Methodological Overview
To better assess the potential of the Neuroeconomic Theory methodology, it is important to understand its conceptual differences with traditional economic theory.
Modern microeconomics is based on two bodies of theoretical research. Decision Theory defines a series of axioms that govern individual decision-making and provides (sometimes unique) utility representations that satisfy those axioms. Game Theory characterizes equilibrium concepts that govern strategic interactions between players who are endowed with these utility functions. Among all axioms and equilibrium concepts, the most standard way of modeling economic agents is to rely on the so-called rational paradigm, an approach that imposes extreme sophistication on individuals (such as perfect memory, perfect capacity to process information, perfect anticipation of othersÂ’ behavior or stable preferences, etc.).
The Experimental Economics literature demonstrates that predictions made under the rational paradigm are often difficult to reconcile with observed behavior, both in individual decision-making problems and in multi-person games. Thus, despite their being excellent first approximations, the standard utility representations and equilibrium concepts need to be refined if we want to understand and predict choices. In the last decades, both Decision Theory and Game Theory have accepted this challenge. New axioms (such as uncertainty aversion or set-betweennes) and new concepts to model strategic interactions (such as quantal response equilibrium or cognitive hierarchy) have been devised to account for some unexplained phenomena. More recently, Behavioral Economics has also started to provide new useful models of decision-making based on findings in Psychology. Note, however, that these models generally presuppose the existence of utility functions without deriving them from axioms, which constitutes a huge methodological difference with respect to Decision Theory.
We see one drawback in the existing literature. Up to now, researchers have only been able to observe choices, so theories of decision-making are built by drawing inferences from those choices. The preference-based approach poses axioms on (non-observable) tastes, while the choice-based approach defines properties of (observable) choices. As long as the rational paradigm holds, consistent choices correspond to rational preferences. Then, both approaches are equivalent and it is irrelevant that preferences cannot be observed. However, when the rational paradigm does not hold, there is not such clear mapping. If the researcher focuses on choices and renounces to draw inferences about preferences, theories are bound to be limited to the situations in which all possible choices are observed. If, instead, the researcher draws inferences about preferences from the observation of choices, the modeling of these preferences becomes speculative.
What Neuroeconomic Theory proposes is to use not only information on choices but also on the brain mechanisms that lead to those choices. This is now possible thanks to the recent improvements in the designs and techniques to measure brain activity. Thus, unlike the previous theories, this approach does not have to choose between studying choices and studying preferences, nor it is bound to speculate on how to model preferences in the latter case.