The Explanation of Inequalities through Generative Models: a Contribution to the Understanding of Social Mobility from the Analytical Sociology's Point of View

The aim of this work is to represent social mobility as a macro-phenomenon generated from a lower level of analysis. Despite Boudon has put a great deal of effort into persuading sociologists of the importance of this explanatory strategy, research on social mobility found however rather difficult formalizing and testing this kind of “bottom-up” theories. It’s with this in mind that the present work intends to address the following purposes: giving a pivotal role to a micro-founded rational theory of social action in the explanation of the phenomenon; gathering various partial mechanisms found in previous research in a more general theory; trying to reconnect two different levels of analysis into the same theoretical framework (the macro-level of social mobility, seen as an emergent phenomenon, and the micro-level of intentional, interdependent, competitive actors); representing social mobility as a dynamic process continuously unfolding through time. To do this, I implement my model by means of agent-based simulation. This technique is indeed well-suited to translate a “bottom-up” theory of social mobility into a non-equivocal algorithmic language, in order to study more closely how a system dynamically emerges from complex interaction between intentional agents. I sequentially introduce the hypotheses into the model to provide, although deductively, better evaluations of the explanatory power of each mechanism. Then, I compare the outcomes generated by the model with the empirical ones (data from a large national survey carried out during 2003 by the Italian national bureau of statistics). Finally, I assess the model on the basis of its ability to reproduce the empirical mobility table, the bootstrapped distribution of its samples, and trying to account for a variety of features of the phenomenon.