Advanced Econometrics

Rating:
Summary: Top choice
Review: Both comprehensive and well-structured this book proves indispensable for anyone delving into the realms of econometrics. Starting from classical least squares the author guides the reader to time series analysis, gls, nonlinear simultaneous equations models up to qr and tobit models. If formulas are a necessary condition for a good study book in econometrics, the clear language of this book fulfills the sufficient condition for any book in this category.

Rating:
Summary: The best book I have read for this subject.
Review: I strongly recommend this book as a textbook for this subject. After reading and using this book, I gained a solid knowledge of econometrics.

Rating:
Summary: Excellent reference, but not an easy read!
Review: This book is justly considered a classic. It has been around for many years, and with some reasons. It provides a very rigorous treatment of many fundamental concepts in cross-section econometrics, such as linear and non-linear models, M-estimation, maximum likelihood, limited dependend variable models. It also has one of the best and more rigorous yet accessible treatments of basic asymptotic theory (the examples and countexamples in this section are uncommonly good). Amemiya is very very rigorous, and this a book where typos and sloppiness do now dwell. Overall, it is not an easy read, though, unless you have a very strong math/stat background, or you are genius. One thing I always liked about this book, indeed, is the very honest title, ADVANCED econometrics, not "Introduction" to econometrics. This book is mostly used as part of the reading list in second-year PhD courses in cross-sectional econometrics. I don't think it would be a good choice for a first year course. But if econometrics is a serious component of your professional life you will be happy to have Amemiya around, and you will keep reaching for it, once in a while. It is a bit too techinical, difficult, and dry to be my ideal textbook, but it is outstanding nonetheless. I find the chapters on asymptotic theory and limited dependent variables particularly well written. On the minus side, it is now a relatively old book, and you will find here many obsolete technical tools, as well as the absence of many important and modern techniques. In particular, note that you will NOT find anything here about nonparametric and semiparametric techniques, panel data, time series. There are many (very short) empirical applications scattered around the book, but most of them are (necessarily, given the publication year) very very obsolete.
Overall, still a great book highly recommended for people who are into advanced econometrics. But if you want an introduction to cross-section econometrics, you may want to look at other textbooks such as Greene, Ruud, Davidson-McKinnon, Hayashi (more time-series oriented) and especially the "graduate" Wooldridge (Econometric Analysis of Cross-Section and Panel Data), which in my humble opinion is currently the very best option around.
P.S. Harvard University Press also has merit of printing Amemiya on top-quality paper and choosing a very nice format for the book. It's a pleasure to browse its pages!

Rating:
Summary: Jackpot/Bingo/Royal Flush
Review: This is THE bible for understanding most empirical econ papers out there. I wished I had found it earlier.