This book is in many ways well-conceived. But it has many errors. I would recommend it to a graduate student or colleague only with reservations. In several places Azariadis comes across as downright sloppy. My edition is a first edition, so if there are any later editions they might be better. It is too bad, because the basic idea of the book is good; it could have been much more valuable. But a methods book that one knows has lots of mistakes is almost useless, since one doesn't know what to trust. ( )