Social sciences are moving to the next stage. One ofthe promising methodologies is agent-based computer simulation. In a series ofworkshops on Agent-Based Ap- proaches in Economic and Social Complex Systems (AESCS), ground-breaking studies of complex economic and social phenomena using computational approaches are being presented and discussed. This volume contains papers selected from pre- sentations at the AESCS '02 held at the University ofTokyo, Japan, on August 16, 2002. The workshop was the second in a series ofPacific Rim activities in interdis- ciplinary areas of social and computational sciences, the first workshop having been held in Shimane, Japan, May 21-22, 2001. The objective of AESCS workshops is to promote worldwide multidisciplinary activities in multiagent computational economics, organizational science, social dynamics, and complex systems. AESCS brings together researchers and practitio- ners from diverse fields, such as computer science, economics, physics, sociology, psychology, and complex theory, in order to understand emergent and collective phenomena in economic, organizational, and social systems.AESCS also provides an opportunity for discussion ofthe effectiveness and limitations ofcomputational models and methodologies for the social sciences. The second AESCS workshop focused on the importance ofcumulative progress in agent-based simulation in the social sciences through discussions ofcommon tasks, standard computational models, replication and validation issues, and evalu- ation and verification criteria for the results.