6.1. Introduction

The first reference to a microsimulation model in social sciences goes back to studies from the American economist Orcutt and his team in the 1960s [ORC 57, ORC 61]. His objective was to provide national forecasts on the evolution of employment in different economic sectors. Disapproving the traditional macroeconomic models, he developed a model at the household level, then aggregating the results in order to produce forecasts at an aggregate level. The necessity of considering basic entities in order to understand the form and evolution of observable organizations and aggregates at a higher level is an old subject of debate in several disciplines such as ecology, biology, economy and sociology. The theoretical debate is old but the increase of computing power has given it a new perspective. In addition, with artificial intelligence methods, it is possible to formalize the behavior of "agents" that have autonomy, adaptation and interaction capabilities [FER 95] and these new perspectives have led to the development of the use of simulation in all social sciences [BAL 00, GIL 99].

A general label for this kind of models is "individual-based". Depending on the disciplines and applications, the term can mean "individuals" of extremely diverse nature (cells, particles, living beings, countries or planets). On the other hand, the term microsimulation mostly refers to basic entities that make sense in economy, demography and sociology, such as the individual, household, family and firm. It is also widely used in the...