption in the process of preference formation. More precisely, the scheme of interpretation used here considers the effect of past consumption on individual preferences both in the case in which the individual is not able to perceive the utility to be obtained from the different alternatives of consumption, not knowing exactly his/her tastes, and in the case in which the utility changes in relation to past choices.
This scheme leads us to examine the most important habit forming models and the role played by the endogenous preferences in case based decision theory. The analysis of habit forming models takes into account two principal factors: i) the methodology used to represent the relation between present and past consumption; ii) the type of rationality that characterises individuals. These aspects do not only represent the lines that permit us to classify different models, but are also the themes that this paper tries to interpret in order to understand the role that they play in the concept of endogenous change of preferences.
The distinction between models based on the change of consumption capital and models based on the change of preferences is considered as a pure methodological question. This, in contradiction with the version that has prevailed in the past that used to counteract these models on the bases of their content: on one hand models that were interpreting habits as a change of preferences and on the other models that, using the idea of consumption capital, were the change of preferences.
As far as individual rationality is concerned, habit forming models can be classified into two categories: myopic and rational habit forming models. According to the first (Stone 1954, 1964, Storz 1956, Gorman 1967, von Weiszacker 1971) economic agents are unable to predict their own future tastes because they are unable to recognise the effect of past consumption on the evolution of their tastes. As a consequence they can not formulate temporally consistent consumption plans. Therefore, there is a divergence between individual choice and individual welfare that generates serious problems for some of the fundamental principles of neo-classical economics such as consumer sovereignty and paretian efficiency. Besides, as showed by Benhabib-Day (1982) and Day (1986), the hypothesis of myopic habit forming may generate chaotic behaviour.
The models belonging to the second category, on the contrary, assume that economic agents are perfectly aware of their own process of taste formation.
The passage from myopic habit forming to rational habit forming is to a certain extent settled by von Weiszacker's paper (1971) in which the author shows the existence of long run utility functions that are free from the influence of past consumption. After having discussed the very restrictive hypotheses that ensure the existence of these functions, the most important rational habit forming models will be examined. Here individuals are able to formulate temporally consistent consumption plans and this permit us to maintain the validity of the principal results of neo-classical theory. Even if, it is necessary to say that the hypothesis of endogenous preferences adds new causes of non-convexity that make the hypotheses that guarantee the validity of those results more restrictive.
Moreover, these models refer to very unreal hypotheses on individual rationality, extremer than those considered in the theory of rational expectations, that at least permit non-systematic errors. In fact, here individuals are supposed to have perfect information on present and future prices and on the availability of goods, in addition they know the effect of any present choice on future preferences and are able to know the level of utility that they will obtain with any possible preference ranking. Some recent models (Orphanides-Zervos 1995, 1998), assume individuals that are aware of their process of preference formation, but unable to foresee the exact impact that the choice to consume a potentially addictive good will have on their future welfare, show that individuals can make choices that they will come to regret.
The idea that individuals are unable to foresee their future tastes is confirmed by abundant empirical evidence. On the other hand if the process of preference formation is perfectly known by the individual from the beginning, is it still possible to speak of endogenous preferences?
The correct anticipation of any outcome from every current choice does not permit us to take into serious consideration the learning process that permits to foresee those outcomes. Moreover it does not permit us to conceive an other type of endogenous change of preferences deriving from the fact that individuals could not know their tastes hitherto inexperienced goods.
This kind of endogenous change of preferences finds its place in the case based decision theory in which the phenomena of changing references is pervasive. This theory considers both the relation between past consumption and present choices and the relation that could emerge between the preferences of any individual and the preferences of other individuals with whom he/she interacts. In this way the comprehension of important socio-economic phenomena becomes easier. However, this theory needs more development which among other thing would bring it to a more precise specification of the similarity function. On the other hand the neo-classical theory can hardly present a general scheme of interpretation of endogenous preferences, at least until the extreme hypotheses above discussed are introduced, as they do not permit us to interpret the case in which individuals do not know their tastes and the case in which their tastes are influenced by the tastes of other persons.
MARIA DE PAOLA is temporary professor of economics at the Università degli Studi della Calabria, Dipartimento di Economia Politica. Via Bucci, Arcavacata di Rende (CS)
m.depaola@unical.it
Go back to Home Page 
Go back to Index
Redazione: econpol@spbo.unibo.it
Webmaster: lrossi@spbo.unibo.it