Econometry and game theory

Economic theory seeks to understand the importance of the motivations underlying the decisions of individuals living in society. Reasoning within the framework of models, theorists formulate hypotheses about these motivations, from which they deduce, through logical analysis, the rational behavior of agents representative of an economic system, such as consumers, producers or various state institutions. The ideal representation of the functioning of market economies thus developed is intended to be descriptive, generating explanations and predictions of human decisions. But do economic agents, with their limited cognitive capacities, behave as theorists have imagined? The results obtained in models must be compared with observed facts to test their validity. Only if these tests are positive can we envisage using formal analyses as a means of acting on reality, in particular as a decision-making tool in an economic and social environment.

The aim of econometrics is to put theoretical models to the test, by applying statistical methods to observations of the behaviour of economic agents involved in market exchange situations. The approach employed consists in constructing empirical relationships between variables suggested by theory, and relating to the motivations and decisions of individuals. Measuring the parameters introduced into these relationships to express their direction and strength, using appropriate statistical techniques, will enable us to assess the fit between the thought experiments developed by theorists and the reality of how an economy works. The data revealing the values of the variables of interest can be generated by field surveys, recording motivations and decisions not influenced by the modeler-econometer. These data have the advantage of incorporating the constraints actually faced by economic players, but they are very often costly and/or difficult to obtain. For this reason, the econometric approach is still based on experimental data, obtained under controlled laboratory conditions, which enable a more detailed assessment of the determinants of individual decisions. These data are generally used as a first step before carrying out an in-depth test on field data. The nature of the variables recorded in the data has consequences for the complexity of the empirical relationships dealt with in econometrics. The individual decisions whose motivations we seek to understand essentially express choices between different options; these choices are naturally translated using qualitative variables and related to their motivations in models belonging to the field of qualitative variable econometrics.

The search for the motivations behind individual decisions has to take into account the fact that economic agents are most of the time immersed in game...