A challenge to many social and economic policies is to tackle the interactive processes between actors and institutions involved, bound to a geographical, economic, and cultural context over which the social mechanisms that policies want to address diffuse. Agent-based modeling is a method to investigate these dynamics of collective action through computer simulations, addressing the interaction of virtual agents mimicking social actors within artificial societies. Synthetic populations are artificial societies built on the integration and harmonization of datasets that can reproduce detailed information on the micro level of agents not provided by the original data for different reasons. The chapter offers a comparison of the main algorithms for the extraction of synthetic populations and illustrates the contribution of research infrastructures to the field. We conclude with three case studies reporting the experience of the authors and conclude with considerations on the role of research infrastructures for policy research.