Social Learning About Consumption

Abstract

This paper applies a social learning model to the optimal consumption rule of Allen and Carroll [Macroeconomic Dynamics 5(2001), 255–271] and delivers convincing convergence dynamics toward the optimal rule. These findings constitute a significant improvement over previous results in the literature, in terms of both speed of convergence and parsimony of the learning model. The learning model exhibits several appealing features: it is frugal, is easy to apply to a various range of learning objectives, and requires few procedures and little information. Particular care is given to behavioral interpretation of the modeling assumptions in light of evidence from the fields of psychology and social science. Our results highlight the need to depart from the genetic metaphor, and account for intentional decision-making, based on agents' relative performances. By contrast, we show that convergence is strongly hindered by exact imitation processes, or random exploration mechanisms, which are usually assumed when modeling social learning behavior. Our results suggest a method for modeling bounded rationality, which could be interestingly tested in a wide range of economic models with adaptive dynamics.

Is the market really a good teacher?

“…By contrast, social learning allows the agents to parallelize the evaluation of the available strategies, so that the larger the population, the quicker the evaluation process. Allen & Carroll (2001) and Palmer (2012) illustrate this difference within the simple framework of the buffer-stock consumption rule; see also Salle & Seppecher (2016).…”

“…However, GAs are not exempt of limitations. Their operators do not always find an easy economic interpretation (Chattoe 1998, Salle & Seppecher 2016. Most importantly, because they have been initially developed to find optima in complicated static problems (Holland 1975), they have been used in economics as a way for agents to learn how to maximize their profits or utility functions, and the focus has been put on the conditions under which agents end up coordinating on the optimal state of the model under GA learning (Arifovic 1990).…”

“…We neither model any counter-part of the cross-over operator as we find little interpretation for it in economics, see Salle & Seppecher (2016) for further discussion.…”

Coordination through social learning in a general equilibrium model

“…Those parameter values are in line with standard calibrations of GAs. As discussed in Section 3.3, parameter ˇ is particularly relevant in ABMs derived from GE models, where performance functions are mainly chosen for their analytical properties in optimization problems, rather than for their ability to facilitate learning by agents with bounded rationality -see also Salle and Seppecher (2016) for a related discussion. Consequently, we adopt ˇ = 4 in the simulations.…”